Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. URINE NITROGEN IN HILL COUNTRY PASTURE SOILS A thesis presented in partial fulfilment of the requirements for a Doctor of Philosophy in Soil Science, Massey University W.M. SAMAN OEEPAL BOWATTE 2003 11 ABSTRACT In New Zealand the traditional way of building up nitrogen (N) fertility in pastures has been to apply phosphorus (P) fertilisers to provide adequate soil fertility for legume growth, which then provides N through biological N fixation. However, the marked responsiveness of hill pastures to N fertiliser indicates that this traditional approach may be placing a serious constraint on hill country production. At the same time, there is concern that the resulting elevated soil P levels may pose some environmental risk. Although the importance of soil N availability to hill country pasture production has long been recognised, there is surprisingly little infonnation available on N cycling in hill country pastures. This is because the limited research funding available has been directed mainly at detennining the requirements for P and suI fur (S) fertilisers, which have constituted the bulk of fertiliser expenditure in hill country. In order to develop best practice in the use of fertiliser N in hill country, infonnation is required on N flows in the soil-plant-animal system on the contrasting topographic land units that comprise hill pastures. The role of grazing animals and particularly the N transformations associated with urine patches are very important components of these N cycles. In this study, two field experiments were conducted at contrasting locations in North Island hill country pastures to investigate the fate of urine N. These field experiments were then followed by a laboratory incubation experiment that sought to clarify the effect of soil properties on subsequent transfonnations of urine N. The experimental results were then used, together with data from the literature, to model the N cycle for hill country pasture. In addition, to assess the N availability in hill pastures, an in situ N measurement technique using ion exchange resin membrane spikes was developed and evaluated. The first preliminary field experiment was carried out at the AgResearch Grassland hill country research site in Waipawa, North Island, New Zealand from 09 June 1 999 to 29 October 1 999. The major soil type was Waipawa Stony Silt Loam (PaIlic Soil). Three synthetic urine treatments (0, 200, 400 kg N/ha) were applied in a randomised complete block design and the experiment was repeated in a flat campsite and a steep site. At 1 III day after urine application (DAUA), the increase in the soil mineral N pool was close to or greater than the quantity of added urine N. The dominant form of mineral N throughout the experiment was NH4 + -N. This suggested that nitrification rates were low and that leaching losses ofN03--N would therefore be low. Only 1 8-27% of the urine N was recovered by the pasture. Estimates of the loss of urine N by ammonia volatiIisation were large, ranging from 2 1 -34% of added urine N. At the end of the experiment ( 1 42 DAUA), 34 -50% of added urine N appeared to have been immobilized into complex organic matter. The second field experiment was carried out at Ballantrae AgResearch hill country research station from 1 4 July 2000 to 1 2 December 2 000. The soil was N gamoko Silt Loam (Brown Soil). Three different rates (0, 280, 5 6 0 kg urine Nlha) of synthetic urine were applied as treatments and the experiment was repeated as a randomised complete block design on a flat campsite and a steep slope. Shortly after application, recovery of urine N as soil mineral N was greater than 1 00% ( 1 1 3 - 1 4 1 %) in the flat site. This increase in mineral N corresponded to a decrease in mineralisable N, suggesting organic matter mineralisation after urine application. During the first month after urine application, N&+-N was the dominant form of mineral N, but during the second month, N03--N was the dominant mineral N form. At the end of the experiment (88 DAUA), urine N recovery as mineral N was very low, ranging from 0-3 % . The rate of nitrification after urine application was higher in flat campsites than in steep slopes. Soil N03 --N levels in the 0- 1 0 cm soil depth in urine- treated plots at both sites decreased considerably between 30 and 45 DAUA. A simple model developed in Microsoft Excel suggested that substantial leaching of urine N (9- 3 3 % of added urine N) was likely to have taken place. Urine N recovery by herbage in this experiment was low ( 1 - 1 4% of added urine N ). Estimates of the loss of urine N through volatilisation were large, ranging from 24-5 1 % of added urine N . At the end of the experiment the amounts of urine N estimated to have been immobilised into the soil organic matter ranged from 8-57% of that added. A laboratory incubation experiment was conducted using four soils collected from the flat and steep sites of the field experiments at Waipawa and Ballantrae together with three other soils collected from lowland sites (Kairanga silt loam, Karapoti silt loam and Manawatu sandy loam (Fluvial Recent Soils)) that had received substantial quantities of IV excretal N over several years. Field moist soil, equivalent to a weight of 1 00 g of dry soil, was placed in each of 3 6 small plastic cups for each soil type. Urine was collected from four cows during m ilking two weeks before the experiment. Urine was applied to 1 8 cups of each soil at the rate of 6 mL of urinel l 00 g dry soil (40 mg urine Ni l 00 g dry soil) . The remaining 1 8 cups were used as controls . No solution was added to the control cups. In contrast to the field experiments, there was little evidence of an initial priming effect, with mineral N levels 3 DAUA ranging from 64-8 1 % of added urine N. Nitrification rates were highly variable (0. 3 to 1 8 .3 Ilg N03--N/g soil/day) across the seven soils. All lowland soils had higher nitrification rates than hill soils, while those soils collected from campsites had higher nitrifi cation rates than soils collected from steep slopes . Although nitrification could account for most o f the disappearance o f soil NH4 + -N from 3-45 DAUA, it was evident that mineralisable N and soil microbial biomass N also increased after urine application. A simulation model of a hill country N cycle developed in Microsoft Excel confirmed the importance of urine N in hil l country pastures. The model indicated that N outputs in animal products, together with losses through ammonia volatilisation and leaching from urine patches were likely to exceed the N inputs to hill pastures by legume N fixation, non symbiotic fixation and atmospheric deposition. This may be the reason for the observed high N responsiveness in hill country pastures. Pasture utilisation and excretal distribution in the paddock were the most important factors influencing the overall N balance in the paddock. M ore work is required to obtain information on these parameters in hill country pastures. The in situ N measurement technique using ion exchange resin membrane spikes proved to be a useful approach to monitoring the continuous changes in soil mineral N in the field experiments as well as in the incubation experiment. Resin spikes were able to detect apparently real differences in the availability of soil N - even when the standard 2 M KCl extraction could detect no differences. The potential of resin spikes to detect spatial variability in soil N status was also demonstrated. v A simple model developed in Visual Basic in Microsoft Excel to simulate the N adsorption by resin spike in soils demonstrated that soil moisture, soil temperature, soil N concentration and the time the resin spike is in the soil are all major determinants of the amount of N adsorbed to resin in soil . Vi ACKNOWLEDGEMENT I would like to thank following people and organisations for their part in seeing this project through to completion. My supervisors, Professor Russ Tillman, Mr Andrew Carran and Dr Allan Gillingham for their supervision, valuable suggestions, constructive criticism, encouragement, support and friendship during my study. Dr Dave S cotter for interesting lessons of modelling. Associate Professor Nanthi Bolan for his warm welcome on my first day at Massey University and the great help in finding accommodation and settling us in. James Hanly, Bob Toes and Phillip Theobald for their enormous support for laboratory and fieldwork, proof reading the thesis and especially the friendship throughout the study. Associate professor Mike Headly, Lance Currie, Ann West, ran Furkett, Mike Bretherton , Ross Wallace, Glenys Wallace, Dr. Loga Loganathan, Hera Kennedy and Denise BrunskiIl for their continuous support . Massey University for granting a Massey University Doctoral Scholarship and Helen E. Akers Ph.D. S cholarship . Managers and technical staff at Ballantrae AgResearch research station and the Waipawa AgResearch research stations for allowing me to conduct field experiments and supplying necessary resources. All past and present post graduate students; Sena, Tin, Asoka, Lui, Aravin, Tony, Steven, Khan, Fabio, Jamie, Rita, Thabo and many others for providing friendly atmosphere at Massey. Mrs. Connie Cathcart and her family for introducing this beautiful country to us and the warm-hearted assistance and friendship for settling to kiwi life . I especially thank to my loving wife Deepa and two lovely daughters D inithi and Tharushi for patient, understanding, support, belief and love through times when I needed it most. Finally I would like to dedicate this thesis to my mother and father and all my teachers for their effort to bring me up to this level. Vll TABLE OF CONTENTS ABSTRACT ........................................................................................................ 11 ACKNOWLEDGEMENT .................................................................................... VI TABLE OF CONTENTS ................................................................................... VII LIST OF TABLES ........................................................................................... Xlll LIST OF FIGURES ......................................................................................... XVI LIST OF PLATES ......................................................................................... XXIV CHAPTER ! INTRODUCTION INTRODUCTION ...................................................... ........................ .................. 1 2.1 2.2 2 .2 . 1 2 . 2 .2 2 .2 . 3 2 . 2 . 4 2.3 2 . 3 . 1 2 . 3 . 2 2 . 3 . 3 2 . 3 .4 2 . 3 . 5 2 . 3 . 6 2 . 3 . 7 2 . 3 . 8 2.4 CHAPTER 2 LITERATURE REVIEW Introduction ......................................................................................................... 4 New Zealand hill country ................................................................................... 4 Topography .. . . . . . .. . . ......... .... . . . . . . .... ..... . .. . . ...... . . . . . . . . ..... ..... . . . '" . . . . . . . . . . . . . . . . . . . . . . . . . 5 Soil .......... . . . . . .. ................................. .. .... . . . . .. .. . . . . .... . . . . ..... .. ....... . ..... . .. . ....... .. ... 5 Pasture composition ... . .... . . ...... . ........ ........ . .... . .. ...... . . . ... . ........ . ....... . ..... . ... .. .... 6 Nitrogen use in hill country .. ... . .... . . . .... . ......... . .... . ..... .. . ............. . . . ..... . . . ....... . . . 7 Nitrogen balances in different topographic units of hill country .................... 8 Pasture N uptake ..... . ........ .. . .... . . ............... . . . .... . . ..... . . . . . . . . . . . . ............ . . . . . . ...... . . 1 1 N fixation .. . . ...... . . . .... . . . . ...... ...... . .. ....... ............... . . . ...... . ....... . ........ . . . .... ... . ...... 1 1 Non-symbiotic N fixation and atmospheric deposition . ..... .. .. . . ...... . .. . ....... . . 1 3 Pasture utilisation . . ....... .. . .... . ... . ............ . ... . ............ . . . ... . . ................. . ... . . . . . . . .. 1 3 N remaining in pasture litter. . . . ..... . ............ . ... . ........ . ..... . . . .. . ...... .... . .... ..... . . . . . 1 4 N in animal products . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4 Dung and urine N . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4 Above ground N balance in hill country . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6 Summary of literature review .......................................................................... 18 Vlll CHAPTER 3 FIELD INVESTIGATION OF THE FATE OF URINE NITROGEN ON A SUMMER DRY HILL COUNTRY PASTURE 3.1 Introduction ....................................................................................................... 20 3.2 Literature review ............................................................................................... 2 1 3 .2 . 1 3 . 2.2 3 . 2 . 3 3 . 2 . 3 . 1 3. 2 . 3 .2 3 . 2 . 3 . 3 3 . 2 . 3 .4 3 . 2 . 3 . 5 3 . 2 . 3 . 6 3 . 2 . 3 . 7 3 . 2 . 3 . 8 Nitrogen cycling through animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 Urine N . . . ......... . . . . . . . . . .. . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 N transformations in the urine patch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Urea hydrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Ammonia volatilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Nitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7 Mineralisation and immobilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7 Plant uptake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 0 Fixation t o clay minerals . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1 3.3 Materials and methods ...................................................................................... 41 3 . 3 . 1 3 . 3 .2 3 . 3 . 3 3 . 3 .4 3 . 3 . 5 3 . 3 . 6 3.4 3 . 4 . 1 3 . 4.2 3 .4 . 3 3 . 4. 4 3 .4 . 5 3 .4 . 6 3 .4 . 7 3 .4 . 8 3.5 3 . 5 . 1 3 . 5 . 2 3 . 5 . 3 3 . 5 . 4 3 . 5 . 5 3 . 5 . 6 Field site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1 Field trial design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Soil and plant sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Ammonia volatilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Results ................................................................................................................ 46 Rainfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Statistical data interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 7 Mineral nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Ammonium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Nitrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1 Mineralisable N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . 5 2 Pasture response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 4 Ammonia volatilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8 Discussion ........................................................................................................... 59 Urine N recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 9 Mineral N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Nitrification and leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Pasture response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Ammonia volatilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Mineralisation and immobilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 IX 3 . 5 . 7 Comparison o f urine N transformations at two sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 3.6 Conclusions ........................................................................................................ 68 CHAPTER 4 DEVELOPMENT OF ION EXCHANGE RESIN MEMBRANE SPIKES FOR CONTINUOUS MONITORING OF AVAILABLE SOIL NITROGEN IN HILL COUNTRY PASTURE 4.1 Introduction ..................... .......... . . . ............... ............................. . . . ...................... 70 4.2 Literature review .............................................. .... ...................................... ....... 7 1 4 . 2 . 1 4 . 2 . 2 4 . 2 . 3 Ion exchange resins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1 Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Ion exchange resin use in recent soil research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 4.3 Development of the ion exchange resin membrane spike ..... ......................... 77 4. 3 . 1 4 . 3 .2 4 . 3 . 3 4 . 3 .4 4 . 3 . 5 Experiment 1 : Preliminary assessment o f the ability of resin strips to adsorb mineral N from Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Experiment 2 : Assessment of variability with resin spikes . . . . . . . . . . . . . . . . . . . . . . . . . 78 Experiment 3: Optimization o f resin spike construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Experiment 4 : Optimization o f resin spike construction II . . . . . . . . . . . . . . . . . . . . . . . . . 80 Experiment 5: Evaluation of resin spike variability in soil . . . . . . . . . . . . . . . . . . . . . . . . . 8 1 4.4 N adsorption to resin membranes .............. ...... .............. ............ .............. ........ 82 4 . 4 . 1 4.4. 1 . 1 4.4. 1 .2 N adsorption to resin spikes over time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Experiment 1 : N adsorption from solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Experiment 2: N adsorption from soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 8 4.5 Resin spikes performance in field ............................................. . . ..................... 9 1 4 . 5 . 1 4 . 5 .2 Field experiment 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1 Field experiment 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 4.6 Modelling of in situ N adsorption to resin membrane spikes ........................ 96 4 . 6 . 1 4 . 6 . 2 4 . 6 . 3 4 . 6 . 4 4 . 6.4. 1 4 . 6 .4.2 4 . 6 .4 . 3 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 6 Basic equations for solute diffusion i n soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Model development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 00 Model output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 0 5 The effect of time o f burial o n N adsorption b y a resin spike i n soil . . . . 1 05 Effect of soil moisture on N adsorption by resin spikes . . . . . . . . . . . . . . . . . . . . . . . 1 09 E ffect of initial soil N concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 1 x 4 . 6 .4.4 Effect of temperature on N adsorption by resin spike in soi! . . . . . . . . . . . . . . . . . . . . l l l 4.6 Recommended procedure ............................................................................... 1 12 4.8 Discussion ......................................................................................................... 1 1 4 CHAPTER S FIELD INVESTIGATION OF NITROGEN DYNAMICS UNDER URINE PATCHES IN NORTH ISLAND HILL COUNTRY PASTURE 5.1 Introduction ..................................................................................................... 1 1 8 5.2 Materials and methods ..... ............................................... ................... ............. 1 19 5 .2 . 1 5 . 2 .2 5 .2 . 3 5 . 2 .4 5 . 2 . 5 5 . 2 .6 5 . 2 . 7 5 .2 . 8 5 .2 . 9 5 .2 . 1 0 Site description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 9 Field layout and soil sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 9 Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 20 Soil sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 1 Ammonia volatilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 1 Soil mineral and mineralisable nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 1 Dissolved organic carbon and dissolved organic nitrogen . . .... .. . . . .. . . . . . . . . . . . 1 22 Resin-adsorbed nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 22 Plant dry matter production and pasture nitrogen uptake . . . . . . . . . . . . . . . . . . . . . . . . . 1 22 Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 23 5.3 Results .............................................................................................................. 1 23 5 . 3 . 1 5 . 3 .2 5 . 3 . 3 5 . 3 .4 5 . 3 . 5 5 . 3 . 6 5 . 3 . 6. 1 5 . 3 . 7 5 . 3 . 8 5 . 3 . 9 5 . 3 . 1 0 Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 23 Statistical interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 25 Mineral nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 5 Ammonium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 8 Nitrate . . . . . . . . . . . . . . . . .. . . . . . . . .. . ..... . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 1 Perfonnance of resin spikes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 3 Relationships between 2 M KCI -extractable N and resin-adsorbed . 1 3 5 Mineralisable N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 7 Pasture response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4 1 Ammonia volatilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 44 Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. .. . ... .. . ... . . . . . . ... .. .. . . . . . . . . . .. . . . . . . . . . . . . . . . . 1 46 5 . 3 . 1 0 . 1 5 . 3 . 1 0. 2 Leaching model development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 46 Leaching model output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1 5 0 5 . 3 . 1 1 Nitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . 1 54 5.4 Discussion ......................................................................................................... 158 5 .4 . 1 5 .4.2 5 .4 . 3 5 . 4.4 5 .4 . 5 5 .4 . 6 5 .4.7 5 . 4 . 8 5 .4 . 9 Xl Urine N recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 8 Mineral N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . .. . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6 1 Priming effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . 1 62 Ammonia volatilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 64 Nitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 65 Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 65 Pasture response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 67 Denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 68 Immobilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 68 5.5 Conclusions ...................................................................................................... 1 69 CHAPTER 6 LABORATARY INCUBATION STUDY OF NITROGEN TRANSFORMATIONS IN HILL COUNTRY AND LOWLAND PASTURE SOILS AFTER APPLICATION OF URINE 6.1 Introduction ..................................................................................................... 1 70 6.2 Materials and methods .................................................................................... 171 6. 2 . 1 6.2.2 6 . 2 . 3 6 . 2 . 3 . 1 6 . 2 . 3 . 2 6 . 2 . 3 . 3 6 . 2 . 3 .4 6 .2 . 3 . 5 6 . 2 . 3 . 6 6 . 2 . 3 . 7 6 . 2 . 3 . 8 6.2.4 Soils used for the incubation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 1 Experimental procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 1 Chemical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 73 Mineral nitrogen (NH/-N and N03--N) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 73 Total dissolved nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 73 Dissolved organic carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 73 Microbial carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . 1 74 Microbial nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 74 Total carbon and total nitrogen in soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 74 Hot water soluble carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 74 Clay fixed nitrogen . . . . . . . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 74 Statistical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 75 6.3 Results .............................................................................................................. 17 5 6.3 . 1 6 . 3 . 2 6 . 3 . 3 6 . 3 . 4 6.3 . 5 6 . 3 . 6 6 . 3 . 7 6.4 6.4 . 1 Organic matter quality of tested soils . . . . . . . . . . . . . . . . . . . . .. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 5 Mineral N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 82 Ammonium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 84 Nitrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 87 Mineralisable N . . . . . . . . . . . . . . . .. . . . . . .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 89 Dissolved organic carbon (DOC) . . . .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9 1 Soil microbial biomass (SMB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . _ . . __ . . . . . . . . . . . . . . . 1 9 1 Discussion ......................................................................................................... 194 Nitrification . . . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . 1 98 6.4.2 Relationships between resin-adsorbed N and O.SM K2S04 - extractable N Xll ..... . . . . ...... . . . . . ........ . . . .. . . . . . .. .. .. . . . . . . . . . . . . . . . ..... . . . . . . .. . . . . . . . . . . . . .... . . . . . .. . .. . . . . . . . . . .. . . . . . . 204 6.5 Conclusion ................... ....................................... .................. ............................ 207 CHAPTER 7 MODELLING THE NITROGEN CYCLE IN SHEEP GRAZED NORTH ISLAND HILL COUNTRY PASTURE 7.1 Introduction ...................................................... ........................................ ....... 209 7.2 Model inputs and development ....................... ................... ........ .................... 209 7.3 Model outputs .......................................................................... .... .................... 214 7.4 Sensitivity of the model to different conditions ............................................ 220 7.4. 1 Impact of excretal distribution . . . . . . .. . . . . . . . . . . . . . . . . ..... . . . . ...... . . . . . . . . . . . . . . . . .. . .. . . . . . . 22 1 7.4.2 Impact of pasture utilisation . . .. . .... . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . .. . . . . . . . . . 222 7 . 4 . 3 Impact o f soil fertility, as affected by P fertiliser addition . . ...... . . .. . . .. . . . . . . . . 226 7.5 Improvement of efficiency of hill country N cycle ... . .. . ..................... . .......... 229 CHAPTER 8 SUMMARY AND IMPLICATIONS FOR FUTURE RESEARCH . . . . . .. ..... . . . . . . . . . 232 REFERENCES . . . ... . . . . . . ... ... . . . . . . . ... . . . . ...... . . . . . . . . . . .... . . . . . ........ . .... . . . . . . . . . ... ... . ... . . .... 241 APPENDIX 1 .................................................................................................. 261 APPENDIX 2 .................................................................................................. 263 APPENDIX 3 .................................................................................................. 271 APPENDIX 4 .................................................................................................. 274 LIST OF TABLES CHAPTER 2 Table 2. 1 Calculation of amount of dung and urine N derived from on each Xlll s lope category in the notional 1 ha paddock. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 Table 2.2 N balance on different land slopes within a notional sheep-grazed hill pasture. (Values are based on Fig. 2 . 1 and Fig. 2 . 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6 Table 2.3 N inputs and N surplus in the notional 1 ha hil l country paddock. Data on N inputs per ha and N surplus per ha are from Table 2.2 and proportions of land in each slope category are from Table 2 . 1 and Gillingham ( 1 97 8 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 CHAPTER 3 Table 3 . 1 The partition of urinary nitrogen (Doak, 1 952) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 Table 3 .2 Constituents of the synthetic urine solution (PH= 7 . 8 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43 Table 3 .3 . Treatments used in the experiment. . . . . . . ... . . ... . . . . . . .. .. . . . .......... . .. .................. .43 Table 3 .4 Apparent recovery of urine N as mineral Nand mineralisable N after urine application (kg N/ha/7 .5 cm depth) in the soil profile (0- 1 5 cm) i .e . all values are treatment minus control . * = did not measure . .............. .49 Table 3 .5 Mineralisable N levels (kg Nlha) in control soil s at different depths. Values are the average of soils sampled 27, 1 00 and 1 42 DAUA. . . . . . . . . . . . . 53 Table 3 .6 Pasture DM production and N uptake as effected by urine application ..... 5 8 Table 3 . 7 Effect of urine treatments on ammonia volatilization ........................ . ....... 59 Table 3 . 8 Apparent fate of urine NH/-N from 1 -27 DAUA (A) and 2 7 - 1 00 DAUA (B). All quantities are expressed as kg Nlha . . ............................... 64 Table 3 .9 Soil fertility indices of the two sites ........... ......... '" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 CHAPTER 4 Table 4. 1 Resin-adsorbed N (flg-N/5 cm2/3 days) from two different pasture plots . ....................................................................................... ...... .............. 78 Table 4.2 Resin-adsorbed N (flg-N/5 cm2/day) from N&N03 solution containing 1 0 flg/mL NH/-N and 1 0 )..l.g/mL N03·-N . ........ ..................... 79 Table 4.3 Adsorption ofN (llg-N/5 cm2/day) by glued spikes and fresh resin membranes from 25 mL samples ofNH4N03 solution containing XIV 1 0 Ilg NH/- N/mL and 1 0 Ilg N03'-N/mL over a day . . . . . . . . . . . .. .... . . . ......... . . 79 Table 4.4 N adsorption (llg-N/5 cm2/day) to resin spikes immersed in NH4N03 solution containing 1 0 Ilg/mL NH/-N and 1 0 IlglmL N03'-N for a day . . 80 Table 4.5 N adsorption to resin spike from homogeneous soil (flg-N/5 cm2/7 days) . ..................................................... .................. . .. . .. . . ... 8 1 Table 4.6 Estimation of resin-ads or bed N03'-N from NH4N03 solution containing different initial quantities ofN03·-N . . . . .. . ..... . . . ........ . ............. ... 8 5 Table 4.7 Estimation of resin-ads or bed NH/-N from NH4N03 solutions containing different initial quantities ofNH/-N . . ....... .......... . . ... . .. . ... ........ 86 Table 4.8 Soil NH/-N and N03'-N assessed by 2 M KC} extraction and resin adsorption, together with pasture N uptake over 1 2 days at 2 sites. Means with common letters are not significantly different (P<0.05) within a column at each site . .. . . . . . .. . . . .. . . . . . . . . . . . . . . . . . . . . . .... . . . . . .. . .. . . . ... . . . .. ... . . ..... 92 Table 4.9 Levels of mineral N in two soils as measured by 2 M KCI extraction and resin spikes. Measurements were made over a 7 day period in the Table 5 . 1 field and also after incubation for 7 days in the laboratory . ................. . . ... 9 5 CHAPTER S Treatments used in the experiment at Ballantrae AgResearch hill country research station investigating the fate of simulated urine N applied to hill country pasture . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . .. . . . .. . . . . . . . ... . . . . . . . . . . . 1 20 Table 5 . 1A Constituents of the synthetic urine solution (PH=7 .8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 20 Table 5 .2 Average meteorological conditions for 1 970- 1 995 at Ballantrae AgResearch Hill country research station (wind speed data was only available for 1 994) . .... . . . ... . . . . .... . . . . . . . . . . . . . . . . .... . . .. . . . . . . . ... . ... . . . . . . . . . . . . . . . . . ... . . . . . . 1 24 Table 5.3 Net mineral N (extracted by 2 M KCI ) in the soil profile (0-30 cm) (i .e. treatment minus control values) . . ... . ....... . . . . . . . . . . . . . . . . . . . . ....... . . ... . . . ... . . . . . 1 27 Table 5.4 Mineralisable N levels (kglha) in control soils at different depths. Values are the average of soils from sampling times at 3, 1 2 and 27 DAUA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 8 Table 5 . 5 Pasture DM production, herbage N concentration and N uptake fol lowing urine application . . . . . . ... . . . .. . . . .... . . . . . . . . . . . . . . . . ... . . . . . . .. . . . . .... . .. .. .... . . ... 1 44 Table 5 .6 Quantities ofNH3 trapped by samplers and estimated losses ofNH3 by volatilisation . .......... ... . . . . . . . ............... ............. . . . . . . . . . . .. .... ....... . . ..... . . ....... 145 Table 5 . 7 Amounts of urine N lost through NH3 volatilisation and overall xv recovery of added urine N at 3 DAUA. # = data from Table 5 . 3 . . . . . . . . . . . . 1 45 Table 5 . 8 Example of leaching calculation by the model. Data extracted from leaching model of the F560 treatment. . . ....... . . . . . . ............. . . .. . . . . .. . . ... . . . .. . .... 1 49 Table 5 .9 Estimated leaching losses from urine treatments by the two models . .. ... 1 5 1 Table 5 . 1 0 Example of nitrification calculation. Calculations up to day 8 are presented for the F560 treatment plots . . . . ... ...... .. . .. . . ... . . .. . . . . . . . . . . . . . ...... . . . . . . . 1 5 5 CHAPTER 6 Table 6 . 1 Apparent fate of urine N at 3 DAUA. * (rate of urine N was 400 ) lg N/g soil) i .e . values are treatment minus control) .. . . .... . .. . . . .. . . . . . .......... ...... 1 94 Table 6.2 Quantitative comparison ofNH/-N decrease and N03--N increase in urine treated soils from 3-45 DAUA. (All values are in )lg N/g soil) . ..... 1 9 8 Table 6.3 Nitrification rates during first 15 DAUA in the experimental soils . ... . . . . . 1 99 CHAPTER 7 Table 7 . 1 Data used to evaluate the model. G= measured data of Gillingham ( 1 978), B= measured data of B lennerhassett (2002), T= Findings from this thesis . . . . . . . . . . . .. . . .. .. . . ..................... . . . . . . . .. . .. . . .. . .. ... . ......... . ......... ..... 2 1 4 Table 7.2 Modelled balances for individual slope categories and for the overall paddocks taking into account that campsites, easy slopes and steep slopes occupy 1 2 .2%, 45 .5% and 42 .3% of the paddock area respectively . . . . . ............ ............. ........ . .... .................. . .... .... . ......... . . ... . ........ 2 1 7 Table 7.3 Data used to evaluate the model on a paddock with a high level of P fertility. G = measured data of Gillingham ( 1 978), B = measured data Table 8. 1 of (Blennerhassett (2002), T = Findings from this thesis . . . . . . . . . . ...... . . . . . . . . 226 CHAPTER 8 Comparison between sustainable levels of pasture production with current N inputs and theoretical maximum pasture production in different slope categories of hill country ..... ....... .. .... . ... . . . . ... . . . . . . . . . ..... . . . . . . 23 9 Fig. 2 . 1 Fig. 2 .2 Fig. 3 . 1 Fig. 3 .2 Fig. 3 . 3 Fig. 3 .4 Fig. 3 . 5 Fig. 3 . 6 Fig. 3 . 7 Fig. 3 . 8 XVI LIST OF FIGURES CHAPTER 2 Above-ground N balances for hill country with a north facing aspect. All values are kg Nlha/yr . . . ... ................ . ....... . ................................ . . 9 Above-ground N balances for hill country with a south facing aspect. All values are kg N/ha/yr . .. ................................................. ........... 1 0 CHAPTER 3 Percentage urine N recovery in urine patches. Data extracted from Ball et al. , 1 979 (Palmerston North) and Carran et al., 1 982 (Gore) ................... ........................................ . . . ..................... . ... . . . . .............. 23 Four general patterns of nitrification observed by Steel et al. ( 1 980) when soils are perfused with .005M (NH4)zS04 ........... . ..... . . . . . . . ... 32 Relationships between INA (Initial Nitrification Activity) and total N (%) and C/N ratio of soil . The data are extracted from soils that showed the Type 1 and Type 2 nitrification patterns of Steel et al. ( 1 980) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 Relationships between INA (Initial Nitrification Activity) and total N (%) and C/N ratio of soil . All the data presented in Steel et al. ( 1 980) were used for the relationships . ........... .. . ...... ....... ...... . . ....... . . 3 3 Daily rainfall during the experimental period . . . . . .. ......... .... ... . . ....... ... .... . . 46 Effect of urine application on soil (0- 1 5 cm) mineral N level. * = significant treatment differences were observed within a sampling at the same site. NS = no significant treatment differences were observed within a sampling at the same site ........... . ........ ............... . . ... . .... . 48 Effect of urine treatments on soil (0- 1 5 cm) NH/ -N. * = significant treatment differences were observed within the sampling time. NS = no significant treatment differences were observed within the sampling time . ...................................................... . . .............. ..................... 50 Effect of urine treatments on soil (0- 1 5 cm) N03 --N. * = significant treatment differences were observed within the sampling time. NS = no significant treatment differences were observed within the sampling time . . . ..... .......... . . . . . ..... . ........... ........ . . ... . .................. ...... . . ............ 5 2 Fig. 3 . 9 Fig. 3 . 1 0 Fig. 3 . 1 1 Fig. 3 . 1 2 Fig. 3 . 1 3 Fig. 3 . 1 4 Fig. 4. 1 Fig. 4.2 Fig. 4.3 Fig. 4.4 Fig. 4 .5 Fig. 4 .6 Fig. 4 .7 Effects of urine treatments on soil (0-7.5cm) mineralisable N levels at 27, 1 00, 142 DAUA. Treatments with common lower case letters do not differ (P<0.05) within a sampling day at the same site. Treatments at the same site with common upper case XVll letters do not differ at (P<0.05) between sampling days . . . . . . .. . . . . . . . . ... . . . . . . . . 53 Effect of urine treatments on pasture DM accumulation. Total DM accumulations at the same site with common upper case letters do not differ at the P<0.05 level . Treatments at the same site and at the same harvest with common lower case letters do not differ at the P<0.05 level. . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Effect of urine treatments on pasture N accumulation. Total pasture N accumulations at the same site with common upper case letters do not differ at the P<0.05 level. Treatments at the same site and at the same harvest with common lower case letters do not differ at P<0.05 level. . . . . 57 Effect of urine treatments on ammonia volatilisation. Treatments with common letters within a site do not differ at the P<0.05 level . . . . . . . . . . 5 8 Total urine N recovery (%) during the experiment. . . . . . . . . . .. . .... . . . . . .. .. . . . .. . ... 60 Urine N recovery during the experiment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1 CHAPTER 4 Polymerization synthesis of a styrene sulfonic acid cation exchange resin (Harland, 1 994) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Relationship between diffusion impedance factor and moisture content. (Logistic curve fit for the data in Fig. 4. 1 of Tinker and Nye (2000) .) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 N adsorption to resin spikes from different concentrations of NH4N03 solution. Each point represents the average of three replicates . . . . . . . . . . . . . . . . . . . . . . . . . . .... . . . . . ...... . . . . . . . . .. . . .. . ..... . ... . . .. . . . . ... . . . . . ..... . .. . . . . . . . .. . . 83 Estimated and measured resin-adsorbed N03'-N from N�N03 solutions containing different initial quantities ofN03'- N . . . . . . . . . . . . . . . . . . . . . . . . 85 Estimated and measured resin-adsorbed NH/-N from NH4N03 solutions containing different initial concentrations ofNH/ -N. n = Selectivity coeffi(:ient of K+ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 N adsorption by resin spikes with time from �N03 solution containing 1 0 �g NH/-N/mL and 1 0 �g N03'-N/mL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 NH4 + -N levels during the incubation as measured by the 2 M KCI - extractable and Resin methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Fig. 4 .8 Fig. 4 .9 Fig. 4. 1 0 Fig. 4. 1 1 Fig. 4. 1 2 Fig. 4. 1 3 Fig. 4. 14 Fig. 4. 1 5 Fig. 4. 1 6 Fig. 4. 1 7 Fig. 4. 1 8 Fig. 4. 1 9 XVlll N 03--N levels during the incubation as measured by 2 M KCI -extractable and Resin methods . . . . . . . . . . . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . 90 Notional box of soi l . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Modelled experimental soil cube with resin spike inserted, viewed from above. The anion exchange resin membrane is facing cells ( 1 0,9), ( 10, 1 0), ( 1 0, 1 1 ), and ( 1 0, 1 2) and the cation exchange resin facing the cells ( 1 1 ,9), ( 1 1 , 1 0), ( 1 1 , 1 1 ) and ( 1 1 , 1 2) . . . . . . . .... . ... . . . . . . . . . . . . . . . . . 1 00 Electrical conductivity of different NH4N03 concentrations at 20° C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 02 Measured and modelled N03'-N adsorption to resin spikes with time . . . . 1 06 Modelled N03'-N concentration (/ lg/cm3 soil solution) in each soil compartment (0.25 x 0.25 x 5 cm) 1 day after placement of the resin spike in soil . The asymmetrical depletion pattern is caused by the placement of the anion resin strip on the side of the spike apparently closest to the top of page. (This diagram shows only half of the system) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 07 Soil N03'-N (/lg/cell) distribution in each soil compartment (0.25x O.25x 1cm) 1 day after placement of the resin spike in soil . The resin spike is placed at 2-3 cm on the X axis and at 2.5 cm on the Y axis. The asymmetrical depletion pattern is caused by the placement of the anion resin strip on the side of the spike apparently closest to the reader. . . . . . . . . . ... .. . . . . . .. . . ............ ......... .... ......... . . . . . . . ... ..... ..... . . . 1 07 Modelled N03'-N concentration (/lg/cm3 soil solution) in each soil compartment (0.25 x 0.25 x 5 cm) 7 days after placement of the resin spike in soil (This diagram shows only half of the system). 1 08 Soil N03'-N (/lglcell) distribution in each soil compartment (0.25x 0.25x l cm) 7 days after placement of the resin spike in soi l . The resin spike is placed at 2-3 cm on the X axis and at 2 .5 cm on the Y axis . . . . . . . l 08 Modelled effect of soil moisture content on N03--N adsorption by resin spikes with time after burial. W = gravimetric moisture content. .. 1 09 Effect of soil moisture on soil N03--N (/ lglcell) distribution 7 days after placement of the resin spike in soil. The resin spike is placed at 2-3 cm on the X axis and at 2 .5 cm on the Y axis. The asymmetrical depletion pattern is caused by the placement of the resin strip on the side of the spike ' apparently closest' to the reader. . . . . . . . . . . . . .. . .. . . . . . . .... . . . . . . 1 1 0 Modelled effect of initial soil N03--N concentration (/ lg N03'-NI g soil) on N03--N adsorption by resin spikes with time after burial. The gravimetric moisture content is 0.26 (w/w) . . . . . . . . .. . . . . . .. . . . ... . . .... . . . ...... 1 1 1 Fig. 4.20 Fig. 4.2 1 Fig. 4.22 Fig. 5 . 1 Fig. 5 .2 Fig. 5 .3 Fig. 5 .4 Fig. 5 . 5 Fig. 5 .6 Fig. 5 .7 Fig. 5 . 8 Fig. 5 .9 Fig. 5 . 1 0 Fig. 5 . l 1 Fig. 5 . l 2 XlX Modelled effect of temperature on N adsorption by resin spikes with time. The moisture content of the soil is 0.26 (w/w) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 2 Schematic diagram of resin spike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 3 Schematic diagram of processes controlling the available soil N . . . . . . . . . . . 1 1 5 CHAPTER 5 Rainfall during the experimental period ( 1 417/2000- 1 9/ 1 0/2000) . . . . . . . . . . 1 24 Air and soil temperature during the experimental period ( 1 4/7/2000- 1 9/ 1 0/2000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 25 Effect of urine application on soil (0-30cm) mineral N levels . . . . . . . . . . . . . . . . 1 26 Effect of urine application on soil NH/ -N levels (O- I Q cm) as determined by 2 M KCl extraction and resin adsorption methods . . . . . . . . . 1 29 Quantities of 2 M KCI extractable NH/-N and N03--N in the 0-30 cm soil depth during the experimental period. Note. Change in scale between control and treated plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 30 Effect of urine application on soil N03--N levels (O- I Q cm) as determined by 2 M KCI -extraction and resin-absorption methods . . . . . . . . 1 32 Effect of urine treatments on resin-adsorbed N03--N at 5 5 to 97 DAUA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 34 Relationship between resin-adsorbed N and 2 M KCI extractable N when both steep and flat site data are used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 36 Relationships between (A) resin adsorbed N03--N and 2 M KCI ­ extractable N03--N. (B) Resin-adsorbed NH/-N and 2 M KCI adsorbed NH4+-N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 37 Effects of urine treatments on soil (0- 1 0 cm) mineralisable N levels at 3, 1 2, 27 days after urine application. Treatments with common upper case letters do not differ at P<0.05 level within a sampling day. Treatments with common lower case letters do not differ at P<0.05 level between sampling days of the same treatment. . . . . 1 3 9 Effect of urine application on soil mineralisable and mineral N levels (O- I Q cm depth) at 3, 1 2 and 27 days after urine application (DAUA) . . . 1 40 Relationship between increase in soil mineral N and decrease in soil mineralisable N from 3 to 1 2 and from 1 2 to 27 DAUA. . . . . . . . . . . . . . . . . . 1 4 1 Fig. 5 . 1 3 Fig. 5. 1 4 Fig. 5 . 1 5 Fig. 5 . 1 6 Fig. 5 . 1 7 Fig. 5 . 1 8 Fig. 5 . 1 9 Fig. 5 .20 Fig. 5 . 2 1 Fig. 5 .22 Fig. 5 .23 Fig. 5 .24 Effect of urine treatments on pasture DM accumulation at the flat and steep sites. Dry matter yields with common letters between treatments within same cut and same site do not differ at P<0 .05 . Total DM accumulation with common lowercase letters between xx treatments within same site do not di ffer at P<0 .05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 42 Effect of urine treatments on herbage N accumulation. Herbage N accumulations with common uppercase letters between treatments within the same cut and site do not differ at P<0 . 0 5 . Total herbage N accumulations with common lowercase letters between treatments within same site do not differ at P<0 . 0 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 43 Effect of urine treatments on ammonia volatilisation during the first 6 days after urine application. Values with common upper case letters between treatments within same site do not differ at P<0.05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 44 Estimated and measured 2 M KCI -extractable N03'-N in the 0- 1 0 cm soil depth during the experimental period. Arrows indicate measured 2 M KCI -extractable N03' N. Other marked data points ( .) are estimated 2 M KC I -extractable N03'-N values from the relationship with resin adsorbed N03'-N illustrated in Fig. 5 . 8 - . . . . . . . . . . . . . 1 49 Estimated quantities of soil N03'-N (g/m2) in the 1 0-20 cm and 20-30 cm depths from the two models during the experimental period . . 1 52 Estimated cumulative leaching of N03--N (g/m2) from 1 0-20 cm and 20-30 cm soil depths from urine treatments from the two models during the experimental period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 53 Schematic diagram to illustrate nitrification rate calculation . . . . . . . . . . . . . . . . . . 1 54 Cumulative daily nitrification during the experimental period . . . . . . . . . . . . . . . 1 5 6 Frequency distribution of daily nitrification rates calculated at each site for the period up to 45 DAUA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 8 Urine N recovery during the experimental period. The dotted line indicates the application rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 9 Urine N recoveries (%) during the experimental period, estimated as the sum of soil mineral N, NH3 volati lisation, plant uptake and leaching ofN03--N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 60 Resin adsorbed NH/-N (A) and N03--N (B) l evels during the experimental period in control treatments at both sites . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 62 Fig. 6. 1 Fig. 6.2 Fig. 6.3 Fig. 6.4 Fig. 6 .5 Fig. 6.6 Fig. 6.7 Fig. 6.8 Fig. 6.9 Fig. 6. 1 0 Fig. 6. 1 1 Fig. 6. 1 2 Fig. 6. 1 3 Fig. 6. 1 4 Fig. 6.1 5 Fig. 6. 1 6 XXI CHAPTER 6 S oil carbon related organic matter properties studied during the experiment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 77 Nitrogen related organic matter properties studied during the experiment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . 1 78 S oil carbon to nitrogen ratios in experimental soiL. . . . . . . . . ... . .. . . . . . . . . . . . . . . . . . . 1 79 Relationships between organic matter quality parameters . . . . . . . . . . . . . . . . . . . . . . 1 8 1 Effect of urine application on 0.5M K2S 04-extractable soil mineral N . (The statistical analysis of the data in this figure is included in Appendix 3) . ......................................................................... 1 83 Percentage of urine N recovered as soil mineral N at the beginning (3 DAUA) and end (45 DAUA) of the experiment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 84 E ffect of urine application on soil 0.5M K2S04-extractabl e NIL; + -N. (The statistical analysis of data i n this fi gure is included in Appendix 3) . ............................................................................................ 1 85 Effect of urine application on resin-adsorbed NIL; +-N. (The statistical analysis of data in this figure is included in Appendix 3) . ...... 1 86 Effect of urine application on 0.5 M K2 S04 -extractable soil N03--N. (The statistical analysis of data in this fi gure is included in Appendix 3) . ........................................................................................ .... 1 88 E ffect of urine application on resin-adsorbed N03--N over time. (The statistical analysis of data in this figure is included in Appendix 3 .) ...................... ...... ........... ......................... ............................ 1 89 Effect of urine application on soil mineralisable N with time. (The statistical analysis of data in this figure is included in Appendix 3) . ...... 1 90 E ffect of urine application on soil DOC levels. (The statistical analysis of data in this figure is included in the Appendix 3 ) . . . . . . . . . . . . . . ... 1 92 Effect of urine application on soil microbial biomass N with time. (The statistical analysis of data in this figure is included in Appendix 3) ............................................................................................. 1 93 The distribution ofNH/-N and N03'-N in urine treated soils with time after urine application . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . 1 97 Comparison of nitrification of urine N and control soil N . . . . . . . . . . . . . . . . . . . . . . 200 Relationship between CIN ratio and nitrification rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 Fig. 6. 1 7 Fig. 6 . 1 8 Fig. 6 . 1 9 F ig. 6 .20 Fig. 7. 1 F ig. 7 . 2 F i g . 7.3 Fig. 7.4 Fig.7 .5 XXll Relationship between nitrification rate and the ratio of labile organic C to TC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Relationship between nitrification rate and soil pH of the experimental soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Relationship between resin-adsorbed N and 0.5M K2S 04- extractable N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Relationships between resin-adsorbed N and 0 . 5M K2 S 04- extractable N in different soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 CHAPTER 7 Modelled N cycle in a hill country paddock with northerly aspect ( 1 2.2% campsite, 45.5% easy slope, 42.3% steep slope) . All values kg Nlha/yr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 5 Modelled N cycle in a hill country paddock with southerly aspect ( 1 2 .2% campsite, 45.5% easy slope, 42.3% steep slope) . All values are kg Nlha/yr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 6 Modelled N balance for two hill country paddocks with contrasting proportions of steep, easy and flat land. All values are kg Nlha/yr. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P.u. = Pasture utilisation, Excretal N = Percentage of excretal N deposited on each slope category in that paddock. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 N Balances for hill country paddocks with different excretal distributions. Values are kg Nlha/yr. Pasture DM production, proportion of clover in herbage and N concentration in herbage were as for the north aspect paddock in Table 7. 1 A. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, AT= Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P.U. = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope category in that paddock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 N Balances for hill country paddocks with different pasture utilisations. Values are kg Nlha/yr. Pasture DM production, proportion of clover in herbage and N concentration in herbage were as for the north aspect paddock in Table 7. 1 A. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P .u. = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope category in that paddock. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Fig. 7.6 Fig. 7.7 N B alances for north aspect hill country paddocks under l o w P and high P conditions. Values are kg N/halyr. F = N input by l egume N fixation, non symbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P.D. = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope XXlll category in that paddock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 N Balances for south aspect hill country paddocks. Values are kg N/halyr. F = N input by legume N fixation, non s ymbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P.D. = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope category in that paddock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 XXlV LIST OF PLATES Plate 3 . 1 Ammonia volatilisation measurement using chamber methods (Ball et al. , 1 979; Theobald, 1 98 3 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Plate 3 .2 Ammonia volatilisation measurement using passive samplers (Carran et al., 2000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 1 CHAPTER ! INTRODUCTION In New Zealand grazed pastures, nitrogen (N) fixation by clover is the traditional method of supplying N to the grasses that make up the bulk of the pasture sward. Nitrogen fixed by clovers is contained initially in the growth of the c lover plant itself. Eventual ly however, most of the fixed N is recycled back into the soil, either through the breakdown of plant litter, roots and nodules, or by urine and faecal return from grazing animals. The N then becomes available for uptake by grasses. The amount of nitrogen fixed by clovers is roughly proportional to clover growth (Ball and Field ( 1 9 82). In many circumstances in hill country, this can b e very low. In an experiment conducted at the AgResearch Ballantrae Research Station, the average quantity of N fixed by legumes in one year on a low-fertility, unimproved, North Island hill country site was estimated to be only 1 3 kg Nlha (Grant and Lambert, 1 979). Although under summer-wet conditions, such as exist at Ballantrae, clover growth and N fixation can be increased markedly by increasing soil fertility, in summer-dry climates white clover may not persist because of moisture stress (Gillingham et al., 1 998) - despite high levels of soil P and S. In such situations pastures can be critically short of available N. Recognising this problem, many experiments have been c onducted in hill country pastures to assess possible N responses. The general conclusion has been that large N fertiliser responses are obtainable in most hill pastures (Luscombe, 1 98 0 ; Ball et aI., 1 982; Lambert and Clark, 1 98 6 ; Clark and Lambert, 1 9 89; Gillingham et al., 1 998; Blennerhassett, 2002). Gillingham et al. ( 1 9 98) considered selective N fertiliser application, taking slope and aspect into account. They suggested that application of "normal" rates of P fertil iser to south aspects and moist north slopes and application of greatly reduced rates of P fertiliser, together with strategic applications of N fertiliser, on steep north- facing slopes, would be the most efficient fertiliser policy. Taking this concept further, Gillingham et al. ( 1 999) suggested that the use of Global Positioning 2 S ystems (GPS ) and Geographic Information Systems (GIS) had the potential to improve greatly the efficiency of fertiliser use in hill country. These technologies also have the potential to reduce the environmental impacts of fertiliser use. However, to adopt these types of advanced, differential fertiliser application on hill country farms requires more detailed infonnation on pasture productivity and associated pasture responsiveness. In particular, more information is required on the N ferti lity on contrasting topographic land units. Detailed studies of N flows in hill country pastures, especiall y in summer-dry hill pastures are scarce. The wide variation in micro climates and topographic units associated with hill country pastures make it difficult to achieve a complete picture of the N cycle in a single experiment. Therefore, in recent years more emphasis has been placed on detailed studies of individual aspects of the main N cycle in hill country. The influence of fertiliser history and climate regime on pasture production and N accumulation in hill soils, was studied by Moir (2000) . The constraints imposed on pasture growth by N deficiency in hill country, were studied by Blennerhassett (2002). Nitrogen recycling through animal excreta has long been considered beneficial to the ferti lity of grazed pastures ( Sears et al. , 1 965; Levy, 1 970) . Nevertheless, recent research has indicated that grazing animals can cause substantial N outgoings from highly productive, flat land pastures through losses from urine patches (Ball et al. , 1 979; Carran et al. , 1 982; Ball and Keeney, 1 983). In contrast, Sakadevan et al. ( 1 99 3 ) indicated that N is conserved in urine patches o f h i l l country pastures in New Zealand and Lambert et al. ( 1 982a) also pointed out that hill soils are strongly N retentive because of their high CIN ratios. The experiments described in this thesis attempt to clarify the apparently contrasting findings described in the previous paragraph, and to explore whether there are processes operating in urine patches of hill pastures that are different from those in more intensively fanned environments. 3 A major obstacle to gammg a good understanding of N dynamics in hill country pastures i s the absence of a convenient way of measuring the pool of available soil N. The widely used extraction with 2 M KCI requires quick analysis after sampling. This is a constraint in experiments conducted at remote sites. In addition, the measure only gives a "snapshot" of the quantity of available N in the soil at the time of sampling. As will be pointed out later in the thesis, although the pool of mineral N extracted by 2 M KCI is usually small, it is extremely dynamic. There are large inflows and outflows of N over short time periods. A simple measurement of the size of the mineral N pool, at one point in time is often therefore, not a good indication of the actual N fertility of the soil. In an attempt to overcome some of these problems, an in situ N measuring method by ion exchange resin membrane spikes was developed. The possible utility of the technique was assessed in preliminary field studies and by using simulation modelling. The overall objectives of this thesis were therefore to examine the fate of sheep urine in hill country pastures, using a range of established and novel measurement techniques . The information so gathered is then used to improve the quantitative modelling of N cycle in hill country pastures. 2.1 Introduction CHAPTER 2 LITERATURE REVIEW 4 This literature review aims to fami liarise the reader with the obj ectives of the series of experiments described in this thesis, and how the obj ectives were developed to fill the information gaps in the available literature. However, coverage of the relevant literature in this chapter is brief, with the main aim being to illustrate the broader concepts inherent in New Zealand hill country pastoral farming, and in particular variations in the above-ground nitrogen balances between different topographic units. More detailed reviews of the literature relevant to each experiment are contained within the introductions of individual chapters . 2.2 New Zealand hill country New Zealand " hill country" i s defined as steep non-arable hills below 1 000 m, which cover about 5 million ha (40% of farmland) in the North Island and 4.9 million ha ( 5 1 % of farmland) i n the South Island (Joblin, 1 98 3 ; White, 1 990). Land i s defined as mo deratel y steep when the ground slope is between 2 1 ° and 25 °, steep between 26° and 3 5° and very steep when the ground slope is > 3 5° (NZLRI, 1 979). Campbell ( 1 95 1 ) classified New Zealand hill country into three different types: i .e., wet hill country, dry hill country and tussock hill country. In the North Island, dry hill country extends the entire length of the east coast in a narrow strip 64 km wide in the Wellington province, narrowing to 1 6 km in the direction of East Cape (C ampbell, 1 95 1 ) . This region has lower rainfall than wet hill country, with hot dry summers and occasional rains from the north-east. At present this land resource carries 3 5 % of the total sheep in New Zeal and and 20% of the total cattle. These animals are on 68 00 farm holdings, representing 3 0% of New Zealand sheep and beef farms (Mackay et al., 1 993). Sheep breeds vary widely but 5 Romney and Romney crosses predominate. Aberdeen Angus is the predominant beef breed although Herefords, Friesians, and various crosses of these breeds are popular. 2.2.1 Topography Hill country topography varies from rolling to steep although most farms have at least a small area of relatively flat land. Hill country slopes may be moderate or steep and they differ in aspect and micro-topography. Contrasting aspects give distinctly different microenvironments. Variation in the amount of direct solar radiation received by sloping surfaces is one of the main reasons for the existence of microclimate differences in hill country. Seasonal temperature and moisture regimes vary markedly between warmer (north-facing) and cooler (south-facing) slopes. Lambert ( 1 973) investigated the differences in climate, soil and pasture between aspects in hill country. He observed different soil nutrient levels on various aspects, arising through nutrient transfer by grazing animals and the differential action of climate during soil formation. Furthermore, he noted different moisture levels on the various aspects, resulting from differential evapotranspiration (ET) rates and leading to different degrees of moisture stress, especially during the summer/autumn period. Lambert ( 1 973) agreed with Suckling ( 1 966), that hill country should, where practicable, be fenced with regard to aspect and where large scale variation due to aspect exists, fertiliser could be spread at varying rates as indicated by soil nutrient status. Gillingham et al. ( 1 998, 1 999) suggested adoption of differential fertiliser application to hill country farms using the information on pasture productivity from contrasting topographic land units and precise aerial fertiliser application with the use of Global Positioning S ystems (GPS) and (GIS) Geographic Information Systems. 2.2.2 Soil Gillingham ( 1 978) described hill country soils as follows. "North Island hill soils range from Yellow-grey earths in parts of Hawkes Bay and Wairarapa to Northern Yellow­ brown earths in Northland. The central North Island was originally covered by various ash showers. Rhyolitic ash blanketed the central and eastern North Island producing "Bush sick" or cobalt-deficient pumice soils while mainly andesitic ash formed Yellow- 6 brown loams in the remainder of the ash-covered area. Ash was lost from steep slopes and so steep land soils are derived almost entirely from underlying mudstone, siltstone, and sandstone parent materials (Steep land yellow-grey earths and Central yellow­ brown earths)" The nutrient status of hill soils is generally poorer than that of intensively farmed lowland soils (Mauger, 1 979). Most North Island hill soils have low nutrient availability, especially N and P and sometimes S, K and Mo (Lambert et al. , 1 982a). Many top soils in the moist hill country of the lower North Island contain very large quantities of nitrogen, rendered largely unavailable by the relatively wide carbon : nitrogen (C/N) ratio of the soil organic matter present (Ball et al., 1 982) . 2.2.3 Pasture composition Pasture botanical composition is one of the most suitable biophysical indicators of sustainability of North Island hill pastures (Lambert et al. , 1 996). Machado ( 1 994) observed high variability of pasture composition in his experiment conducted at seven different sites at the Ballantrae AgResearch hill country research station. The sites varied from undeveloped, unfertilised hill pasture containing only low-fertility adapted grasses and weeds to a highly productive sward with a high proportion of high fertility responsive grasses and white clover. The differences in plant species composition and production are a product of soil and environmental factors as influenced by topography and animal grazing and treading effects (Gillingham and During, 1 97 3 ; Grant and Brock, 1 974). Grant and Brock ( 1 974) indicated that differences in pasture composition due to soil properties were not great, except where they appeared to be related to differences in soil-moisture characteristics. They observed that Lolium perenne (Perennial ryegrass) and Trifolium repens (White clover) were most frequent on flat to low slope areas, Notodanthonia common on steeper, drier sites and Holcus lanatus (Y orkshire fog), Lotus pedunculatus (Lotus), rushes and sedges on wetter sites. M acfarlane and S heath ( 1 984) outlined the importance of flexibility in legumes on dry hill country, in terms of a wider range of flowering dates and high seed production, so that they can persist and produce under the variable soil moisture conditions. Ledgard et al. ( 1 987) observed marked seasonal differences in legume growth. They observed 7 high growth of subterranean clover in spring, white clover in summer and autumn in campsites, and lotus in winter. The optimum content of legumes in hill pasture should be up to 20-25 % of the sward (Suckling, 1 959). 2.2.4 Nitrogen use in hill country Until recently, most New Zealand hill country pastures have relied completely on legumes (mostly clover) to supply N by atmospheric N fixation. Phosphorus fertilisers are applied to pastures to produce satisfactory clover growth. Nitrogen fixed by clover is initially used in the growth of the clover and most of the fixed N is then recycled back to the soil, either through breakdown of pl ant litter, roots and nodules or by urine and faecal return from grazing animals (Tillman, 1 99 3 ) . S ears ( 1 953) indicated that ingestion by stock with subsequent return in animal excreta is the major pathway of transfer of N from legumes to associated grasses . However, N supply from this low cost source is often limited in hill country by inadequate moisture for vigorous clover growth and persistence, especially during spring and summer. In summer-dry climates, white clover does not persist, and annual clovers form the basis of pasture legume content (Gillingham et al., 1 998). These annual clovers make a small contribution to annual pasture production and associated N fixation (Brock, 1 9 73). Consequently, direct al leviation ofN deficiency by fertiliser appl ication has been suggested. Several studies have observed large N fertiliser responses in hill country pastures; 43 kg DMlkg N (Gillingham et al., 1 998), 3 7 . 8 kg DMlkg N (Ball et al. , 1 976), 28 kg DMlkg N (Lambert and Clark, 1 9 86), 5-25 kg DMlkg N (Luscombe, 1 9 80). Morton et al. ( 1 993) reported that many Wairarapa hill country sheep and beef farmers apply N fertilisers at a mean rate of 20 kg Nlha/yr. This appears to be a strategy that is increasingly popular, though driven by product prices . Gill ingham et al. ( 1 998) suggested adoption of a differential fertiliser appl ication policy on hil l country pastures. They suggested application of P and S to south aspects and moist north slopes, which have adequate clover content in the pasture. In contrast, steep, north-facing slopes should receive N ferti liser, and only very limited P, because of low legume contents. Furthermore, as mentioned in Section 2.2. 1 , use of GPS and 8 GIS for precIse aerial fertiliser application on hill country was proposed. They highlighted from their desktop study that greater p asture and animal production can be achieved in this way than from the same amount of fertil i ser appl ied uniformly over a block of typical hill land. However to adopt this type of advanced fertiliser policy there needs to be detailed information on N fertility in each micro-topographic unit of hill country. 2.3 Nitrogen balances in different topographic units of hill country Farming is a profession that needs a detailed understanding of the interactions between plants, climate and soil as wel l as the influence of biotic factors above ground (man and grazmg animals) and the soil fauna and flora (worms, bacteria, fungi etc .) below ground. The main challenge the hill country farmer faces is to provide adequate nutrients to match pasture demands whilst maintaining the soil fertility for future growth. No ecosystem, whether natural or managed is completely ' leak-free ' but the opportunities for loss increase with increasing input, especially of N (in its various forms) where animal production is involved (Jarvis, 1 999). A nutrient balance provides detailed information on the inputs, outputs and internal recycling of the nutrient. The cycle usually includes a number of compartments, e.g. soil, crop, animal , and depending upon the l evel of detail required, these may be further subdivided into different pools. At the simplest l evel, a budgetlbalance is a nutrient accounting process that sums all the inputs and outputs to a given defined system. The need for detail and accuracy in budget compilation depends solely on the purpose of the balance (Jarvis, 1 999). The aim of this section is to develop simple above ground N balances for the main topographic units in summer-dry hill country (north aspect-easy, north aspect-steep, south aspect- easy, south aspect-steep and flat campsite) as a way of integrating available data in the literature (Fig. 2 . 1 and 2.2). These N balances are developed for a notional 1 ha hill country paddock with the same slope distribution as that studied by Gillingham ( 1 9 7 8 ) ; 1 2 .2% campsites, 45.5% easy slopes (250 slope) and 42 . 3 % steep slopes ( 450 slope). The same division of slopes was considered for notional 1 ha north and south facing paddocks. The paddocks were assumed to be located on summer dry hill country in Hawkes Bay at the AgResearch Waipawa research station. A) Campsites • L I_N_fi.::.;:.:�;_io_n_..J - Animal products 1 1.9 t Animal 1 18.5 -- Dung & urine pool 106.7 B) North easy s lopes N fixation 29.4 -- --. Animal 150.8 Dung & urine --+­ pool 1 3 5.7 � �/ �� 3 1.3 / Soil solution + Organic matter C) North steep slopes Animal N fixation o products 1 't I::l-C-L:iiJ � ��\ t � Dung & -- Dung & urine to the site 492 Dung & urine to the site 56.6 Dung and urine to the site 10.2 9 Fig. 2. 1 Above-ground N balances for a sheep grazed hill country paddock with a north facing aspect. All values are kg N/halyr. N fixation 8.3 A) Campsites t Dung & - Dung & urine pool __ urine to the Animal 1 18.5 ���7 N fixation 3.7 B) South easy slopes - Animal 94.7 Dung & urine pool 85.2 site 372.6 Dung & urine to the site 42.9 ���7 C) South steep slopes Animal N fixation products 2 4.3 1 t t Dung & urine Dung & to the site �- Animal --.. urine --.. 56.7 43.2 pool 7.7 38.9 � 13.5 7 Soil solution + Organic matter 1 0 Fig. 2.2 Above- ground N balances for a sheep grazed hill country paddock with a south facing aspect. All values are kg Nlhalyr. 1 1 2.3.1 Pasture N uptake The values for the pasture N uptake at each site were estimated from me asured pasture production and pasture N concentration (N%) data of Bl ennerhassett (2002) at the Waipawa AgResearch research station. The values were taken from the control plots of low P fertility farmlets, which received sufficient superphosphate to maintain the original Olsen P status of 9 Jlg/g soi l . The measured DM production by Bl ennerhassett (2002) consisted of both grass and clover DM. Hence to estimate the N uptake from soil the amount of fixed N (see Section 2 . 3 .2) was deducted from the estimated total plant N uptake calculated by multiplying DM production and pasture N concentration. On both the north and south aspects, N uptake was less on steep slopes than easy slopes and campsites. This reflects the differences in pasture production between steep slopes and campsites. Previous studies have also shown that flatter campsites produced higher DM production than steep slopes (Suckling et al. , 1 959; Gillingham, 1 97 8 ; Ledgard et al., 1 9 87). Ledgard et al. ( 1 982b) observed greater pasture growth on warmer, north-facing slopes than on colder, south-facing aspects. This was in agreement with Blennerhassett ' s (2002) data for easy slopes. However, Gillingham ( 1 974) reported that slope accounted for more of the variability in pasture growth than did aspect. 2.3.2 N fIxation Symbiotic fixation of atmospheric nitrogen by Rhizobia in nodules on the roots of clovers is a key factor in the grassland systems of New Zealand. No country is more dependent upon the grass/clover association than New Zealand (Lee, 1 97 1 ) . The level of biological N fixation by pasture l egumes can vary greatly. Annual N fixation in developed lowland pastures is around 1 84 (range 1 0 7-392) kg N/ha (Hoglund et al., 1 979) and N fixation through symbiotic fixation in unimproved North Island hill country was around 13 kg N/ha (Grant and Lambert, 1 979). S easonal correlations observed by Hoglund et al. ( 1 979) indicated temperature to be the most influential component (negative in summer and positive in winter) of climate on N fixation. 1 2 Factors affecting N fixation have been extensively reviewed by Ledgard and Steel ( 1 992). Nitrogen fixation (kg N/ha/yr) by legumes was assumed to be proportional to clover growth. The proportionality constants varied between campsites and sloping sites (Ledgard et al., 1 987) . The N fixation in the current N balances (Fig. 2 . l and 2.2) was estimated by multiplying legume dry matter production by the factors (0. 040 for steep slopes and 0 .03 0 for camp sites) suggested by Ledgard et al. ( 1 987). The multiplication factors were developed by Ledgard et al. ( 1 98 7) based on legume N concentration (mean 4. 9%) and the proportion of legume N fixed from the atmosphere (PN) (82% for steep and 62% for campsite) . Lower PN values in campsites were associated with high inorganic N levels, which were largely a result of transfer of dung and urine by grazing sheep to these sites. The data on annual legume DM production (kg/ha) in control plots of low P farmlets of Blennerhassett (2002) were used in the current N balance. Interestingly, Blennerhassett (2002) did not observe any legume production on dry North-steep slopes . However, the highest legume production was observed in the North-easy s lopes. S imilarly, on the southerly aspect, more legume production was observed on campsites than slopes (Table 7 . 1 of Chapter 7). The annual values for N fixation estimated in the current N balance (Fig. 2 . 1 and 2 . 2) were much less than the estimations (54-8 5 kg N/ha/yr) by Ledgard et al. ( 1 9 8 7). In their experiment, they observed high legume DM production as the experimental area had received 3 5 0 kglha of superphosphate annually since development into p asture, about 20 years previously. In addition, the experimental site of Ledgard et al. ( 1 987) was in moist hill country. In a recent review Ledgard (200 1 ) reported annual legume N fixation values of 1 5 , 5 5 and 3 0 k g Nlha/yr for steep (>20° slope), easy ( 1 0-20° slope) and campsites (0- 1 0°) respectively within hill country. 2.3.3 Non-symbiotic N fixation and atmospheric deposition 1 3 Grant and Lambert ( 1 979) suggested that in a poorly developed hill country pasture, N fixation by non-symbiotic organisms and N in rainfall are together as important as clover N fixation. They estimated about 2 1 kg Nlha/yr for non-symbiotic N fixation and atmospheric deposition using the intercept of a regression between N fixation and legume DM production. Values for non-symbiotic fixation ( 1 3 kg Nlha/yr) and atmospheric deposition (3 kg Nlha/yr) in the current N balance, were obtained from the N balance reported for unimproved hill pasture by Lambert et al. ( 1 982a). 2.3.4 Pasture utilisation Under sheep grazing systems, poor pasture utilisation can occur because sheep tend to preferentially graze pasture on tracks or easy slopes rather than on banks or steep land zones (Gillingham, 1 982; Clark et al., 1 984). Clark et al. ( 1 984) found in their study that ryegrass til lers on slopes < 1 2° were grazed more frequently (number of grazing events of a marked grass tiller during a given period) and severely (leaf length grazed per tiller) by sheep than those on slopes >25° . Under New Zealand conditions, the uti lisation or consumption of the pasture that is produced can range from 40% to 90% depending on the stocking rate and method of grazing (Quin, 1 982). The pasture utilisation (%) for the current N balance was taken from Gillingham ( 1 978). He estimated the annual pasture utilisation as 79.2%, 8 2 . 8%, and 76.2% for campsites, easy slopes and steep slopes respectively. This pasture utilisation was measured in a paddock that had a topography of 1 2 .2% campsites, 4 5 . 5 % easy slopes and 42 . 3 % steep slopes. Pasture utilisation in a second paddock, that had 20. 1 % campsites, 5 5 .7% easy slopes and 24 . 1 % steep slope was 77.2%, 8 6 . 5 % and 8 0 . 8 % respectively (Gil l ingham, 1 97 8 ) . 2.3.5 N remaining in pasture litter 1 4 Gillingham ( 1 978) defined litter as dead plant material present in the pasture below grazing level at the end of grazing. This included both loose and attached material . It also included any whole plants that had been uprooted during grazing, then rej ected. Nitrogen remaining in the litter was estimated for the current N balance by deducting the animal utilisation from the pasture N uptake. Radcliffe ( 1 982) pointed out that herbage on flatter areas and wetter southerly aspects tended to decay faster than on north and steep slopes. Thus, accumulation of pasture litter was greater on northern steep slopes than on wetter, flatter areas. This would again lead to grazing discrimination between micro-sites as animals choose green herbage. 2.3.6 N in animal products The N removed in animal products was estimated usmg the proportion of animal ingestion (4 1 kg Nlha/yr) that was removed in animal products (4 kg Nlha/yr) reported by Lambert et al. ( 1 982). They calculated the N removal in animal products using the average N (2 . 5 %) content of products (meat and wool). Gillingham ( 1 978) also assumed that 1 0% of P intake by animals was lost from the grazing area in animal products. 2.3.7 Dung and urine N As noted in the previous section, only about 1 0% o f the pasture N ingested by grazing animals is retained in animal products. The reminder is returned in excreta to the paddock. The total amount of excretal N returned to the notional 1 ha paddock is calculated in Table 2 . 1 for the north-facing paddock. 1 5 Table 2 . 1 Calculation of amount of dung and urine N derived fro m on each slope category in the notional 1 ha paddock. Site (% area) Ingested N N retained in N in dung & Dung & urine N derived from (kglha) animal products urine pool each slope category within (kg N/ha) (kg Iha) the whole paddock Campsites ( 1 2 .2%) 1 1 8.5 1 1 .9 106.6 1 06.6 x 0. 1 22 = 1 3 .0 kg N Easy (45 .5%) 1 50.8 1 5 . 1 1 35 .7 1 35 .7 x 0.455 = 6 1 .7 kg N Steep (42.3%) 40.3 4.0 36.3 36.3 x 0.423 = 1 5 .4 kg N Total dung & urine N deposited to 1 ha paddock 90. 1 kg N In hill country pastures the stock tend to camp on flat areas of land and significant quantities of nutrients are transported to these areas through dung and urine from the steeper slopes where the sheep graze (Rumble and Esler, 1 96 8 ; Gillingham and During, 1 97 3 ; Rowarth and Gillingham, 1 990). Gillingham ( 1 97 8 ) measured the dung P distribution in two hill country paddocks ( Section 2 . 3 .4). His data were used to estimate the proportion of the total excreta deposited on the whole paddock that were deposited on each slope category. As noted earlier, the present N balance was developed for a notional 1 ha paddock, that had 1 2 .2% campsites, 4 5 . 5 % easy slopes, 42 . 3 % steep slopes. This was the same topography as the paddock studied by Gillingham ( 1 978). The measured proportions of total excreta deposited on each of the slope categories in the paddock of Gillingham ( 1 978) were 66.6%, 2 8 .6% and 4 . 8% for campsites, easy slopes and steep slopes respectively. These same proportions were assumed for the present study. Using these data for excretal return, the quantities of dung and urine N returned to each slope category within the notional 1 ha paddock were calculated as fol lows. Quantity of dung & urine N deposited on campsite =(66 .61 1 00) (90 . 1 ) kg N Quantity of dung & urine N per ha of campsite = ((66. 6/ 1 00) (90 . 1 » / 0 . 1 22 =49 l . 8 kg N/ha on campsites Quantity of dung & urine N deposited on easy slope= (28 .6/ 1 00) (90. 1 ) kg N Quantity of dung & urine N per ha of easy slope = ((2 8 . 61 1 00) (90. 1 » /0.4 5 5 =56.6 k g N/ha on easy slopes Quantity of dung & urine N deposit to steep slope =(4.8/ 1 00) (90. 1 ) kg N Quantity of dung & urine N per ha of steep slope =((4.81 1 00) (90. 1 » /0.423) = 1 0.2 kg N/ha on steep slopes 2.3.8 Above ground N balance in hill country 1 6 The above ground N balance (Fig. 2 . 1 and 2 . 2) developed in this section indicates that N cycling in hill country is more controlled by slope than aspect. However, seasonal variations were not considered in this balance. In winter more p asture growth can be expected on north aspects due to better sunlight while in summer more pasture growth can be expect on southerly aspects due to less moisture stress. Table 2.2 N balance on different land slopes within a notional sheep-grazed hill pasture. (Values are based on Fig. 2. 1 and Fig. 2.2). North aspect South aspect Campsite Easy Steep Campsite Easy Steep slope slope slope slope Inputs (kg/ha/yr) Legume N fixation 8 29 0 8 4 2 Non-symbiotic N fixation 1 3 1 3 1 3 1 3 1 3 1 3 Atmospheric. deposition 3 3 3 3 3 3 Total Input 24 45 16 24 20 1 8 Outputs (kg/ha/yr) Animal products 1 2 1 5 4 1 2 10 4 Animal transfer -385 79 26 -266 42 3 1 Total output -373 94 30 -254 52 35 N surplus (kg/ha/yr) +397 -49 - 14 + 278 -32 - 1 7 1 7 Estimates of N inputs, removal by grazing, and transfer in excreta reveal the existence of two extreme topographic units in hill country pasture. They are net N-gaining flat campsites and net N-Iosing sloped sites. Ledgard (200 1 ) reported a N balance for a hill country sheep-grazed pasture system with 45% of the area as steep slopes (>20°), 40% as easy slopes ( 1 0-20°) and 1 5 % as campsites. He reported that steep slopes have a negative N balance (- 25 kg N/ha/yr) while easy slopes (+ 1 5 kg N/ha/yr) and campsites (+ 1 90 kg N/ha/yr) have a net gain of N. These results will be discussed in detail in Chapter 7 . When the N balances for each slope category are adjusted for the proportion of the paddock they occupy, and are then summed to give an overall N balance for the notional 1 ha paddock as a whole (Table 2 .3), it appears that about 60-70% of the N input accumulates in the paddock (Table 2 . 3 ) and this accumulation is in the flat sheep campsites. Table 2 . 3 N inputs and N surplus in the notional 1 ha hill country paddock. Data on N inputs per ha and N surplus per ha are from Table 2.2 and proportions of land in each slope category are from Table 2 . 1 and Gillingham ( 1 978) Site N inputs N surplus kg N/slope category/yr kg N/slope category/yr North aspect paddock Campsite 24 x 0. 122=03 + 397 x O. I 22 = + 48 Easy slope 45 x 0.455=2 1 - 49 x 0.455 = - 22 Steep slope 1 6 x 0 .423=07 - 14 x 0.423 = - 06 TotaVl ha paddock/yr 3 1 + 20 South aspect paddock Campsite 24 x 0 . 1 22=03 + 278 x 0 . 1 22 = + 34 Easy slope 20 x 0.455=09 - 32 x 0.455 = - 1 5 Steep slope 1 8 x 0.423=08 - 1 7 x 0.423 = - 07 TotaVl ha paddock/yr 20 + 1 2 1 8 The data presented above demonstrate importance o f the spatial distribution of excreta within the paddock in determining the above ground N balance. Excretal N can also influence the overall N balance through the leaching losses from urine patches (Carran et al., 1 982; Ball and Keeney, 1 983). However, although such leaching losses are well recognised in intensive, flat land pasture systems, Sakadevan et al. ( 1 99 3 ) indicated that N appeared to be conserved in urine patches of hill country pastures in New Zealand. Lambert et al. ( 1 982a) also pointed out that hill soils are strongly N retentive because of their high C/N ratios ( 1 4 . 5 average, but 1 2- 1 7 depending on micro topography). Thus, whether the apparent N surplus observed from the above ground N balance (Table 2 . 3 ) actually eventuates wil l depend o n the fate of excretal N. Research on the fate o f urine N deposited in these l andscapes will provide important information on the N cycle in hill country pastures. 2.4 Summary of literature review Hill pastures have highly variable micro-topographic units with associated variation in pasture production and pasture composition . The physiographic factors of aspect and slope are maj or determinants of this variation through their influence on microclimate, stock grazing behaviour and soil fertility. Most studies suggest that N cycling in hill country pasture is very inefficient. Recently, some scientists have suggested improving hill country N e fficiency by differential N fertiliser application. In order to optimise this approach, there is a requirement for detailed information on N cycling in soil-plant-animal systems within the micro­ topographic units of hil l country. Estimated above-ground N balances for different topographic units (Section 2 . 3) revealed the existence of two major units in hill country pasture in terms of N c ycling. They are flat stock campsites and steep slopes. Significant quantities of N are transported to campsites from the sloped sites resulting in net N accumulation in campsites and net N loss from slopes. However, when the total paddock is considered, the above ground N balance revealed an apparent surplus of 60-65% of the N input per year. The actual extent of this apparent N surplus will depend on the fate of excretal N. The traditional assumption has been that New Zealand hill pastures are N conservative 1 9 systems, unlike intensive dairy farms that lose much o f the N input. However, little experimental evidence is available on the fate of urine N in hill country pasture to j ustify this assumption. The aim of the subsequent experimental chapters of thi s thesis is to study the fate of urine N in these hill pastures and to investigate the sustainability of the N cycle. Thus, in Chapter 4 a new technique for measuring soil mineral N using ion exchange resin is developed and evaluated. In Chapter 3 and 5 the fate of urine N applied to hill country pastures is measured at contrasting locations. In Chapter 6 a glasshouse incubation study is conducted to assess the effect of differences in soil organic matter on the transformations of urine. Finally in Chapter 7, the information derived in earlier chapters is incorporated in a quantitative model of the N cycle in sheep-grazed hill country pastures. 20 CHAPTER 3 FIELD INVESTIGATION OF THE FATE OF URINE NITROGEN ON A SUMMER-DRY HILL COUNTRY PASTURE 3.1 Introduction The literature review presented in Chapter 2 revealed that North Island hill pasture systems are characterised by steep slopes with marked aspect and micro-topographical variation. These topographical variations complicate studies of nutrient cycling in hi ll country pasture. However, most studies emphasised the dominant effect of grazing animals on the movement of nutrients through the soil-plant-animal system. Variations in land slope influence animal grazing and camping behaviour and lead to net accumulation of N in campsites and net N depletion in steep slopes. However, there are questions with respect to the efficiency of N cycling through the excreta of animal s. Though there are some studies on the fate of urine, they are mainly confined to lowland dairy-based systems and very little research has been done on hill country sheep-based systems. In recent years, Gillingham e t al. ( 1 9 98, 1 999) have questioned the cost effectiveness of large additions of P fertili sers to sunny aspects of hill country pasture where legume growth is restricted severely by moisture stress. Instead, these workers have suggested that application of nitrogen fertiliser to these areas may be beneficial. It is however necessary to understand completely, the nitrogen cycle operating in the system to recommend such idea. The fate of urine N will be a vital component of this cycle. The aim of the experiment described in this chapter is to investigate the fate of urine N in hill country pastures. 3.2 Literature review 3.2.1 Nitrogen cycling through animals 2 1 The grazing animal affects N balances in pasture by first consuming much of the herbage produced. Then the N contained in the consumed herbage is either retained within the animal, transferred to non-productive areas, or returned to the soil in a spati ally non-uniform fashion in dung and urine. For sheep grazing grass/clover pastures in New Zealand, 70-75% of the excreted N occurred in the urine (S ears, 1 950). B arrow and Lamboume ( 1 962) found that for sheep ingesting herbage containing more than 4% N, 80% of the N was excreted as urine whereas with herbage containing 0.8% N, the proportion of the excreted N present in the urine was only 43%. 3.2.2 Urine N Urine is a concentrated aqueous solution containing N, sodium (Na), potassium (K) and sulfur (S) as the dominant inorganic elements and typically has a pH value of 8 to 9 (Haynes and Williams, 1 992). Haynes and Williams ( 1 993) comprehensively reviewed the information about urine N. They reported that sheep urinate 1 8-20 times per day. A single urination is normally 0. 1 to 0. l 8 litres and the area covered by single urination is approximately 0 . 0 3 -0.05 m2. They further indicated that the N concentration in urine is around 1 0 9/litre and the equivalent application rate is 500 kg N/ha. Doak ( 1 952) described the constituents of urine N as outlined in Table 3 . 1 . Table 3 . 1 The partition of urinary nitrogen (Doak, 1 952). Constituents % Total urine N Urea N 76.4 Allantoin N 4. l Hippuric acid N 2.6 Creatine and Creatinine N 1 . 5 Amino N 1 2 .4 Ammonia N 0 . 7 3.2.3 N transformations in the urine patch 22 Urine affected areas are characterised by large inputs of biologically labile N, with transformations and transport processes operating at high rates (Carran et al. , 1 9 82). Nitrogen transformations that occur in urine patches cause large fluctuations in the pH, ionic composition and ionic strength of the soil solution in the urine patch (Carran, 1 9 8 8 ; Haynes and Williams, 1 9 92). Early work stressed the positive effects of urine N return to pasture herbage yield (Sears, 1 956). This increase in herbage yield results from a speeding up of the N cycle through the conversion by the animal of organic nitrogen in plant material into readily available N in urine. More recently, the potential for large losses of N from pasture via animals has been reported (Ball et al. , 1 979; Carran et al., 1 9 82; Ball and Ryden, 1 9 84). In New Zealand, Ball et al. ( 1 979) at Palmerston North in the North Island and Carran et al. ( 1 982) at Gore in the South Island attempted to construct a N balance in a urine patch. Fig. 3 . 1 i llustrates the urine N recovery (%) up to 32 days after urine application (DAUA), using data extracted from Ball et al. ( 1 979) and Carran et al. ( 1 9 82 ) . In the Palmerston North experiment, Ball e t al. ( 1 979) had two urine treatments (300 and 600 kg urine N/ha) and in Gore, Carran et al. ( 1 982) used one urine treatment (300 kg urine N/ha) under two moisture regimes. In Fig. 3 . 1 , data from the 3 0 0 kg urine N/ha treatment from Ball et al. ( 1 979) and the 3 00 kg urine N/ha treatment on dry plots by Carran et al. ( 1 982) were considered. Constructing a mass balance for N is not easy, as it is difficult to measure all the possible pools in the soil and all the possible pathways of N loss. In the work of Ball et al. ( 1 979) and Carran et al. ( 1 9 82) measurements were made of N taken up by pasture and lost by volatilisation. Within the soil, mineral N in the forms of � + and N03- were measured by both groups and Carran et al. ( 1 982) also measured the clay-fixed NH/-N. Because the pool of organic N in the soil is so large, changes in its size through immobilisation of urine N cannot be detected in the short term. Thus, in these studies 23 urine N that could not be accounted for would include N immobilised in soil organic matter as well as N lost by processes such as leaching and denitrification. In PaImerston North, Ball et al. ( 1 979) observed an initial steady N recovery of 70-80% which then decreased to about 50% 32 DAUA. In contrast Carran et al. ( 1 982) observed an initial increasing N recovery from 55% to >90% 1 8 DAUA, followed by a decrease in recovery to 85% of that added 32 DAUA. 100% 90% 80% � ., 70% > 0 80% � z 50% ., c "t: 40% :> 30% � 20% 10% 0% Palmerston North 4 11 18 25 32 Days after urine application IJ Unreco...,red • Plant Uptake • Nitrate ClAmmonium IJAm.vol � ., > 0 � z ., c "t: :> ;t. Gore 100% 90% 80% 70% IJ Unrecovered 60% • Plant Uptake 50% Ammonium 40% 30% • Clay Fixed 20% 10% 0% 4 1 1 1 8 25 32 Days after urine application Fig. 3 . 1 Percentage urine N recovery in urine patches. Data extracted from Ball et ai. , 1 979 (PaJmerston North) and Carran et aI. , 1 982 (Gore) . Surprisingly, 1 DAUA neither BaU et al. ( 1 979) nor Carran et al. ( 1 982) observed full recovery. I n Palmerston North, urine N recovery at 1 DAUA was approximately 70% while at Gore it was approximately 50%. Both groups commented that this initial unrecovered urine N might have been immobilised into microbial biomass. This v iew is supported by the progressive increase in recovery of urine N in Gore ( Fig. 3 . 1 ) . This increase in urine N recovery over time could be due to remineralisation of immobilised urine N . However, this was not apparent in the Palmerston North experiment (F ig. 3 . 1 ) . The volatilisation losses varied between the two sites. In Palmerston North, 1 5% of added urine N was lost as NH3 and at Gore it was 40%. 24 Mineral N changes over time were very different in the two experiments . At Gore, N� + -N was the dominant mineral N form until about 25 DAUA and nitrification wasn't a major N transformation process. However at Palmerston North, by 1 1 DAUA N03--N was the dominant mineral N form. Thus, urine N losses through leaching and denitrification could be possible N loss mechanisms in the Palmerston North experiment (Fig. 3 . l ). During that experiment Ball et al. ( 1 979) observed that all of the N03--N had disappeared by day 46. The urine N recovery by herbage was also different at the two sites (Fig. 3 . 1 ) . In the Palmerston North experiment, Ball et al. ( 1 979) observed that urine N recovery in herbage was 37% while at Gore, Carran et al. ( 1 982) observed only 1 5%. Inspection of Fig. 3 . 1 makes clear that the fates of urine N added in similar amounts in two experiments at two different sites were different. The differences observed in N transformation processes at the two sites (e.g. plant uptake, volatilisation, nitrification) are due to the differences in soil and environmental conditions prevailing at the two experimental sites . This demonstrates why it is difficult to directly apply research work done on fate of urine N in flat dairy soils to extensively-managed -sheep-based hill country pastures. 3.2.3.1 Urea hydrolysis Urea accounts for about 76% of urine N (Table 3 . 1 ) . After urine deposition to the soil, the urea is rapidly hydrolysed to NH/-N (Haynes and Williams, 1 993) . ___ .. � N�+ + COt (3 . 1 ) (3 .2) The rate of hydrolysis of urea N in urine patches varies, mainly with soil temperature and to a lesser extent with soil moisture (Quin, 1 977). However, even under unfavourable conditions urea hydrolysis is complete within a few days. Quin ( 1 977) reported that under warm moist conditions, complete conversion to NH4 + took place 2 5 within 2 4 hours. S herlock and Goh ( 1 984) calculated half-lives of urine urea as 3 and 4.7 hours, under summer and autumn conditions respectively. Haynes and Williams ( 1 993 ) indicated that urea in animal urine hydrolyses extremely rapidly after release from the animal. The urea in urine is hydrolysed by urease, a microbial enzyme that is widespread in soil (Bremner and Mulvaney, 1 978) and on plants and plant litter (Freney and B lack, 1 988). Due to urea hydrolysis, rapid rises in pH in the surface soil of up to two pH units occur within 48 hours of urination (Holland and During, 1 977). Whitehead et aI . , 1 989 reported that hydrolysis of urine urea is more rapid than that of pure urea. The major reason for this is that hippuric acid, a minor nitrogenous constituent of animal urine, has a stimulatory effect on urea hydrolysis. 3.2.3.2 Ammonia volatilisation Volatilisation of NH3 is favoured by the presence of large quantities of NH/-N and a high soil pH, according to the following equation. ( 3 . 3 ) A s noted above, under suitable conditions, enzymatic urea hydrolysis w i l l be rapid i n urine spots, creating high pH conditions, and large volatilisation losses could therefore occur shortly after urine deposition (Doak, 1 95 2 ; Carran et al., 1 9 8 2 ; Ball and Keeney, 1 98 3 ) . Carran e t a l . ( 1 982) observed that 8 0 % of total volatilisation occurs over the first 3 days. However, a wide variation in the extent of NH3 volatilisation can be expected as many soil and environmental factors are involved. Soil pH is a maj or factor affecting volatilisation. Higher pH (increase in hydroxyl ion concentration) tends to increase production of NH3. NH3 volatilisation tends to increase with increasing temperature (Lyster et al., 1 980) . Maximum loss can be expected in high temperatures where evaporation is prominent. 26 Lowest losses would occur when it is cool. Ball and Keeney ( 1 983) found losses of NH3 from urine patches that averaged 5, 1 6, and 6 6% of added urine N under cool moist, warm moist, and warm dry conditions, respectively. Haynes and Sherlock ( 1 986) pointed out that increasing wind speed should tend to increase NH3 volatilisation rate by promoting more rapid transport of NH3 away from the air-soil interface. Thus, in hill country the effect of wind speed could play a vital role . Soil moisture content has an important influence on the rate of NH3 volatilisation since it affects the concentration of NH/ and therefore NH3 in soil solution. Vallis et al. ( 1 9 82) observed relatively low NH3 loss from dry soil and speculated that the lack of water for evaporation would have restricted volatilisation. Haynes and Sherlock ( 1 9 86) pointed out that in dry soils, dissolved NH4 + may be adsorbed onto soil colloids while in initially wet soils, NH4 + might tend to remain in macropores . Then convection to the soil surface would tend to proceed through macropores, thus transporting more NH4 + to the surface of initially wet soils. However, the data of Carran et al. ( 1 982) indicated the reverse, with 20% volatilisation loss from wet plots and 40% loss from dry plots. Haynes and Sherlock ( 1 986) cited the results of several workers suggesting that increasing soil cation exchange capacity (CEC) has a negative effect on NH3 volatilisation. Haynes and Williams ( 1 993) cited a range of field measurements of NH3 volatilisation losses (9% - 46%) from urine patches. Among them, under New Zealand conditions, volatilisation losses were 1 5 - 1 8 % (Ball et al., 1 979), 1 7-36% (Carran et al., 1 982) and 1 2 -25% (Sherlock and Goh, 1 9 84). However, for hill country there are no field measurements for NH3 volatilisation losses. This is mainly because of the technical obstacles associated with hill country research. It can be seen in Plate 3 . 1 that the instruments used by Ball et al. ( 1 979) for volatilisation measurement were highly complex. This NH3 trapping system consisted of a steel frame (40 cm x 25 cm) embedded 1 0 cm into the soil immediately after urine application. Then a plastic chamber was attached to the frame. Ambient air was drawn Plate 3 . 1 : Ammonia volatilisation measurement using chamber methods (Ball et al. , 1 979; Theobald, 1 983). 27 Thus i t i s d ifficult to set up these instruments under hill country conditions. Carran et al. (2000) developed a passive sampler (Plate 3 .2) for assessing NH3 volat ilisation from difficult sites, which is quite useful in hill country studies. Plate 3 . 2 : Ammonia volatilisation measurement using passive samplers (Carran et aI. , 2000). 28 through the chamber from polyethylene tubing with the inlet on a 1 . 5 cm mast placed outside the plot area. The air was drawn through a high capacity vacuum pump. The air flow from chamber was passed through acid traps containing 1 00 mL of 0 . 2 N H2 S 04• In the absence of measured values, Lambert et al. ( 1 982) assumed 5% of excretal N for volatilisation losses from hill country. He explained that the low value was used because he b elieved that the hill country environment he studied was unlikely to be conducive to volatilisation. 3.2.3.3 Nitrification After the hydrolysis of urea, the NH/ produced can be converted to N03-. This process is called nitrification and is defined as the process whereby NH/ is oxidised via N02- to N03-. These reactions are carried out by chemoautotrophic nitrifying bacteria. In soil , fi v e genera of autotrophs are known t o b e able t o oxidize NH/ t o N02- : Nitrosomonas, Nitrosococcus, Nitros osp ira, Nitrosolobus, and Nitrosovibrio. And one genus: Nitrobacter is known to oxidise N02- to N03- (Haynes and Williams, 1 993). However, Haynes ( 1 9 8 6a) also put emphasis on the significance of heterotrophic nitrification mainly by fungi . In addition, he pointed out evidence from several workers for nonmicrobial chemical nitrification in soil . Oxidation of NlLt + to N02- by Nitrosomonas occurs as follows (Norton, 1 999). (3 .4) Oxidation ofN02- to N03- by Nitrobactor occurs as follows (Norton, 1 999). ( 3 . 5) Wrage et al. (200 1 ) described the enzymes involved in nitrification. The oxidation of NH3 to NH20H is catalysed by ammonia monooxygenase. This enzyme has a broad range of substrates for c atalytic oxidations. Thus these substrates can inhibit the NH3 29 oxidation function of the enzyme. Conversion of NH20H to N02- is catalysed by hydroxylamine oxidoreductase. Production of N03- from N02- is catalysed by nitrite oxidoreductase. Where nitrification is high, pasture growth could be limited by loss of N through leaching and denitrification, which would not occur in a soil of low nitrification activity. Consequently, nitrification is an important process to consider in a urine patch, because large quantities of NH4 + are available to act as a substrate for nitrifying bacteria. Nitrification is heavily dependent on soil and environmental factors. This is clearly exemplified in Fig. 3 . 1 where two different soils received the same amount of urine N. Nitrification is primarily affected by the NH4 + substrate supply, soil temperature, moisture, and soil pH (Gilmour, 1 984; Paul and Clark, 1 988). The autotrophic nitrifiers are dependent on either NH4 + or N02- as specific energy sources so substrate concentration can be a very important factor influencing nitrifier activity. Several studies have shown that addition of NH/ (McLaren, 1 97 1 ; Ardakani et al. , 1 974) can increase the population of Nitrosomonas and addition of N02- (Ardakani et al. , 1 973) can increase Nitrobacter. Malhi and McGill ( 1 982) studied nitrification in three Alberta soils fol lowing the addition of NH/ at concentrations of 5 0 , 1 00, 200 and 300 flg NH/-N/g soil . They observed that nitrification increased up to 200 flg NH/-N/g soil but was inhibited at 300 flg NH/-N/g soi l . The depression at 300 flg NH/-N/g soil was thought to be due to the combined effect of lowered pH, due to the NH4S04 addition, and an increase in salt content. Application of urine can increase the soil NH/-N content up to 3 90-440 flg NH/-N/g soil (Wil liams and Haynes, 2000). These large amounts ofNH/-N are surplus to plant requirements and therefore nitrifiers do not have to compete with other organisms for NH4 + substrate. The time taken to completely oxidise the NH/-N in a urine patch varies - 3 5 days in dry conditions and 2 1 days in moist conditions (Carran et al., 1 982), 2 1 days (Ball et al., 1 979), 14 days (Vallis et al., 1 982), and 40 days (Williams and Haynes, 2000). 30 Temperature is an important factor regulating nitrification (Gilmour, 1 984; Haynes, 1 986a; Bramley and White, 1 990). The optimum temperature for nitrification is usually between 25° and 3SO C (Haynes, 1986a), but nitrification can occur over a wide temperature range. Selvarajah ( 1 996) reported N03--N production in moist soils stored under refrigerated conditions. Haynes ( 1 986a) cited some evidence that indigenous nitrifiers have temperature optima adapted to their climatic regions . Under warm temperate conditions N03- i s often the major form of mineral N present in a urine patch after 3-5 weeks (Holland and During, 1 977 ; Ball et al., 1 979; Carran et al. , 1 982). In contrast, low levels of nitrate have been observed after urine application in cooler conditions (Thomas et al., 1 988). Soil moisture content and degree of aeration are fundamentally important to nitrification as autotrophic nitrifiers are strictly aerobic organisms (Bramley, 1 989). Under very high soil moisture contents, water logging limits the diffusion of 02 and nitrification is suppressed. Similarly, nitrification can be suppressed by insufficiency of water by retarding the bacterial proliferation (Paul and Clark, 1 988). Haynes ( 1 986a) mentioned that the maximum rate of nitrification occurs at soil moisture potentials in the range of - 1 0 to -33 kPa. There have been number of studies on the effect of pH on nitrification. Paul and Clark ( 1 988) suggested that optimum pH values might vary between 6.6 to 8 . In agreement with this, Sarathchandra ( 1 978) reported much greater nitrification in a Wharekohe silt loam at a pH of 7.5 than in the same soil at a pH of 5 . 5 . Bramley and White ( 1 990) carried out some interesting work in which they showed that across a range of soils the optimum pH for nitrification was generally close to the pre­ existing soil pH, suggesting that the indigenous nitrifier populations adjusted to the prevailing soil pH. Haynes and Williams ( 1 992) studied urine transformations in urine patches and observed that nitrification was occurring across a range of pH values down to a pH of at least 5 . Monaghan and Barraclough ( 1 992) reviewed the effect of urinary cr, osmotic potential and N content on the rate and dynamics of nitrification in urine-affected soi l . They 3 1 concluded that in most temperate grassland soils at near neutral pH, urinary er and N are unlikely to reduce nitrification rates, except where urine N concentrations exceed 16 g N/litre . There is some evidence of an effect of clay minerals on nitrification. Soils containing high amounts of allophanic minerals have been found to enhance the nitrification process (Baber, 1 978; Sarathchandra, 1978). Although the exact reason for this is unclear, Selvarajah ( 1 996) regarded this as an important observation and indicated that Waikato soils have a greater potential for generating N03- than other soils in New Zealand. Nitrification activity and the number of nitrifying bacteria were measured in nine New Zealand soils by Sarathchandra ( 1 978) as Short Term Nitrification Activities (SNA). This was the mean rate of nitrification between 1 and 1 7 hours, from 5 g of oven-dry soil perfused with 50 mL of (NH4)2S04 ( 14 mg/l). The nitrification activities of nine soils were widely different ranging from 0.25 to 3 .3 1 Ilg N03--N producedlg soillhour. Steel et al. ( 1 980) investigated the nitrification activity of soils from sixty-eight sites in New Zealand grassland. Nitrification activity was measured as Initial Nitrification Activity (INA), which was the mean rate of nitrification between 1 and 1 7 hours, in 1 0 g of oven dry soil mixed with 20 g of acid washed, autoclaved quartz sand, perfused with 1 00 mL 0.005M (NH4)2S04 previously adjusted to the pH of the soil measured in a 1 :2 .5 soil : water suspension after 1 6 hours equilibration. Similar to Sarathchandra ( 1978), Steel et al. ( 1 980) also observed considerable variation « 0.02 to 5 .7 Ilg N oxidisedlg soillhour) of INA among the studied soils. They observed four general patterns of nitrification (Fig. 3 .2) In Type 1 , NH/-N rapidly oxidised to N03--N and the rate of oxidation was linear. Type 1 nitrification was observed in soils with high INA values, which were yellow brown and red and brown loams. In Type 2, nitrification was linear but at a very much slower rate compared to Type 1 . Type 2 nitrification, occurred in four yellow brown earths, one of the yellow grey earths, and one of the yellow brown pumice soils. In Type 3 soils, NH/-N was oxidised slowly to N03--N at a rate that increased 32 exponentially with time after an initial lag phase. Most soils exhibited Type 3 nitrification. In Type 4 soils, nitrification occurred with a temporary accumulation of N02- at the start. This pattern was observed only in soils of high pH. Type 2 600 - � 400 (71 :t - ' ''' Type 4 0 Type :3 z 600 � 4) - 0 400 � I NO'2 E � 200 0 • N O"5 (,) < 0 0 lOO 200 300 100 200 300 Hours perfusi on Fig. 3 .2 Four general patterns of nitrification observed by Steel e t al. ( 1 9 8 0) when soils are perfused with .005M (NH4hS04 . Fig. 3 .3 and 3 .4 illustrate the relationships of INA to total N (%) and C/N ratio developed from the data presented in Steel et al. ( 1 980). When soils exhibiting Type 1 and Type 2 nitrification behaviour are considered alone (Fig. 3.3) it appears that soil total N (%) and C/N ratio have strong rel ationships with nitrification. However, when all the data presented in Steel et al. ( 1 9 8 0) are considered, the general pattern of the graphs is similar though the relationships are not very strong (Fig. 3 .4). � " o 3.0 .,..---------------, 2.5 y = -5.65x' + 10.75x - 2.65 R' = 0.83 § 2.0 o VI Cl> � � 1 .5 - .. � j( � 1 .0 Cl> :1 0.5 0.0 +-. --II�I__,_-_,.._-_,.._-__,_---1 0.00 0.25 0.50 0.75 1 .00 1.25 1 .50 Total N (%) 3.0 .,..-------------, � " o 2.5 § 2.0 o .. « E! 2 1\ 1 .5 - .. � j( � 1 .0 Cl> :1 0.5 y = -0.34x + 5.41 R' = 0.81 0.0 +----r-_,.._-..-----.-...!.:......, ...... ..:. • ..--� o 3 6 9 12 15 18 21 C/N ratio 33 Fig_ 3 .3 Relationships between INA (Initial Nitrification Activity) and total N (%) and CIN ratio of soiL The data are extracted from soils that showed the Type 1 and Type 2 nitrification patterns of Steel et al. ( 1 980)_ 3.5 -.-------------, 3.5 -.-----------� .... 3.0 Y = -1 . 10x2 + 2.74x - 0.25 3.0 ,; = -1 . 35Ln(x) + 4.23 :l • .... 0 R2 = 0.26 :l R2 = 0.20 :§ 0 2.5 :§ 2.5 '0 • • • '0 •• ", '" � 2.0 • • • '" � • « � 2.0 « '0 • � Cl) • � � • • '" • • :c 1 .5 '" 1 .5 • • :c )( )( 0 0 z 1 .0 z 1 .0 Cl Cl :l. :l. 0.5 0.5 • • • 0.0 0.0 0 0.5 1 .5 2 2.5 3 0 4 8 1 2 1 6 20 24 28 32 Total N (%) C/N ratio Fig_ 3.4 Relationships between INA (Initial Nitrification Activity) and total N (%) and CIN ratio of soiL All the data presented in Steel et al. ( 1 980) were used for the relationships. 34 According to the graphs, nitrification activity increased up to total N contents of about 1 % and then decreased. Soils with total N contents beyond this level are mostly peaty soils. Although they have a high N content, the N is unavailable to microbes due to a high C/N ratio. This may retard the nitrifier population. Hence, substrate quality is a maj or determinant of nitrification. Lambert et al. ( 1 982a) pointed out that hill country soils have high C/N ratios ranging from 1 2- 1 7. Also, in Chapter 2, N depletion in steep hill soils due to animal transfer was discussed. In steep soils, the main C source is from decomposing plant materials and roots. Root materials have a rel atively high C/N ratio. Thus, with low N levels plus a high C/N ratio in steep hill soils, it can be anticipated that there will be relatively low nitrification. This might be the reason that S akadevan et al. ( 1 993) didn't observe any accelerated leaching of N03--N in urine patches on hill country soil. 3.2.3.4 Leaching Mineral N concentrations occurring in urine patches are often in the range of 1 00-440 �g N/g soil (Ball and Ryden, 1 984; Thomas et al., 1 98 8 ; Haynes and Williams, 1 992; Williams and Haynes, 2 000). Once most of this mineral N has nitrified, leaching of N03--N will occur when water moves down through the soil after excess rainfall. Thus, leaching could be a maj or N loss mechanism responsible for the unaccounted-for urine N discussed in Fig. 3 . 1 . Both Ball et al. ( 1 979) and Carran et al. ( 1 982) commented that leaching of N03--N could be one of the probable loss mechanisms. The movement of N03--N in soil i s affected by a large number of physical, chemical, and microbiological processes. Hence, a range of experimental, mathematical and physical sciences are required to study and describe solute transport in soi l . Transport of a dissolved substance (solute) depends on the magnitude and direction of the solvent (water) flux. Cameron and Haynes ( 1 986) reviewed the principles of solute movement in relation to the process of N03- leaching. They summarised the main transport processes as being convection (solute transport due to mass flow of water alone), diffusion (solute movement from areas of high concentration to areas of low concentration) and dispersion (mechanical action of a solution fl owing through soil that 3 5 causes mixing and equalises the solute distribution). In addition, they paid attention to anion exclusion, anion adsorption, and transformations and plant uptake of nitrogen. The combined effects of convection-diffusion, dispersion and the rate of N production and disappearance can described b y where c = concentration of N03- (llg/mL ) t = time (days) U = average pore velocity (cm/day) x = linear distance in direction of flow (cm) E = dispersion coefficient (Ds+ mU) Ds = effective diffusion coefficient in soil (cm2/day) m = dispersivity S = index for rate of N03- production or disappearance ( llglmL /day) ( 3 . 6) Bums ( 1 975) developed an equation to predict the leaching of surface applied nitrate. Scotter et at. ( 1 993) identified the usefulness of the Bums equation and also the problems associated with it. They modified the equation for use in a wider range of conditions as follows. x = exp(-zB / 1) x = fraction of solute leached below depth z z = depth f) = volumetric water content at field capacity I = net rainfall (3 .7) Season and climate play a maj or role in l eaching. Summer rainfall is generally used to satisfy the evapotranspiration deficit and leaching is therefore usually minimal. Winter rainfall readily leaches any nitrate present in the soil profile since there is excess of 3 6 rainfall over evapotranspiration and low plant N uptake. Also, a dry summer can result in the accumulation of soil nitrate due to poor pasture N uptake and significantly higher than average leaching losses then occur over the subsequent winter (Cameron and Haynes, 1 986). Steel and ludd ( 1 984) estimated that leaching losses from an intensively grazed pasture over one year were 88 and 1 93 kglha for a no N fertiliser treatment (control) and a plus N treatment (3 applications of 5 7 . 5 kg /ha as urea) respectively. The direct impact of the grazing animal on N03 - leaching was demonstrated by Ryden et al. ( 1 984). They observed an annual loss of N03--N of 29 and 1 62 kg Nfha/yr from cut swards and grazed grass swards (both receiving 420 kg Nfha/yr) respectively. Thus, the leaching losses from the grazed sward were 5 .6 times greater than from the cut sward. They pointed out that the enhanced loss of N03--N below the grazed sward was mainly due to return in urine and dung which accounted for as much as 90% of the N in the herbage consumed by cattle. Grazing animals deposit dung and urine in localised patches at average rates equivalent to 500 and 1 000 kg Nlha/yr for sheep and cattle respectively (Haynes and Williams 1 993). These rates are in excess of plant demands and can l ead to significant N leaching as Ryden et al. ( 1 9 84) pointed out. Quin ( 1 979) demonstrated that 80% of the urine urea-N was eventually l eached as N03--N from a typical urine patch under surface irrigation. Ball et al. ( 1 979) observed considerable amounts of applied urine N lost from the 0-45 cm soil profile after 5 3 days and indicated that leaching of N03--N was the probable loss mechanism. In contrast, Fraser et al. ( 1 994) in a lysimeter study observed only 8% of the applied urine 1 5N leached below 1 200mm after 1 year, although the experiment was conducted under intense rainfall and irrigation conditions. This may be due to the presence of a low hydraulic conductivity layer in the monolith at 20-30 cm depth. This could have led to anaerobic conditions leading to high denitrification losses which they have identified as a probable loss mechanism of 28% of the added urine N. There are few quantitative estimations in the literature of leaching losses from hill country. This is mainly because of the technical obstacles associated with hill country 3 7 measurement. However, scientists generall y believe that the leaching loss from hill country is minimal because the soils are strongly N retentive (Lambert et al., 1 982a). This view was confirmed b y Sakadevan et al. ( 1 993) who did not observe any accelerated leaching ofN from urine-affected soil in hill country pasture. 3.2.3.5 Denitrification Microbial reduction of N03- to N02- and then to gaseous N20 and N2, which are commonly lost to atmosphere, is known as denitrification. Ryden ( 1 986) pointed out that denitrification also can occur in urine patches. Theoretically the potential for denitrification from pastures would appear to be high due to high amounts of organic C in the surface soil and high concentrations of N03--N present in the soil under urine and dung patches (Haynes and Williams, 1 993). Sherlock and Goh ( 1 9 8 3 ) suggested that any attempt to quantify global N20 production should recognize the possibility of enhanced N20 production from urine patches in grazed grassland. In contrast, Carran et al. ( 1 9 9 5 ) reported that nitrification or other transformations of urine derived N do not contribute to overall emissions of N20 in an important way. In the same way, Luo et al. (2000) reported an annual nitrogen loss of only 4 . 5 kg Nlha through denitrification from a legume based intensive dairy pasture in New Zealand. 3.2.3.6 Mineralisation and immobilisation Mineralisation is the general term for the conversion of organic N to inorganic N as either N1Lt + or N03 -. Immobilisation is the conversion of inorganic N to organic N. This can be mediated by microorganisms or by chemical interactions with complex carbon compounds. Direct immobilisation of N in the soil organic matter provides a mechanism whereby urine N is retained to be again recycled. The rate of incorporation of N into the microbial biomass and other soil organic fractions will depend on factors influencing the mineralisation/immobilisation turnover. These factors include the C :N ratio of organic residues and environmental conditions. 3 8 The incorporation o f N into microbial biomass and organic N occurs through numerous enzymatic and abiotic pathways. Norton ( 1 999) reviewed the microbial N immobilisation process extensively. The preferred inorganic N source for assimilation by bacteria and fungi is NH3INH/ although N03- is also used under appropriate conditions. Ammonia enters microbial cells by rapid diffusion across cytoplasmi c membranes . Biological NH4 + immobilisation b y soil microorganisms i s mainly accomplished by two enzymatic pathways. They are the Glutamate dehydrogenase (GDH) system and the Glutamine synthatase (GS) - Glutamate synthase ( GOGA T) system. In soils with low NH4+-concentrations the GS-GOGAT system is operative and the GDH pathway immobilizes N at relatively high concentrations. The two systems are shown below as described by Moat and Foster ( 1 999) . GDR system a- ketoglutarate + NH/ + NADH + H+ +-+ L-glutamate + NAD+ a- ketoglutarate + NH/ + NADPH + H+ +-+ L-glutamate + NADP+ GS-GOGA T system ( 3 . 8 ) ( 3 . 9 ) L-glutamate + N H/ + ATP GS � L-glutamine + ADP + P i (3 . 1 0) a- ketoglutarate + L-glutamine + NADPH + H+ GOGA T � 2 L-glutamate + NADP+ (3 . 1 1 ) Nitrate may be immobilised directly by assimilatory nitrate reduction by both bacteria and fungi (Paul and Clark, 1 988). (3 . 1 2 ) The enzymes responsible for reduction are assimilatory n itrate reductase and assimilatory nitrite reductase. Broadbent and Tyler ( 1 9 62) observed nitrate immobilisation when N03- was the only form of N avail able. Rice and Tiedj e ( 1 989) also observed nitrate assimilation when there was a demand for N (excess C) and low concentrations of NH/. They also observed the inhibition of N03 - assimilation by N H/ . Ammonium inhibited the N03- assimilation immediately « 1 min) and when present in very low concentrations (0. 1 J.lg N/g soil). 3 9 Urine patches i n a soil with high nitrification activity nonnally contain both NH4 + and N03- ions. Thus, N assimilation by microbes could take place by uptake of either N� + or N03-. As N H/-N is the preferred source, immobilisation in urine patches would nonnally occur in the initial days after urine application because N� + -N is the dominant N fonn at that time. Keeney and MacGregor ( 1 978) observed that apparent l sN immobilisation was more rapid in urea and (NH4)S04 treated plots compared to KN03 treated plots . Soils with very high nitrification rates could produce high amounts of N03- ions within a day or two after urine application. Assimilation of this N03--N by microbes could b e inhibited b y the presence of N H/-N. Once most of the NH/-N has been nitrified, the N03--N could be assimi lated by microbes . Broadbent ( 1 9 6 5 ) observed an increased rate of soil organic N mineralisation with higher initial concentrations of added 1 5N-(NH4hS04 . This is called a "Priming effect" or added nitrogen interaction (Molina et al., 1 990) . Several studies have observed the priming effect due to added fertiliser or salts. (Agrawal et al., 1 97 1 ; lenkinson et al. , 1 98 5 ; Molina et al., 1 990; Leon et al., 1 995). However, the actual cause for this so called priming effect is not clear. Broadbent and Nakashima ( 1 97 1 ) postulated that osmotic effects contributed to the salt­ stimulated mineralisation of soil organic N, partially due to the extraction of organic N by the salt solution. The organic N thus rendered soluble would constitute a pool o f easily mineralisable N. Another possible contributing factor could be that the high salt concentrations result in the death and breakdown of microbial cells with the release o f readily available N , for subsequent mineralisation b y the remaining microbial population (Haynes, 1 98 6b). B al l and Tillman ( 1 994) suggested stripping of resident organic N from urine patches. They pooled the data from several studies, and viewed different seasonal runs as replications over time and observed a sign ificant decline in total N in topsoil, which had received urine. They commented that urine had imparted some "priming effect", thereby stripping some of the resident organic N from the soil, during the period when its benefit to plant growth was being perceived aboveground. 40 3.2.3.7 Plant uptake Plant uptake is another critical mechanism whereby urine N is recycled through the system. The two major forms of nitrogen taken up by plants are N03- and NH/. Although the most common form of N adsorbed by most plants is N03-, in acidic soils NH4 + predominates because nitrification is largely inhibited (Pilbeam and Kirkby, 1 992). Haynes ( l 986c) pointed out that the majority of plant species produce the greatest growth when the soil has a mixture of N03- and NH/. Pilbeam and Kirkby ( 1 992) also cited similar evidence of increased growth rate when both forms of nitrogen are supplied simultaneously. Reisenauer et al. ( 1 982) observed maximum yield of ryegrass by suppl ying low levels of NH4 + (36 pM) with adequate N03 - (72 pM). The reason for the stimulating effect of N� + on the growth of plants supplied with N03- is not clear. However, the total energy requirement for N uptake is less when both forms are taken up than when N03- is taken up alone because N03- uptake needs extra energy to be reduced to � + after uptake (Pilbeam and Kirkby, 1 992). There is some evidence of restriction ofN03- uptake by the presence ofNH/ (Youngdahl et al. , 1982; Pilbeam and Kirkby, 1992). In a urine patch, the abundance of NH4 + and N03 - varies with time. Ammonium is predominant immediately after urine application and later N03- become prominent in soils with high nitrification activity. In high nitrification activity soils, both forms of mineral N can be found up to about four weeks after application. In low nitrification soils, NH/-N is the dominant form in a urine patch for a longer period. These variable combinations of mineral N forms existing in a urine patch could be the reason for the resulting highly variable urine N recoveries by pasture herbage in previous studies. It has been observed that the urine N recovery by herbage ranges from 8-5 5% (Holland and During, 1 977; Ball et al., 1 979; Carran et al., 1982; Ledgard et aI . , 1 982a; Ball and Keeney, 1 983; Thomas et al., 1988) Many workers in New Zealand have observed increased pasture growth in urine patches (Ball et al., 1 979; Carran et al. , 1 982; Sakadevan et al., 1 993 ; Williams and Haynes, 4 1 2000; Theobald and Carran, 2000) . Dale ( 1 96 1 ) noted that response is greatest in spring and autumn. The responses to urine N in midsummer and winter are generally restricted by environmental conditions (i .e. , low soil moisture and low temperature, respectively) that are not conducive to rapid pasture growth. Many workers have observed that N fixation by the clover component of the sward is markedly depressed in urine patches (Ball et aI. , 1 9 79; Ledgard et al. , 1 982a). This is due to the inhibitory effects of high soil mineral N on N2 fixation by the Rhizobium bacteria (Haynes and Williams, 1 993). 3.2.3.8 Fixation to clay minerals N H/-N is generally held in soils as an exchangeable cation, but in soils that contain 2 : 1 c lay m inerals (e.g., illites, vermiculites, and montmorillonites) it can also be held in a non-exchangeable "fixed" form. In a Waimumu (Southland) soil, Carran et al. ( 1 982) (Fig. 3 . 1 ) reported that fixed NH/-N concentrations increased from 42 to 85 mg/kg in plots treated with urine. They regarded NH4 + fixation as environmentally desirable, as the NH4 + ions were quickly removed from soil solution, thereby inhibiting losses, and were then released slowly. Crush and Evans ( 1 9 8 8) measured fixed NH/-N in artificial urine patches in four Manawatu soils containing high quantities of 2 : 1 minerals. They found that less than 8% of urine N applied in spring and less than 3% in summer were fixed to clay minerals. 3.3 Materials and methods 3.3.1 Field site This experiment was carried out at the AgResearch hill country research site near Waipawa, North Island, New Zealand. Waipawa is approximately 300 m above mean sea level and receives an annual average rainfall of about 800 mm, most regularly falling in winter and early spring. Warm, dry summers and much cooler winters are typical of the site. The maj or soil type at the experimental site is Waipawa Stony Silt Loam (Pallic Soil) which has a low water holding capacity. The soil is low in P, S, and 42 N (Gil lingham et al., 1 998). The pastures of the steep northerly facing slopes contain low proportions of clover due to the regularly dry summer conditions. 3.3.2 Field trial design The experiment was carried out at a site within an ongoing AgResearch trial. The AgResearch trial occupies 48 ha divided into 4 farmlets of 1 2 one-hectare paddocks each receiving contrasting P and N fertilizer treatments. This experiment was carried in a north-facing paddock receiving high phosphorus and no nitrogen (Paddock # HP2 5 ) . Two experimental sites; a steep sloping site ( 1 5-25°) and a lower fl atter sheep campsite (0- 1 5 0) were selected from the paddock for this experiment. Some soil fertility indices of the two sites are listed in section 3 . 5 . 7 (Table 3 . 9). The whole paddock was grazed by sheep b efore the plots were selected. The locations of recent urine patches were checked in the experimental plot area by a field pH meter as increased pH values could be observed immediately after urine deposition due to urea hydrol ysis (Section 3 .2.3 . 1 ) . These patches were avoided in selecting the experimental plots. In each experimental site, twelve (0.5m x I m) plots were established and arranged into 4 b locks of 3 plots . Three treatments were assigned at random within each of the 4 blocks to give a randomized complete block design. The plots were established with enough space around the edges to prevent interference from other treatments b y runoff etc. 3.3.3 Treatments A synthetic unne solution (Table 3 .2) was used in the experiment and was made according to Muller ( 1 995). The N application rate in urine patches greatly varies depending on diet and animal. Therefore a range of application rates has been found in previous experiments ; 300 and 600 kg Nlha (Ball et al. , 1 979), 300 kg N/ha (Carran et al. , 1 982), 280 kg Nlha ( Sakadevan et al., 1 99 3 ) . Hence, in this preliminary experiment three N rates (0, 200 and 400 kg Nlha) were used as treatments (Table 3 . 3 ) to represent the wide range of application rates found in the literature. However, these application rates were further reviewed in next experiment (Section 5 .2.3) . The 200 and 400 kg 43 Nlha treatments were applied as 2 liters of 0 . 5 %N and 1 % N urine solution respectively. The 2 l iters of urine solution ( 0 . 5%N or 1 %N) was added evenly by a small watering can to the appropriate plots. The application rate was such that there was no visible surface runoff. In natural conditions, urination adds both water and nutrients to the soil. In the current experiment the application rate of water was broadly comparable to that suppl ied by urine. Some consideration was therefore given as to whether the control plots should have 2 liters of water added- so as to isolate the effect of adding N. In the end however, it was decided that the control plots would receive nothing. Table 3 . 2 Constituents of the synthetic urine solution (PH= 7 . 8). Constituent 0.5% N solution (�/l) 1 % N solution {gll) Urea 9 . 8 1 9. 6 Glycine 2.7 5 .4 KC} 2.6 5 . 2 K2S04 1 .0 2 . 0 KHC03 7 . 5 1 5 .0 Table 3 . 3 . Treatments used in the experiment. Rate Treatment N kg/ha Flat 0 FO 200 F200 400 F400 * Apphed as 2 hters of O . 5 %N artIficlal unne so lutlOn ** Applied as 2 liters of a 1 %N artificial urine solution 3.3.4 Soil and plant sampling Steep SO S200 S400 Soil samples were collected 1 , 6, 27, 1 00, and 1 42 days after urine application (DAUA). At each sampling, five soil cores (25 mm diameter) were taken from each plot and bulked into a polythene bag to provide a plot sample. S amples were taken from 0-75 mm and 7 5 - 1 5 0 mm depths. Soil samples were taken from a 0.25 m2 area in the down­ slope half of the plot and the rest of the area was kept undisturbed for pasture yield 44 measurements. The first soil core samples were collected from near the bottom of the slope and the next sampling taken from further up slope to minimize the effect of early samplings on subsequent measurements. All soil samples were kept in a refrigerator at 4°C to minimize microbial action, and samples were analysed within a day. A sub­ sample of soil was used to determine soil moisture content by drying at 1 05°C for 1 6 hours. Soil cores in each plot sample were crumbled and well mixed, then 4 grams of field moist soil was weighed into a 50 mL centrifuge tube. Thirty mL of 2 M KCl was then added and the tubes shaken for 1 hour. The mixture was centrifuged at 9000 rpm for 2 minutes, then the solution was decanted off and filtered through Whatman No. 41 filter paper. Extracts were analysed NH/ -N and N03--N contents colorimetrically by a Technicon auto-analyser (Searle, 1 975; Blakemore et al. , 1 987). Herbage yields were taken from an area of each plot undisturbed by soil sampling. Herbage cuts were made using a hand cutter at 27 and 1 03 DAUA. Dry matter yield was measured by drying at 60°C and weighing. The total N content of herbage was determined by a semi-micro kjeldahl method. A herbage sample (O. l g) was digested with 4 mL of kjeldahl digest acid (McKenzie and Wallace, 1 954) in a 1 00 mL test tube in a drilled aluminium block (350 °C) for 7 hours. The digested solution was allowed to cool down, diluted to 50 mL with distilled water, and mixed using a vortex mixer. The NH/-N content of the solution was analysed colorimetrically by a Technicon auto analyser (Searle, 1 975 ; Blakemore et al., 1 987) and expressed as the total herbage N. Soil potentially mineralisable N content was measured using an anaerobic incubation (SparIing and Schipper, 1 999) from soil samples collected at 27, 1 00, and 1 42 DAUA. The incubation carried out was similar to the method used by Blennerhassett (2002) . Five g of field moist soil was weighed into a 50 mL centrifuge tube, 20 mL of distilled water was added, tubes were sealed and incubated for 4 weeks at 30°C. After the incubation period, N was extracted by adding 1 0 mL of 3M KCI solution and shaking for 1 hour. Extracts were centrifuged at 9000 rpm then filtered (Whatman 4 1 ) and the NH/-N contents of the extracts were analysed colorimetrically by a Technicon auto 45 analyser ( Searle, 1 975; Blakemore et al., 1 987). The difference b etween NH/-N content before and after the incubation was considered as soil mineralisable N. Total C and N contents of control soils were determined by combustion using a Leco FP-2000 CNS analyser. 3.3.5 Ammonia volatiIisation Ammonia volati lisation after unne application was estimated usmg the method described by Carran et al. (2000). Two cylindrical polycarbonate vials with internal dimensions 2 1 x 90 mm (Plate 3 . 1 ) were placed in the middle of each p lot to assess the effect of urine on ammonia volatilisation. Inside the polyc arbonate vials three layers o f stainless steel mesh (0 . 5 mm) were secured against the bottom o f the tubes. The mesh was loaded with oxalic acid by dipping in a solution of 1 0% oxalic acid in acetone and drying. The each vial can trap a maximum of 2.6 mg NH3-N (C arran, P ersonal communication). After exposure in the field for 6 days, tubes were recapped until analysis. Analysis involved adding 1 0 mL distilled water and 0. 1 mL of 1 0M NaOH to each tube and determining the NH3-N concentration with an NH3 gas-sensing electrode. Carran (Personal communication) has developed the following relationship between ammonia volati lisation values obtained from the method used in this experiment and the amount of ammonia volatilised on a per hectare basis. NH3-N Volatilised (kg Nlha) = (0.8) ()lg NH3-N/sampler) (R2 = 0.6) (3 . 1 3) To develop this relationship, the current method and a quantitative method (Schoj erring et al., 1 992) were used to measure ammonia volatilisation from a range of urine application rates on a Manawatu silt loam. These measurements were made over a period of months from late summer to winter. The above-mentioned relationship was used to estimate the amount of ammonia volatilised on a per hectare basis from this experiment in the absence of any more directly relevant information. 3.3.6 Statistical analysis 46 Analysis of variance was carried out using SAS for Windows (Version 8). The 2 M KCl extracted NH/-N, N03--N (0- 1 5 cm soil depths), total mineral N (0- 1 5 cm depth) and mineralisable N data were analyzed using a repeated measures model in mixed procedure (Little et al. , 1 998) to examine and compare response trends over time. Autoregressive co-variance structure was used in the model. Means were compared using the Least Mean Square Multiple Comparison method at 5% significant level. Analysis of variance for data on NH3 volatilisation, herbage DM, and herbage N accumulation was carried out using the General Linear Model (GLM) procedure. Mean compansons were done using Fishers Least Signi ficant Difference (LSD) at 5 % significance. 3.4 Results 3.4. 1 Rainfall The rainfall data was obtained from the Waipawa AgResearch research site. The total amount of rainfall during the experimental period was 1 8 1 mm (Fig. 3 . 5 ) . 45 40 35 30 ;.. co � E 25 E "i 20 ... c: "i Cl: 15 10 5 0 0 10 20 30 40 50 60 70 80 90 100 1 10 120 130 140 1 50 DAUA Fig. 3 . 5 Daily rainfall during the experimental period. 47 There was little rainfall during the first 30 days. At 3 9 DAUA heavy rainfall (42 . 5 mm) occurred. During the period of 27 to 1 00 DAUA, frequent and substantial amounts of rainfall ( 1 3 0 mm) were received. 3.4.2 Statistical data interpretation The repeated measures model in a mixed procedure made it possible to compare the treatment means at a p articular sampling time as well as treatment means between sampling times within a site . No comparisons were made between sites, as the site locations were not replicated appropriately. The least squares mean procedure for mean comparison supplied mean comparisons for all possible comparisons . This caused some difficulties in tabulating the statistical data. Thus, the existence of statistically significant differences at specific sampling times are pointed out in Fig. 3 .6, Fig. 3 . 7 and Fig. 3 . 8. When there i s at least one statistically significant difference between treatments at a particular sampling time it is indicated by " * " and " NS " is used when no significant treatment differences were observed. However, there was no convenient way of indicating exactly which treatments were significantly different from which at each sampling time. This detailed information is tabulated in Appendix 1 . When necessary, any important statistically significant differences within and between samplings are identified and discussed in the text. 3.4.3 Mineral nitrogen M ineral N comprises N H/-N and N03--N. These two components are discussed separately in S ections 3 .4.4 and 3 .4 . 5 . This section provides general comments on the combined quantities of N H/-N and N03--N extractable by 2 M KCI in the soils. Mineral N in soil, to a depth of 1 5 cm, was significantly increased (Table 1 of Appendix 1 ) at both sites, after urine application. Unsurprisingly, these increases were higher when urine was applied at 400 kg Nlha than applied at half the rate, 200 kg Nlha, at both the sites (Fig. 3 . 6). For example, the soil (0- 1 5 cm) mineral N levels were increased 8, and I I -fold relative to control at 1 DAUA for the S200 and F200 4 8 treatments while the mineral N levels in the S400 and F400 treatments increased 1 5 and 3 3 -fold respectively. The elevated mineral N levels in urine treated plots decreased with time. As a result, only 3 6- 5 3 % of mineral N found at 1 DAUA was present in the soil by 27 DAUA. B y 1 00 DAUA, the mineral N levels i n urine treated plots were not statistically different from the controls. This pattern of mineral N change after urine application was similar at both flat and steep sites. 600 550 500 IV ,:: 450 III .It Z 400 ii � 350 C E 300 r 250 an "":' 200 2- :g 1 50 III 100 50 0 0 . . • . . FO . . • . . F200 . . • . . F400 -O-SO -Cr-S200 -o-S400 NS 10 20 30 40 50 60 70 80 90 1 00 1 10 120 1 30 140 150 Days after urine application Fig. 3 . 6 E ffect of urine application on soil (0- 1 5 cm) m ineral N level. * = significant treatment differences were observed within a sampling at the same site. NS = no significant treatment differences were observed within a sampling at the same site. At I DAUA in the F400 urine treatment, the urine N recovery as mineral N was greater than 1 00% (Table 3 .4) suggesting a possible priming effect. The mineral N recoveries in the other urine treatments were also close to 1 00% (Table 3 .4). Soon after the urine application, movement of urine N to the 7 . 5 - 1 5 cm-soil depth was evident (Table 3 .4). This is discussed in more detail in S ection 3 .4 . 4 . 49 Table 3 .4 Apparent recovery of urine N as mineral N and mineralisable N after urine application (kg NIha/7 .5 cm depth) in the soil profile (0- 1 5 cm) i .e . all values are treatment minus control. * = did not measure Soil depth Treatment DAUA 0-7.5 cm 7.5- 1 5 cm % recovery in soil NH.+ -N NO)' -N Mineralisable NH: -N NO)' -N Mineralisable N N S200 1 1 39 1 * 5 1 1 * 96 6 1 7 1 2 * 1 6 0 * 95 27 50 2 0 4 9 34 50 1 00 1 0 0 0 0 5 3 142 0 0 0 0 0 3 .5 2 S400 1 348 1 * 45 1 * 99 6 328 2 * 2 1 0 * 88 27 1 39 2 5 1 22 2 7 56 100 2 1 1 9 2 0 29 1 3 142 2 0 1 2 1 0 14 7 F200 I 156 1 * 14 0 * 86 6 1 1 7 1 * 1 1 0 * 65 27 76 4 44 6 0 0 65 100 6 0 35 6 1 1 6 32 142 0 0 9 1 0 4 7 F400 1 462 2 * 72 1 * 1 34 6 359 3 * 39 I * 1 0 ! 27 140 24 59 1 8 3 0 6 1 1 00 2 1 39 2 1 28 1 8 142 8 1 1 7 1 0 22 1 2 At the last two samplings ( 1 00 and 1 42 DAUA), very little urine N was recovered as mineral N but reasonable amounts were recovered as mineralisable N (Table 3 .4 and Section 3 .4.6). 3.4.4 Ammonium 50 The amount of 2 M KCI -extractable soil NH/-N ()lg NH/-N/g soil) to a depth of 1 5 cm followed a similar pattern to the mineral N changes after urine application (discussed in the previous section). This was because NH/-N was the dominant form of mineral N in both treatments at both sites over the whole experimental period (Table 3 .4). Soil (0- 1 5 cm) NH/ -N contents were significantly increased (Table 2 of Appendix 1 ) by urine treatments in both sites (Fig. 3 . 7). Steep site 600 -,---------., "f 500 i! ! 400 :a $ $ � =5, 300 :: "" ;: U 200 " � 100 DAUA NS NS ......... 50 ""'-5200 -+-5400 "f i! z . :a • $ 600 500 400 � :5,300 :: "" ;: 200 U " � 100 Flat site -o-FO -o-F200 -o-F400 NS � 0 25 50 75 100 125 150 DAUA Fig. 3 . 7 E ffect of urine treatments on soil (0- 1 5 cm) NH/ -N. * = significant treatment differences were observed within the sampling time. NS = no significant treatment differences were observed within the sampling time. At 1 DAUA, 1 90, 393, 1 70, and 534 kg Nlha were recovered as NH/-N in soil (0- 1 5 cm) (Table 3 .4) from the S200, S400, F200 and F400 treatments respectively. These amounts were 9 5 %, 98%, 85% and 1 3 3 % of added urine N for the S200, S400, F200 and F400 treatments respectively. The large amount of NH/-N observed 1 DAUA indicates rapid urea hydrolysis after urine application. In addition, the > 1 00% recovery in the F400 treatment suggests again a possible priming effect . 5 1 The elevated NH/-N levels in urine treatments then declined with time. Within the period of 1 -27 DAUA, 254 and 390 kg NH/-N/ha disappeared from the S400 and F400 treatments respectively. These amounts were 6 1 and 7 1 % of soil NlLt + -N present at 1 DAUA in the S400 and F400 treatments respectively. Thus, the rate of soil NH/-N disappearance was faster in the flat site than the steep site in the 400 kg urine Nlha treatment. In contrast, in the 200 kg urine N/ha treatment, the rate of soil NlLt +-N disappearance was faster in the steep site than the flat site. From 1-27 DAUA, 1 6 1 and 1 02 kg soil NH/-N/ha disappeared from the S200 and F200 treatments respectively. These amounts were 74 and 54% of soil NH/-N present at 1 DAUA in S200 and F200 treatments respectively At 1 00 DAUA, the soil (0- 1 5 cm) NH/-N levels in urine treatments had reached the levels of the controls. Downward movement of NH/-N occurred immediately after urine application. By 1 DAUA, 26%, 12%, 7% and 1 8% of the added urine N in the S200, S400, F200 and F400 treatments respectively, was recovered as NH/-N in the 7 .5- 1 5 cm soil depth (Table 3 .4). This is probably due to preferential flow through soil macro-pores, as there had been no rainfall on the first day (Fig. 3 . 5). 3.4.5 Nitrate After urine application, the 2 M KCI -extractable N03--N (Ilg N03--N/g soil) levels in the soil to a depth of 15 cm showed a different pattern to the soil ammonium (discussed in the previous section). At both sites, urine treatments did not significantly affect the soil (0- 1 5 cm) N03--N levels at 1 and 6 DAUA (Table 3 of Appendix 1 ). Then soil N03- -N levels began to increase due to nitrification. Though there was a trend of increasing N03--N content over the first 4 weeks, the increase was small relative to the total quantities ofN added. For example, by 27 DAUA, only 2 to 7 % of added urine N was recovered as N03--N in soil (0- 1 5 cm) at both sites (Table 3 .4). In addition, there was a lack of quantitative agreement between the decrease of NH/-N and the N03'-N production in the soil . The increases in N03--N levels in soil by 27 DAUA were only 8, 52 7, 9 and 3% of the amounts of soil NH/-N that had disappeared between 1 and 27 DAUA from the F200, F400, S200 and S400 treatments respectively. However, whether further increases in N03--N occurred shortly after 27 DAUA could not be determined as there were no sampling times between 27 and 1 00 DAUA. No statistically significant differences in N03--N levels were observed between control and urine treatments at 1 00 and 142 DAUA at both sites. Steep site 60�---------------, :; o z 50 • 40 :;; � € 30 • CD � '" • � 20 U :0: :E 10 N NS NS 25 50 75 100 125 150 DAUA -lr-SO -o-S200 -o-S400 Flat site 60�----------------� 50 z :s z • 40 :;; · . � � 30 � .. • � 20 U :0: :E 10 N NS � 0 25 NS t 50 75 100 125 1 50 DAUA """- FO -e-F200 --F400 Fig. 3 . 8 Effect of urine treatments on soil (0- 1 5 cm) N03--N. * = significant treatment differences were observed within the sampling time. S = no significant treatment differences were observed within the sampling time. These results appear to indicate that although some nitrification occurred during the experiment it was relatively small compared to the total amount of NH/-N in the urine treated plots. 3.4.6 Mineralisable N The urine N recovered as soil mineral N was less than 50% of that added by 27 DAUA in all the urine treatments. To investigate whether this unaccounted-for-N had been converted into readily mineralisable organic N, an anaerobic incubation was conducted to investigate any changes in soil mineralisable N levels in soils sampled 27 DAUA. Thereafter anaerobic incubation was also conducted on soils from samplings at 1 00 and 142 DAUA. 53 At both sites, in all treatments, higher mineralisable N levels were observed in the 0-7.5 cm soil depth than the 7.5 -1 5 cm soil depth (Table 3 .5 ). At both sites, mineralisable N in the 7.5- 1 5 cm soil depth, was not significantly affected by urine treatments. Therefore, only changes in mineralisable N in the 0-7 .5 cm soil depth are discussed (Fig. 3 .9). Table 3.5 Mineralisable N levels (kg N/ha) in control soils at different depths. Values are the average of soils sampled 27, 100 and 142 DAUA. Soil Depth Flat site Steep site 0-7 .5 cm 1 07.7 84.5 7.5- 1 5 cm 5 1 .3 45 .9 Total (0- 1 5cm) 1 59 1 30.4 A)Steep site E u 250 III ,..: 200 e- A z � 1 50 AB � I» -= A A A :a � 1 00 ftI .!! i ... I» 0 C � 50 5200 5400 50 5200 5400 50 5200 5400 27 DAUA 100 DAUA 142 DAUA B)Flat site E u 250 III ,..: 200 A A A e- A Z IV 150 A A B � B I» - B _ 1:) 100 .12 � � .!! 50 b b a b b i a ... 0 I» C � FO F200 F400 FO F200 F400 FO F200 F400 27 DAUA 100 DAUA 142 DAUA Fig. 3 .9 Effects of urine treatments on soil (0-7.5 cm) mineralisable N levels at 27, 1 00, 142 DAUA. Treatments with common lower case letters do not differ (P<0.05) within a sampling day at the same site. Treatments at the same site with common upper case letters do not differ at (P<0.05) between sampling days. 54 Urine treatments significantly increased the soil (0-7.5 cm) mineralisable N levels. At the steep site this was observed only in the S400 treatment at 27 DAUA. In contrast, at the flat site, mineralisable N levels significantly increased in both treatments at 27 and 1 00 DAUA. It is interesting to note that at the flat site, in all treatments including controls, the mineralisable N levels were similar from 27 to 1 00 DAUA but then decreased by 1 42 DAUA. This was also observed in the S400 treatment (Fig. 3 . 9). At 27 DAUA, mineralisable N had increased by more than 56 and 5 1 kg/ha, compared to controls, in the F400 and S400 treatments respectively. During the period 1 -27 DAUA, the decrease in soil (0-7 . 5 cm) NH/-N was 327 and 2 1 8 kglha in the F400 and S400 treatments respectively (data not presented). Therefore, the increases in mineralisable N levels accounted for 1 7 and 23% of the loss ofNH/-N in the F400 and S400 treatments respectively. 3.4.7 Pasture response Pasture yields were measured on two occasions, at 27 DAUA on 06 July 1 999 (Cut 1 ) and at 1 00 DAUA on 1 7 August 1 999 (Cut 2) . At the completion of the trial, 1 42 DAUA, on 29 October 1 999 there had been no measurable pasture regrowth in any of the plots after the cut at 1 00 DAUA. Over the two pasture growth periods, unne treatments significantly increased the combined (27 DAUA + 100 DAUA) pasture dry matter accumulation (Fig 3 . 1 0). The total pasture dry matter accumulation on the S200, S400 and F200 treatments increased by 2 fold, and on the F400 treatments 3 fold, compared to the control treatments. At the first cut (27 DAUA), no statistically significant increases in DM accumulation were observed at either site. This could be due to poor pasture growth because of cold winter conditions. As a result of this, a low uptake of urine N was observed, although the pasture N concentration was markedly increased by urine addition. At the first harvest, the lower dry matter production and lower pasture N uptake on the 400 kg urine 55 Nlha treatment compared to the 200 kg urine Nlha treatment, could be due to pasture burn and other indirect effects. At the second cut, unne treatments significantly increased (P<0.05) the DM accumulation at both sites. This may be a combined result of better environmental conditions for pasture growth in spring and the extended growing period. At both sites DM accumulation during the second pasture growth period was 2 times higher in the 200 kg urine N/ha treatment and 3 times higher in the 400 kg urine Nlha treatment, compared to the control. The N uptake by pasture was significantly increased by the urine application (Fig. 3 . 1 1 and Table 3 .6). Total pasture N uptake was increased 2 - 4 fold compared to controls. At 27 DAUA, pasture N uptake at the higher urine N rates at both sites was lower than at the lower urine N rate treatment. As noted previously this could be due to pasture bum. Total urine N recoveries in pasture were 26%, 24%, 22% and 1 8% in the S200, S400, F200 and F400 treatments respectively. Urine N recovered as soil mineral N at 27 DAUA had declined by 64 (S200), 1 60 (S400), 73 (F200) and 1 79 (F400) kg Nlha by 1 00 DAUA (Table 3 .4). The urine N uptake by pasture from 27- 1 00 DAUA was 36, 86, 34 and 69 kglha from S200, S400, F200 and F400 treatments respectively (Table 3 .6). These amounts of pasture N uptake accounted for 5 5%, 54%, 47% and 39% of the decrease in soil mineral N from 27- 1 00 DAUA. c 0 .. ca "S E ::::I ca � .e ca Cl Q) � .... ::::I -VI ca Co :!: C c 0 .. ca ::::I E ::::I (,) RI (,) ..1: RI -Q) Cl ... � ::::I -VI RI Co :!: C 4000 3500 3000 2500 2000 1 500 A 1 000 a 500 0 a 50 Steep site B b 5200 Treatment C c 5400 ID Cut 1 - 27 Days (09/06/1999-06/07/1 999) OCut 2 - 72 Days (06/07/1 999 - 1 7109/1 999) I 4000 3500 3000 2500 2000 1 500 A 1 000 500 a a 0 FO Flat site B b a F200 Treatment C c a F400 Ilfi:Cut 1 - 27 Days (09/06/1 999 - 0610711 999) lfi:Cut 2- 72 Days ( 06/07/1 999- 1 7/09/1 999) I 56 F ig. 3 . 10 Effect of urme treatments on pasture OM accumulation. Total DM accumulations at the same site with common upper case letters do not differ at the P 1 00 0 u Cl .. Z 80 Gj == '': ::) 60 40 20 o 10 20 30 40 50 60 70 80 90 100 1 10 120 130 140 1 50 DAUA Fig. 3 . 1 3 Total urine N recovery (%) during the experiment. __ F200 ---+- F400 -0-5200 -0-5400 - - 'Added The proportion of urine N accounted for at the beginning of the current experiment was high compared to other studies discussed earlier in Section 3 .2 .3 . In an experiment conducted at a flat land site in Palmerston North (Ball et al., 1 979), 30-35% of the added urine N could not be accounted for 1 day after the experiment began. Carran et al., ( 1 982) at a flat site in Gore, also found that about 50 and 30% of the applied urine N could not be accounted for in dry and wet plots respectively, a day after application. Both these sites in Palmerston North and Gore were well developed as well as being relatively flat. They also had a history of high production and intensive use. At the end of the experiment at 1 42 DAUA, 66% (S400), 62% (F400), 54% (S200) and 50% (F200) of added urine N was accounted for in the soil as mineral or mineralisable N, or lost by volatilisation or plant uptake (Fig. 3 . 1 3 ). It was assumed that urine N not accounted for in any of these pools during the experiment was immobil ized to complex organic matter in soil and converted to non-mobile N. 5400 800�--------------------------� 700 600 .! 500 z 400 t------�� .......... � 300 200 1-______ •• 100 O +-�--�--�-r--��--�--�� 1 6 12 30 45 60 80 103 120 142 DAUA 5200 800 �--------------------------, 700 600 .! 500 Z 400 � 300 200 . ___ __ _ 6 1 2 30 45 60 80 1 03 1 20 1 42 DAUA F400 800 �--------------------------� 700 600 .! 500 Z 400 � 300 200 100 J---------­ O +-�--,_--�_r--��--,_--r_� 800 700 600 '" 500 .r:. Z 400 � 300 200 100 0 6 12 30 45 60 80 103 120 142 DAUA F200 III • immobilised D mineralisable • Nitrate • immobilised El Mineralisable • Nitrate D Ammonium . plant o Am.Vol • immobilised o mineralisable • Nitrate DAmmonium . plant o Am.volat • immobil ised o Mineralisable • Nitrate DAmmonium • plant DAm.volat 1 6 12 30 45 60 80 103 120 142 DAUA F ig . 3 . 14 Urine N recovery during the experiment. 6 1 62 The apparent distribution of urine N in the pools described earlier is illustrated in Fig. 3 . 14 . The urine N in the mineral N pool was mainly in the form of NH4 + -N throughout the experiment. This suggested that nitrification rates were low and that leaching losses of N03--N would therefore be low. Only 1 8-27% of the added urine N was recovered in the plant. The urine N loss by ammonia volatilisation was large, ranging from 2 1 -34% of added N. At the end of the experiment ( 1 42 DAUA), 34 -50% of added urine N was assumed to be immobilized into complex organic matter. Urine N transformations in these various pools are discussed separately in the following sections. 3.5.2 Mineral N This experiment demonstrated that increased soil mineral N levels could occur during the first 2-3 months after urine application (Fig. 3 . 1 4). Large amounts of NH/-N were observed at 1 DAUA in both urine treatments at both sites (Fig. 3 .7). This observation agrees with other reports of rapid urea hydrolysis after urine application (Section 3 .2. 3 . 1 ) . As noted earlier, NH/-N was the dominant mineral N form found in urine-treated plots throughout the experiment. This is in contrast to many other studies that have found that N03--N is often the major form of N present in the urine patch 2-5 weeks after application (Ball et al., 1 979; Carran et al. , 1 982; Haynes and Williams, 1 992). Although NH4 + -N was the dominant form of mineral N throughout the trial, the large quantities of soil NH/ -N that were present at 1 DAUA in the urine-treated plots, declined rapidly with time. Between 88 and 376 kg/ha of the NH/-N present 1 DAUA had disappeared by 27 DAUA (Table 3 .4 and 3 . 8) . The NH/-N levels in urine treated plots then continued to decline after 27 DAUA to reach levels similar to the controls by 1 00 DAUA. 63 In most treatments except F400, the urine NH/-N (0- 1 5 cm) that had disappeared within the first 27 DAUA could be substantially accounted for by ammoma volatilisation, the increase in mineralisable N and plant uptake (Table 3 . 8) . Of these three pathways of loss during the first 27 DAUA, ammonia volatilisation appeared to be the greatest, ranging from 34-59% of the decrease in NH/ -N. However, the quantities of urine N lost through volatilisation were measured from the day of urine application (Day 0) to 6 DAUA and the decrease in urine N in the form of NH/-N (0- 1 5 cm) in Table 3 .8A was calculated from 1 to 27 DAUA. Therefore, not all the measured amount of ammonia volatilisation would have contributed to the decrease in urine NH/-N observed between 1 and 27 DAUA, i .e . some volatilisation would have occurred between 0- 1 DAUA. In the F400 treatment, the disappearance of urine NH/-N could not be accounted for by ammonia volatilisation, plant uptake, the increase in mineralisable N and nitrification. Approximately 160 kg N/ha remained unaccounted for. From 27- 1 00 DAUA, although urine NH/-N in soil continued to decline, plant uptake could only account for approximately half of the apparent N loss in all treatments. It is commonly observed (see Chapter 6 of this thesis) that decreases in soil 1\TH/-N are mirrored by corresponding increases in soil N03--N, through the process of nitrification. The lack of quantitative agreement between the rates of NH/-N decline (Fig. 3 .7) and N03--N accumulation (Fig. 3 . 8) during this experiment can be explained by one of two scenanos. The first possibility is that nitrification rates were indeed low during the experiment, and as noted above, significant amounts of NH4 + -N were lost through NH3 volatilisation (Section 3 .4.8), by preferential downward movement (Section 3 .4.4), by uptake of NH/-N by plants and by immobilisation ofNH/-N into the soil organic matter. The second possible scenario is that large amounts of nitrification did take place in the long interval between samplings at 27 and 1 00 DAUA, but the resulting N03--N was then leached (see later Section 3 . 5 .3) . 64 Table 3 . 8 Apparent fate of urine NH/-N from 1 -27 DAUA CA) and 27- 1 00 DAUA CB). All quantities are expressed as kg N/ha. A) 1 -27 DAUA Decrease in Apparent fate of urine NB/-N Treatment urine NB/-N Plant NB3 Increase in Increase from uptake volatilisation mineralisable in urine 1 -27 DAUA N N03'-N S200 1 36 1 7 5 1 34 9 S400 232 1 1 1 3 8 58 2 F200 88 9 4 1 44 3 F400 3 76 4 1 29 59 24 B) 27- 1 00 DAUA Decrease in Apparent fate of urine NH/-N Treatment urine NH/-N Plant NH3 Increase in Increase from uptake volatilisation mineralisable in urine 27- 1 00 N N03'- DAUA S200 53 36 - -29 0 S400 1 57 86 - - 1 0 -3 F200 70 34 - 7 -3 F400 1 54 69 - 8 -25 3.5.3 Nitrification and leaching Quantity ofN unaccounted for 25 23 -9 1 60 Quantity ofN unaccounted for 46 84 32 1 02 The results of this experiment appear to suggest that l ittle nitrification occurred fol lowing urine N application. However, as noted above, there was a possibility that further nitrification was occurring, but the N03--N so formed had then been leached from the profile before it could be detected. 65 The results of this experiment show some similarities to the work of Ball et al. ( 1 979) conducted at a lowland site at Palmerston North. Ball et al. ( 1 979) however had more frequent sampling times and observed marked nitrification followed by leaching to end up with low N03--N values in the 7 .5- 1 5 cm depth, 53 days after urine application. This final outcome at 5 3 DAUA was similar to that observed in the current experiment. It was therefore possible that the infrequent sampling carried out in the current experiment might not have detected the nitrification and leaching loss . This possibility was explored in more detail as follows. Leached N03--N could not be observed in the 7 .5 - 1 5 cm soil depth, 1 00 DAUA. So if leaching had occurred, the N03--N must have moved beyond 1 5 cm depth. A simple water balance calculated for the experimental period suggested that by 1 00 DAUA only 52 mm of drainage would have occurred. The field capacity in the 0- 1 50 mm soil depth considered for the water balance was 53 mm. Therefore, even if all the NH/-N that had disappeared was assumed to be nitrified; large amounts of 03-­ should stil l be in the 75- 1 50 mm soil depth 100 DAUA. The apparent absence of any such N03--N therefore suggests that the nitrification rates at these sites were indeed low and there had been little leaching ofN03--N. The current experiment agrees with the previous work of Sakadevan et al. (1 993), who observed no accelerated leaching of N from urine-affected soil in grazed hill pasture in New Zealand. However, although Sakadevan et al. ( 1 993) measured leaching using in situ mini lysimeters with ion exchange resin traps, they did not measure soil N03--N with time after urine application. Therefore, it is not clear whether or not the lack of leaching was due to low rates of nitrification, as was thought to be the case in the current experiment. 3.5.4 Pasture response Urine treatments markedly increased the pasture dry matter production, pasture N concentration and pasture N accumulation (see Section 3 .4.7), highlighting the value of urine N to hill country pasture production. These responses lasted only up to 100 DAUA. Haynes and Williams ( 1 993) also indicated that pasture response to added 66 urine N normally lasts for 2 to 3 months . However, Theobald and Carran (2000) reported that responses in herbage production to urine application, persisted for periods exceeding 1 2 months in an experiment conducted in flatland in Palmerston North. Pasture DM response to urine was higher at the steep site than at the flat campsite. This could be due to higher sunlight on the north facing steep site than the flat campsite, which was in a valley bottom. The urine N recoveries by pasture, ranging from 1 8-27%, are within the range cited in literature (Section 3 .2 .3 .7). Although the low nitrification that appeared to occur in this soil may have decreased N losses through preventing leaching, it could also reduce N uptake by pasture. This is because plants prefer to take up a mixture of N forms or N03--N (Section 3 .2 .3 .7). In addition, microbes prefer NH/ -N for N assimilation (Section 3 .2 .3 .6). This would lead to competition between plants and microbes for N. 3.5.5 Ammonia volatilisation Estimation of NH3 volati lisation after urine addition in this experiment indicated this to be a substantial loss of N from the system. In this experiment, 2 1 - 34% of added urine N was lost as ammonia. Many other workers have also recorded considerable N losses by NH3 volatilisation in urine patches (Ball et al., 1 979; Carran et al., 1982; Vallis et al., 1 982). The technique used to estimate NH3 volatilisation in this experiment (Carran et al., 2000) greatly overcomes the technical difficulties associated with hill country volatil isation measurements and the technique is simple to use. The present experiment clearly showed the potential for high NH3 volati lisation in hill country pasture, even during the wet, cooler winter period. 3.5.6 Mineralisation and immobilization At both sites, the quantities of NH/-N at 1 and 6 DAUA remained high (Fig. 3 .7), although it appeared that a substantial amount of NH3 volatilisation occurred during this period. This resulted in an apparent recovery of urine N of greater than 1 00% as 67 described earlier. This suggests that additional NH/-N may have been produced by mineralisation of organic matter in some sort of priming effect. Ball and Til lman ( 1 994) have also commented on recoveries of urine N in excess of 1 00%. The exact cause of the priming effect is not clear. High salt concentrations in urine could dissolve some organic matter resulting in release of N, or cause the death of microbes (Section 3 .2 .3 .6). The F400 treatment showed a larger priming effect than did the S400 treatment. This could be due to differences in the mineralisation potential of the two soils as shown in Section 3 .4.6; flat site soils had slightly higher mineralisable N ( 1 59 )lg N/g soil in 0- 1 5 cm) than soil from the steep site ( 1 30 )lg N/g soil in 0- 1 5 cm). IfN mineralisation is a common process when urine is added to the soil, the N loss from urine patches will include losses of N from the original organic matter as well as urine N. As noted in Section 3 .4. 1 , it was assumed that urine N not accounted for by soi l mineral N or mineralisable N, or urine N lost from the soil through plant uptake, and volatilisation during the experiment was immobilized into complex organic matter in the soil and converted to non-mobile N. The increase in mineralisable N in urine treatments was a hint that immobilization of urine N was occurring. However, the exact source of the increased mineralisable N is not clear, as some of the N released by the priming effect could also be reimmobilized. The estimated amount of immobilization at the end of the experiment was large. Values were 46% (S200), 50% (F200), 34% (S400) and 38% (F400) of added urine N. Ball et al. ( 1 982) reported that many top soils in the summer wet hill country of the lower North Island contain very large quantities of N, rendered largely unavailable by the relatively wide C:N ratio of the soil organic matter present. Sakadevan et al. ( 1 993) also indicated, from an experiment conducted at Ballantrae AgResearch hill country research station, that the majority of added urine N was immobilized in hill country soil. 3.5.7 Comparison of urine N transformations at two sites 68 The review of l iterature (Chapter 2) indicated that in hil l country the N cycle is complex due to the existence of contrasting landscapes. This experiment was carried out on two contrasting landscapes, a flat site that showed some evidence of sheep camping and a steep slope. Surprisingly, the general p atterns of urine N transformations over time in the two sites were quite similar (Fig. 3 . 1 4). However, when compared using a number of soil fertility indices the two sites in fact, did not appear greatly different (Table 3 .9). Table 3 . 9 Soil fertility indices of the two s ites . Soil fertility index Flat site Steep site Total N 0-7 . 5 cm (%) 0 . 3 4 0 . 3 7 Total C 0-7 . 5 c m (%) 4 . 7 4 . 9 CIN (0- 7 . 5 cm) 1 4 1 4 Olsen P (0-7 . 5 cm) ()lg/g soil) 44 83 pH 5 . 3 5 . 3 This may explain the similar behaviour between the sites that topographically were very different. 3.6 Conclusions Urine application markedly increased the soil mineral N availability. A priming effect after urine application was indicated in both experimental sites . Increased mineral N could be found about 2-3 months after urine application. NH/-N was the dominant mineral N form throughout the experiment indicating that a low level of nitrification existed in the experimental soil s . Therefore, the potential for leaching of added urine N was low in the experimental sites. 69 Pasture responses to the added urine were observed. Pasture dry matter production was increased 2-3 fold in the urine treated plots, compared to controls, during the experimental period. However, the estimated urine N recovery by pasture was only 1 8% to 27%. Ammonia volatilisation after urine application was a major N loss mechanism, ranging from 2 1 % to 34% of added urine N. Immobilisation of urine N seems to be high in hill country pastures. During the first 27 DAUA 1 6% to 50% (mean 29%) of the loss of urine NH/-N, could be accounted for by increased mineralisable N and at the end of the experiment, estimated immobilisation was 34% to 50% of added urine N . CHAPTER 4 DEVELOPMENT OF ION EXCHANGE RESIN MEMBRANE SPIKES FOR CONTINUOUS MONITORING OF AVAILABLE SOIL NITROGEN IN HILL COUNTRY PASTURE 4.1 Introduction 70 The results of Chapter 3 revealed that more than 40-50% of added urine N could not be recovered as mineral N by 27 DAUA. Probable mechanisms of loss of urine N were assumed to be immobilization into microbial biomass and other complex organic carbon materials, and ammonia volatilisation. However, the widely spaced sampling times during the experiment resulted in a number of questions such as whether nitrification occurred or not, the detai led pattern of mineral N change with time, and exactly at what stage did most of the urine N start to disappear. Answers to these questions are extremely important in determining the fate of the unaccounted-for urine N. The 2 M KCl -extractable N method, which was used in the experiment detailed in Chapter 3 , has several disadvantages. First, the method is laborious. Five cores were taken from each plot, leading to 1 20 cores per sampling in what was a relatively small experiment. In hill country, taking this number of cores involves a lot of labour. In addition, soil samples needs to be extracted with 2 M KCl as soon as possible because mineral N does change due to microbial reactions. Hence, frequent sampling is not practicable using this technique. In addition, the data obtained are only a measure of available mineral N at the time of sampling. The availability of N in the soil is regulated by numerous interacting processes including microbial mineralization and immobilization, diffusion, mass flow and plant uptake. Thus, the available N pool is highly variable over time and space. Over 90% of soil N is held in soil organic matter and its release by mineralisation and nitrification is dependent on a number of soil environmental factors. So the rate of 7 1 release of N is difficult to predict. In addition, nitrate which is produced from mineralisation is susceptible to leaching, denitrification and immobilization. Consequently, soil tests for N availability have proved difficult to develop, and no single test has been universally adopted (McLaren and Cameron, 1 996). Methods involving the incubation of ion exchange resins in soils and their subsequent removal and extraction for nutrient ions have shown particular promise for providing integrative indices of available soil N. Anion exchange resins contain positively charged surface functional groups, that attract anions including phosphate, sulphate and nitrate by electrostatic attraction. Cation exchange resins contain negatively charged surface groups, that attract cations such as potassium and ammonium to their surfaces. Thus , ion exchange resins act as a sink for ions when placed in a suspension of soil and water. The obj ective of this chapter is to develop a new method for N measurement using ion exchange resin membranes which may be suitable for use in hil l country situations to enable continuous N measurements over time. 4.2 Literature review 4.2.1 Ion exchange resins In recent years, interest in the use of ion exchange resins to study natural soil systems, has increased. The theoretical descriptions of ion exchange resins, discussed in this section, are based on the information in; BDH ( 1 98 1 ), Harland ( 1 994) and Skogley and Dobermann ( 1 996). Many specialty resin products have been developed. Macroporous ion exchange resins are the most common resins used in soil studies. They are generally uniform in external shape (spherical), with individual particles referred to as beads. Ion exchange resins in the form of sheets are called ion exchange resin membranes. During manufacture, membranes are extruded into sheets and combined with reinforcing material to provide dimensional stability and mechanical strength. 72 Most ion exchange resins are solid organic polymers with an electrostatic charge that is neutralized by a selected counterion of opposite charge. Hence, they function in a manner analogous to charged soil colloids . Most ion exchange resins are made from styrene polymerized with itself to form long chains. These chains are reacted with divinylbenzene to produce cross-linkages. Ion exchange membranes are made "cation permeable" or "anion permeable" by chemical treatment, whereby sulphonic acid (­ S03-) or quaternary ammonium (-NR3 +) groups respectively are attached to the membranes (Fig. 4. 1 ). Cation exchange resms are classified as either strong acid or weak acid and amon exchange resins are either strong base or weak base. Strong acid cation exchange resins have active groups having a greater affinity for cations other than H+ and weakly acidic cation exchange resins have a strong affinity to H+. Resin studies involving the release of cations from soils and subsequent accumulation by resins, as a measure of cation mobility or bioavailability, should involve use of a strongly acidic type of cation exchanger. Strong or weak base anion exchange resins differ similarly by relative affinity of the active group for OH-, compared with other anions. Resins exhibit preferential selectivity for various ions. Knowing the relative affinity of the resin for each ion in the medium is critical to understanding the results of a particular study. When ion-exchange resins are equilibrated with a solution containing a mixture of ions, the proportions adsorbed by the resin will not be the same as the ionic proportions in the bulk solution. The following sequence (Harland, 1 994) represents the order of adsorption usually found for dilute solutions of commonly encountered ions with standard resins. Strong Acid Cation Resin (styrenic-sulfonate) Ag+ > Cs+ > K+ > NH/ > Na+ > H+ > Li+ ; Ba2+ > Pb2+ > Sr2+ > Ca2+ > Ni2+ > Cd2+ > Cu2+ > C02+ > Zn2+ > Mg2+ > Cs + Strong Base Anion Resin (styrenic-quaternary ammonium) SO/- > HS04- > r > N03- > Br- > cr > HC03- > HSi03- > F > OR 73 Heal catalysl s1yrene linear polystyrene divi y lben.zene crosslinked copolymer styrenic su l fonic acid resin Fig. 4. 1 Polymerization synthesis of a styrene sulfonic acid cat ion exchange resm (Harland, 1 994). Characteristics of commercially available BDH Ion exchange resm membranes are described as : • High perrnselectivity: This describes their ability to allow, for example, a cation to pass through much more easily than an anion. • Low electrical resistance 74 • High mechanical strength : The polymeric fibre cloth used in these membranes give them adequate strength for any normal application. • Resistance to chemical attack : These membranes are stable in alkalis and acids and in inorganic salt solutions. • High thermal resistance: Generally suitable for temperatures up to 60 cC, but for continuous use a maximum working temperature of 40 cC is recommended. 4.2.2 Diffusion Ion exchange involves the redistribution of ions from the solution to the resin. This redistribution occurs through diffusion. Diffusion means ' spreading out ' , and is caused by random thermal motion, as is the Brownian movement of colloi dal particles observable under the microscope (Wild, 1 98 1 ) . In unstirred liquids, all the molecules and ions of the solvent and solute have this random movement. As a result of this spontaneous molecular motion, diffusion intermingles the ions and molecules in gases, liquids and solids without the participation of external forc es. The motion of individual ions or molecules is irregular, and at equal concentrations, no net movement occurs. However, if concentrations are different, substances move from regions of higher to those of lower concentration. The driving force is the existence of a concentration gradient. Diffusion is the processes by which nutrients m igrate to the depletion zone near roots to maintain soil solution equilibrium levels (Wild, 1 98 1 ; Jungk and Claassen, 1 997). Tinker and Nye (2000) extensively reviewed the subj ect of solute flux through soil. The following formula was used to calculate the diffusion flux. where F = flux density of solute (kg/m2/s) D, = the diffusion coefficient of the solute in free solution (m2 Is) (4. 1 ) e = the fraction of the soil volume occupied by solution; and gives the cross section for diffusion through solution (m3/m3 ) f dCldx 75 = an impedance factor (dimensionless) = the concentration gradient of solute in the soil solution (kg/m 4) = the excess flux created by the possibility of surface diffusion ( kg/m2/s) The impedance factor needs to be measured indirectly. It varies with the moisture content of the soil because, as soil becomes drier the diffusive pathway b ecomes more tortuous. The relationship between the impedance factor and soil moisture for chloride ion in a sandy loam is shown in Tinker and Nye (2000) . The same data were used to draw the logistic curve fit (Fig. 4.2) for the relationship between impedance factor and moisture. It can be seen that under very dry conditions the impedance factor (1) is very low. 0.7 0.6 ... 0 0.5 .. u � ... 0.4 CII u c � " 0.3 CII Q. 0.2 E 0.1 y= 0.66/(1 +77.24 EXP(-14X) 0 0 0.2 0.4 0.6 0.8 Volumetric moisture content Fig. 4.2 Relationship between diffusion impedance factor and moisture content. (Logistic curve fit for the data in Fig. 4 . 1 of Tinker and Nye (2000) .) 4.2.3 Ion exchange resin use in recent soil research Recently, ion exchange resins have been used in a number of ways in soil analysis. S ibbesen ( 1 978) used resin bead-water-soil suspensions to extract soil phosphate. The suspensions were shaken and transferred to a sieve (mesh size 0 . 3 5 5 mm), which retained the resin and allowed the soil suspension to pass through. Then the resin was washed to remove adhering soil and resin-adsorbed P was extracted with 1 M HCl. 76 However, Sibbesen ( 1 977) had earlier suggested a better method to separate soil and resin after shaking. In that method, the resin was placed in a nylon netting bag which was immersed in a soil suspension and shaken. Compared to the method (S ibbesen ( 1 978) where the resin beads are freely suspended in the soil-water mixture, the bag procedure allowed quicker separation of resin from the soil suspension. Subsequently, researchers used these resin- fil led bags directly in the fi eld to assess N availability in soil (Binkley and Matson, 1 983 ; B inkley, 1 984; Carlyle and Malcom, 1 986), to measure N mineral isation and nitrification (Distefano and Gholz, 1 986; Hubner e t al. , 1 99 1 ), nitrate movement (Schnabel et a l . , 1 993) and as a phytoavailability soil test (Skogley et al. , 1 990). A maj or disadvantage of the buried resin bag technique is the shape of the resin bags. The three dimensional nature of the resin bags means that the results are likely to be highly dependent upon the placement methods because of differences in the flow of water and nutrients through or around the bags and surrounding soil. S aggar et al. ( 1 990) suggested that when used in pasture soil, the resin bags trap fine root material and soil particles. If not removed by washing, these entrapped materials interfere with further analysis. Furthermore, with normal wear and tear the sealed edges of the bags may rupture resulting in the loss of resin beads to the soil suspension during shaking. Sakadeven et al. ( 1 994) used mixed resins in an in situ mini lysimeter for measuring nutrient losses by leaching from grazed pastures. Saggar et al. ( 1 990) introduced the use of an ion exchange resin membrane technique for extracting phosphorus from soil. Strips of anion and cation exchange membrane were shaken with suspensions of soil in deionised water for 1 6- 1 7 hours. Then phosphate retained on the anion exchange resin strip was determined by shaking the strip directly with a colorimetric reagent (Murphy and Riley, 1 962). Subsequently, some studies did simultaneous extraction from soil of N, S, Ca, Mg, K, Mn, Al and P by shaking soil with a resin membrane in distilled water (Qian et al. , 1 992; Mclaughlin et al. , 1 993). The resin was then separated from the soil and ions desorbed from the resin using an acid/salt solution. 7 7 Recently some work has been done o n the use o f ion exchange membranes for in situ measurement of soil nutrients (Cain et al., 1 999). Subler et al. ( 1 995) and Qian and Schoenau ( 1 995) directly inserted the membrane strip into soil that was incubated in containers. Cain et al. ( 1 999) constructed a tool to place ion exchange resin membranes directly in the fi eld. Cain e t a l . ( 1 999) studied spatial and temporal variation in soil N availability in coastal dunes. Many of the disadvantages of the buried resin bags technique may be overcome by the in situ use of ion exchange resin impregnated membranes. Their essentially two dimensional structure ensures more surface area will be in contact with the soil. Compared to the resin bags, resin membranes c an be inserted into the soil with minimal disturbance. This technique could also offer additional advantages compared to their use in laboratory water-soil suspensions. Direct burial of resin strip s in soil will more closely resemble a plant root in its natural environment. Subler et al. ( 1 995) emphasized that in situ use of ion exchange membranes needed more work before reliable interpretations can be made under a wide variety of field and laboratory conditions. In New Zealand, use of ion exchange resin membranes in soil analysis has received much attention after the method was introduced by S aggar et al. ( 1 990). However, there has been no attempt to use ion exchange membranes for in situ measurements of nutrients. 4.3 Development of the ion exchange resin membrane spike For direct soil burial of resin membranes to be effective, a technique that allows rapid placement, removal and handling, needs to be developed. A series of studies were carri ed out to identify the most effective technique for direct burial of resin membranes. 4.3.1 Experiment 1 : Preliminary assessment of the ability of resin strips to adsorb mineral N from Soil At the first attempt, both anion and cation resin membranes were cut into 1 x 5 cm pieces and glued, using methacrylatebase epoxy glue (Cain et al., 1 999), to two sides of a rigid acrylic spike. These spikes were placed in saturated KCI solution for 24 hours, rinsed with deionised water and air-dried. The spikes were then buried into the soi l of 78 two pasture plots, which were known from earlier studies to have contrasting levels of soil fertility and pasture growth. After 3 days, the resin spikes were removed from the soil , washed free of adhering soil with deionized water and then placed in centrifuge tubes containing 25 mL 2 M KCl and shaken for 1 hour. All nutrient ions adsorbed from the soil by the resin were then displaced into KCl eluent and the NH/-N and N03- -N concentrations determined by Technicon Auto Analyzer (Searle, 1 97 5 ; Blakemore et al. , 1 987). Amounts of NH/-N and N03--N were expressed as �g-N/5 cm2resin 13 days. Table 4 . 1 . Resin-adsorbed N (�g-N/5 cm213 days) from two different pasture plots. Plot 1 : Poor Growth Plot 2 : Good Growth Spike NH4+-N N03--N NH4+-N N03--N 1 2. 1 6.9 6.9 1 6 .2 2 5 . 4 4 . 5 8 . 1 6.9 3 7 . 5 6 . 0 9 . 6 1 2 .6 4 5 . 1 2.7 1 3 . 5 1 0 .2 5 1 3 .2 7.2 8 . 7 1 5 .3 Average 6.7 5.5 9.4 12.2 CV% 62 34 27 3 1 The results (Table 4. 1 ) show that the resin membranes successfully adsorb both NH/-N and N03--N when directly buried in the soil. This was the main conclusion to be drawn from this small experiment. An additional observation was that on average, resin adsorbed N values were higher in the fertile plot with good pasture growth than in the plots with poor pasture growth although there was considerable variation in N adsorption among the resin spikes. These results were sufficiently promising to move on to the next stage. 4.3.2 Experiment 2: Assessment of variability with resin spikes Next, another simple experiment was carried out to find out whether the variation i n resin-adsorbed N between spikes, observed i n the previous experiment, was due t o variability i n the soil or variability between the spikes. Five resin spikes were placed i n a large volume o f a solution of NH4N03 containing 1 0 � g NH/-N/mL and 1 0 � g N03-­ N/mL for a day, and the resin-adsorbed N analyzed (Table 4.2). 79 The results suggested that there was a considerable variation in N adsorption among the spikes (Table 4.2) even though they had been immersed in a uniform solution. Thus these resin spikes were thought to be not suitable for use in the field. Table 4.2 Resin-adsorbed N ()lg-N/5 cm2/day) from Nf4N03 solution containing 1 0 )lglmL NH/-N and 1 0 )lglmL N03--N. Spike NH/-N 1 5 6 . 8 2 7 1 . 3 3 5 2 . 0 4 1 29 . 3 5 1 3 5 . 3 Average 88.9 CV% 45 4.3.3 Experiment 3: Optimization of resin spike construction N03--N 203 . 8 1 1 9 . 5 326.0 1 62 . 8 323 .3 227.1 41 The resm membranes used in the prevlOUS experiment were double sided-that is exchange resin was on both sides of the supporting fabric. It was suspected therefore that by gluing the membranes to the acrylic spikes, variable proportions of the cation lanion exchange sites were inactivated or rendered inaccessible by the glue. To test this, glued spikes and unglued pieces of resin membrane cut to the same size were immersed in separate 25 mL samples of Nf4N03 solution containing 1 0 Ilg NH/-N/mL and 1 0 )lg N03--NI mL for a day. Table 4 . 3 Adsorption of N (llg-N/5 cm2/day) by glued spikes and fresh resin membranes from 25 mL samples of Nf4N03 solution containing 1 0 )lg Nf4 +-N/mL and 1 0 )lg N03- -N/mL over a day. Spike Resin membranes (5 cmL) glued Fresh resin membranes (5 cmL) to spikes NH4+-N N03--N NH/-N N03--N 1 34. 8 86.3 1 22 . 8 2 1 1 .3 2 1 24.0 1 82 . 0 1 3 5 . 3 2 2 8 . 3 3 7 5 . 5 1 5 8 . 0 1 1 6 . 3 1 74 . 3 4 8 1 . 8 1 08 . 8 1 2 1 . 8 222 . 5 5 3 8 . 3 9 5 . 8 1 1 6 . 8 220.5 Average 70.9 1 26.2 1 23 . 5 2 1 1 .4 CV% 52 33 6 1 0 80 The results (Table 4.3) clearly suggest that the unglued portions of resin membranes adsorbed more NH/-N and N03--N and had greater uniformity than the resin membranes glued to spikes, therefore supporting the hypothesis that the variabi lity in N adsorption of resin membrane spikes was due to the use of glue. This finding led to an alternative design of the resin membrane spikes. 4.3.4 Experiment 4: Optimization of resin spike construction 11 The next experiment involved using a water-resistant adhesive tape instead of glue to attach the resin membranes to the acrylic spikes. Strips of cation and anion membranes ( I x 6 cm) were taped to two sides of an acrylic spike (2 x 0 . 5 x 8 cm) using water resistant tape, leaving 5 cm2 of membrane area for adsorption. Initially 1 6-resin membrane spikes were made and the variation among the spikes was again examined in NH4N03 solution. Table 4.4 N adsorption (llg-N/5 cm2/day) to resin spikes immersed in NH4N03 solution containing 1 0 Ilg/mL NH/-N and 1 0 IlglmL N03--N for a day. Spike Adsorbed N03--N Adsorbed NH4 +-N 1 1 49.3 44 . 0 2 1 67 . 5 5 5 .3 3 1 76.3 5 8 . 5 4 1 52.0 5 8 . 3 5 1 86.0 4 9 . 0 6 1 54.5 5 7 . 3 7 1 82 . 5 64. 5 8 1 89 . 5 5 0 . 0 9 1 74.0 64. 3 1 0 1 8 1 . 8 5 6 . 5 1 1 1 8 1 . 5 5 3 . 3 1 2 1 67.3 5 8 . 5 1 3 1 80 . 0 5 1 .3 1 4 1 82 . 3 5 0 . 5 1 5 1 65 .3 5 4 . 8 1 6 1 72 . 5 6 2 . 3 Average 1 72.6 55.5 CV% 7 1 0 8 1 The results (Table 4.4) indicated that there was relatively little variation in the amount of N adsorbed by resin membrane spikes when the membranes were attached to the spike using tapes. It was interesting to note that from a solution containing equal amount of NH/ -N and N03--N, adsorption of N03--N to resin spikes was greater and less variable than NH4+-N adsorption. 4.3.5 Experiment 5 : Evaluation of resin spike variability in soil These resin spikes were also examined in a homogeneous soil to check suitability for field use. Soil (Manawatu Sandy Loam) was sieved « 2 mm) uniformly moistened (23 .67%w/w) and placed in a container at a bulk density approximately similar to that in the field. Then eight KCl-saturated resin spikes were buried in the container of compacted soil. After seven days in the soil, spikes were removed and the resin­ adsorbed NH/-N and N03--N were analyzed (Table 4.5). In addition in this experiment, after the standard single extraction with 2 M KCl , resin membranes were extracted a second time with another 25 mL of fresh 2 M KCl to check further recovery ofN from the resin. In the second extract, only negligible amounts of N were detected. This suggests that one extraction is sufficient to recover a high proportion of adsorbed N from the resin membrane. Table 4.5 N adsorption to resin spike from homogeneous soil (J.lg-N/5 cm2n days). Spike Adsorbed N03--N Adsorbed NH4 +-N 1 66.5 1 5 .0 2 32 .0 1 8 .8 3 55 . 3 1 1 .3 4 39.3 9.5 5 66.8 1 0.0 6 5 1 . 8 7 .0 7 63 .8 3 .8 8 5 1 .8 6.0 Average 53.4 1 0.2 CV% 24 48 The results in Table 4.5 indicated that the resin membranes attached to spikes with tape still showed some variation in the amounts of adsorbed N when inserted into relatively homogeneous soil. Adsorption of N03--N showed lower variation than NH/-N. From 8 2 this experiment, i t was not clear whether the variation in N adsorption was due to soil variation or due to differences in the N adsorption capacity of each resin spike. Even though the soil used in the laboratory was relativel y homogeneous there will inevitab ly be some variability remaining and this may account for some of the variation between resin spikes. 4.4 N adsorption to resin membranes The N adsorption behaviour of the 5 cm2 ion exchange membrane spikes was examined using different concentrations of N�N03 solutions. The KCI saturated resin spikes were kept for 24 hrs in 25 mL of NH4N03 solutions containing 20, 40, 60, 80, 1 00, 1 5 0, 200, 250 and 3 00 )lglmL o f N�+-N and N03--N. Adsorbed N was then extracted by shaking the resin spikes for 1 hour with 25 mL of 2 M KCI . The concentrations of resin-adsorbed NH/ -N and N03--N were analysed using a Technicon Auto analyser (Searle, 1 97 5 ; Blakemore et al., 1 987). Both �+-N and N03-- adsorption to resin increased linearly with increasing initial solution N concentration up to 200 and 1 50 )lg N/mL respectively. At these initial solution concentrations, the resin membranes (5 cm2) adsorbed 1 63 5 )lg of 03--N and 1 1 1 1 )lg of N� +-N. Thereafter, N adsorption to resin was relatively constant indicating that these values were the maximum adsorption capacities of the resin spikes under these conditions. As previously noted, N� + -N adsorption to resin membrane was less efficient than N03--N adsorption. The actual maXImum N adsorption by resm spikes was determined by repeatedly placing the resin spike in fresh NH4N03 solutions containing 300 )lglmL of NH/-N and N03--N until no further adsorption took place. 8 3 1 800r-----------------------------------------------------------� 1 600 _ 1400 >. co 'tl NE 1200 u on C, 2. 1000 Z 'tl .e 800 o III 'tl 'l' 600 c ·in G> a: 400 200 o 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 Initial NH:,N and N03·,N concentrations ( 119 N in 25 ml of NH.N03) 1 ___ Nitrate,N -0-Ammonium, N 1 Fig. 4 . 3 . N adsorption to resin spikes from different concentrations of NH4N03 solution. Each point represents the average of three replicates. Under these conditions, the maximum quantity of NH/- N that was adsorbed to the resin spikes (5 cm2 of cation membrane) was 2393 ).!g ( 1 7 1 J.lmoles) NH/-N and the maximum quantity N03'-N adsorption to the resin spike was 2 1 79 ).!g ( 1 5 6 J.lmoles) N03'-N. In the previous experiments, NH/-N adsorption to resin was much lower than the N03'­ N adsorption to resin. Thus, it was surprising that in the results described in the previous paragraph, the maximum adsorption capacity of the 5 cm2 cation membrane was higher than the 5 cm2 anion membrane. S imple arithmetic calculations were used to model the amount of N adsorption to KC l­ saturated resins from 25 mL samples of N�N03 solutions of different concentrations. Maximum capacity of 5 cm2anion membrane Maximum capacity of 5 cm2cation membrane = 1 5 6 J.lmoles = 1 7 1 J.lmoles 84 It was assumed that er and N03 - do not show any differential selectivity in adsorbing to resin. Thus; er (resin) NO; (resin) = ----'=-- -- er (solution) NO; (solution) (4 . 1A) Example calculation of the amount of N03--N adsorption to resin from the 25 mL sample ofNH4N03 solution containing 20 JlglmL ofNH/-N and N03--N. Initial amount of er on resin surface Amount ofN03--N in 25 mL sample of 20 JlglmL NH4N03 solution Total amount of anions in the system Fraction ofN03- on the resin surface Fraction of er on the resin surface a ( 1 56) + a (36) = 1 56 a= 0 .8 1 = 1 56 Jlmoles = 36 Jlmoles = 1 92 Jlmoles = a = a Thus, 8 1 % of each anion will be adsorbed on the resin surface and 1 9 % will be in the solution. Similar calculations were done usmg the other concentrations (Table 4.6) and the modelled values for resin-adsorbed N were plotted against the initial concentration of N03--N in 25 mL of N�N03 solution in Fig. 4.3 . It can be seen that the measured resin-adsorbed N03--N and the estimated values were very close (Fig. 4 .4), suggesting that the measured maximum sorption capacity of 5 cm2 resin membrane was reasonable and the assumption of the resin 's non-selectivity between N03- and er was valid. Similarly, the quantity of resin-adsorbed NH/-N was estimated assuming the resin showed no selectivity between K+ and NH4 + (Fig. 4.5). It can be seen that the measured resin-adsorbed values for N�+-N were much lower than the estimated values. This suggests that resin is showing higher selectivity for K+ than NH4 + . Thus , NH: (resin) K + (resin) -----'----- = n ---'------'-NH: (solution) K + (solution) 85 (4. 1 B) where n i s the selectivity coefficient. Table 4.6 Estimation of resin-adsorbed N03--N from �N03 solution containing different initial quantities ofN03--N. �moles of NO]--N in solution (x) 3 6 7 1 1 07 1 43 1 79 268 357 446 5 3 6 1800 >; 1600 IV N'!!. E 1400 U III c;, 1200 -= � 1000 cS Z 800 " Cl .c 600 .. 0 1/1 " � 400 c: Vi Cl 200 a:: 0 Total anions in Proportion of anions adsorbed on Resin-adsorbed NO]·-N system reSIn (�g) (x+ 1 56) a =( 1 56/ (x+ 1 5 6)) ( 1 4) (x) (a) 1 92 0 . 8 1 407 227 0.68 686 263 0.59 889 299 0_52 1 044 3 3 5 0.47 1 1 66 424 0.37 1 3 80 5 1 3 0.30 1 5 20 602 0.26 1 6 1 8 692 0.23 1 69 1 • o 500 1 000 1 500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 Initial NO,--N in 25 ml of NH4NO, solution ; 1 800 � � N E 1600 U !!! 1400 en � � 1200 + ... % 1000 Z 'a GI 800 .a .. 0 600 UI 'a � 400 C UI 200 GI a:: 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 I 0 Measured -Modelled (n=1 ) -Modelled (n=2.5) I Fig. 4.5 Estimated and measured resin-adsorbed NH/-N from NI£,N03 solutions containing different initial concentrations ofNH/-N. n = Selectivity coefficient ofK+ 4.4.1 N adsorption to resin spikes over time 4.4. 1 . 1 Experiment 1 : N adsorption from solution The rate of N adsorption by resin membrane spikes in soil depends on the kinetics of the adsorption process itself and also the rate of diffusion of N03 --N or NH4 + -N through the soil to the resin membrane. In this simple preliminary experiment, the rate of adsorption ofN03--N and NH/-N from a solution ofNH4N03 was investigated. The KCl-saturated resin membrane spikes were put into centrifuge tubes containing 25 mL of solution ofNH4N03 containing 1 0 f.lg NH/-N/mL and 1 0 f.lg N03--N/mL . At 5 , 1 0, 20 and 3 0 minutes and 1 , 2, 6 and 1 2 hours after immersion in the NH4N03 solution, 3 replicate spikes were removed and resin adsorbed NI£,+-N and N03--N contents were analysed (Fig. 4.6). 88 z - 175 C " .- 1 50 en Q) Q) .Q .. 1 25 .. N 0 E 1 00 en " u 75 nI U) - 50 C Z .-en en 25 Q) � a:: - 0 0 2 4 6 8 1 0 1 2 14 1 6 18 20 22 24 26 Adsorption time (Hours) I � Nitrate-N -Ammonium-N I Fig. 4 . 6 N adsorption by resin spikes with time from NB4N03 solution containing 1 0 )lg NH/-N/mL and 1 0 Ilg N03--N/mL . N adsorption by the resin membrane spikes was non linear with time. Initially rapid uptake was fol lowed by a relatively s low rate of adsorption as the maximum adsorption was approached. It took 2-6 hours to achieve maximum adsorption. Up to 2 hours both NB/-N and N03--N were adsorbed at the same rate. From then on NB/-N adsorption was much slower while N03--N adsorption continued up to 6 hours. As seen previously, approximately twice as much N03--N was adsorbed to resin membranes than N� +-N at their maximum adsorption under these conditions. 4.4.1 .2 Experiment 2: N adsorption from soil To study the rate of N adsorption from soil the following experiment was conducted using air-dried, 2 mm sieved, homogeneously moistened Manawatu sandy loam. At the beginning of the experiment the moisture content (26.6% w/w) and 2 M KCI - extractable NB/-N ( 1 5 Ilg/g soil) and N03--N ( 1 3 1 )lg/g soil) contents were analysed. Twenty-one plastic cups ( 1 43 cm3) were filled with moist soil equivalent to 1 90 g air­ dry soi l . These cups were incubated at room temperature in the dark with the cups covered by a polythene sheet to minimise evaporation. From time to time, the polythene sheets were opened to maintain aerobic conditions. The soil 2 M KCI - 89 extractable NH/-N and N03--N contents were analysed in three cups at day 3 and in a further 3 cups at day 7. A major difficulty in determining the rate of adsorption of soil nutrients to resins is ensuring that the nutrient level in the soil is constant with time. Any change in soil nutrient level over time will confound the apparent rate of sorption. In this experiment it was expected that any initial flush of mineralisation would have ceased after 7 days and the levels of NH/-N and N03--N in the soil would then be reasonably constant. The samples on day 3 and 7 were used to check that this was indeed the case. At day 7, KCI-saturated resin spikes were inserted into the rest of the cups. On each of days 8, 9, 1 1 , 1 5 and 22 ( 1 , 2, 4, 7,and 1 4 days after resin spikes burial) three cups were taken and the resin adsorbed NH/-N and N03--N contents and 2 M KCl -extractable soil NH/-N and N03--N contents were analysed. "0 - UI N tn E _ U tn Il) :::L -,.... Cl Z ..; + , z .. , :J: + .. Z :J: Qj Z :g " nI Qj .. .c u .. nI 0 .. I/) .. " )( nI .! C: u I/) � Qj =: a:: N 1 6 1 4 1 2 1 0 8 6 4 2 0 0 2 4 6 8 10 12 1 4 1 6 1 8 20 22 24 Days -2M KCI extractable ammonium --- Resin-adsorbed ammonium Fig. 4 .7 NH/-N levels during the incubation as measured by the 2 M KCl -extractable and Resin methods. The 2 M KCl -extractable soil NH/-N levels decreased over the first 8 days and thereafter stayed at steady levels . This could be due to nitrification, during the initial 90 period of the incubation. This decrease was also indicated by the resin spikes from 9 to 1 1 days. Once the soil NH/-N levels achieved the steady levels, resin adsorbed NH/­ N started to increase and then stayed constant to the end of the experiment. It seemed that the resin spikes had achieved equilibrium within 2 days of insertion in the soil. Relatively uniform 2 M KCI -extractable N03--N levels were observed throughout the incubation. The N03-·N adsorption to resin spikes was non l inear with time (Fig. 4.8) . Initially N03--N adsorption to resin was rapid. Thereafter, gradual changes in membrane bound N03-·N were observed. 2000.--------------------------------------------. � � 1 800 III E � u 1600 = I/') ::L - -- � 1400 � - o ::z;: 1200 Z 0 GI Z 1 000 :g -g '0 of 800 III 0 � � 600 1 � U III 400 :r:: GI :E a:: 200 L--�----� ......... -�-----<---------­N o 2 4 6 8 10 1 2 14 16 1 8 20 22 24 Days -2M KCI·extractable nitrate --- Resin-adsorbed nitrate Fig. 4.8 N03--N levels during the incubation as measured by 2 M KCI -extractable and Resin methods. The key point of these results was that the resin spikes took a number of days before equilibrium of N03-·N was observed. Seven days after resin spike burial, resin­ adsorbed N03--N was constant indicating that a 1 week burial period in the soil may be appropriate for field measurements. 4.5 Resin spikes performance in field 9 1 Two simple field experiments were conducted to check the resin spikes perfonnance in field conditions. 4.5. 1 Field experiment 1 The resin spikes were used to assess the mineral N status at sites, where ongoing N trials were being conducted by AgResearch, Palmerston North. The same experiment was repeated at two different sites (Site 1 and Site 2). The soil types at the two sites were Manawatu sandy loam and Kairanga silt loam (both Fluvial Recent soils) for site 1 and site 2 respectively. Each site contained control (C) plots that did not receive any N and treated plots that had received more than 400 kg Nlha/yr of fertil izer and 3 inputs of 1 000 kg urine N over 4 years (U+N). Each of the (C) and (U+N) treatments had four replicate plots. The procedure for resin spikes burial is detailed in Section 4.7. One resin spike was buried in each plot C l x 2 m). Before resin spikes were buried, the pasture was mowed and soil cores (0-7 . 5 cm) were collected for NH/-N and N03--N analysis, using the standard 2 M KCI method. After 1 2 days in the soil the resin spikes were removed and extracted for adsorbed NH/-N and N03--N. Soil cores were also collected for 2 M KCI -extractable NH/-N and N03--N detennination at day 1 2 . Pasture was cut to determine dry matter production and then analyzed for total N to assess the N uptake on each plot over the 1 2-day period. Analysis of variance of data was carried out using the GLM procedure in SAS for Windows (Version 8). Mean comparisons were done using Fishers LSD at 5% level of significance. This first in situ resin spike experiment produced promising results (Table 4 .8). 9 2 Table 4 . 8 . Soil NH/-N and N03--N assessed by 2 M KCl extraction and resm adsorption, together with pasture N uptake over 1 2 days at 2 sites . Means with common letters are not significantly different (P 1 00% at both sites. This indicated the resin spike's potential to detect the high soil spatial variabil ity associated with soil mineral N. 9 5 Table 4 . 9 Levels of mineral N in two soils as measured by 2 M KCl extraction and resin spikes. Measurements were made over a 7 day period in the field and also after incubation for 7 days in the laboratory. A) C S ' amp Ite Field 2 M KCI -extractable N Ilg N/g soil Rep Da 0 Dav 7 NH/-N N03'-N NH/-N N03'-N I 5 .4 9 . 7 1 0 0 2 7 . 5 9 . 7 3 . 8 0 3 7 . 5 9 . 7 3 . 7 0 Average 6.8 9.7 5.8 0 Resin adsorbed N Ilg N/5 cm217days Rep NH/-N N03'-N I 0 33 2 0 60 3 3 83 4 8 205 5 3 85 6 3 93 7 0 3 8 1 3 2 5 9 0 5 8 1 0 3 73 Average 3.3 71 .8 B) Steep site Field 2 M KCI -extractable N Ilg N/g soil Rep Da 0 Dav 7 NH/-N N03'-N NH/-N N03'-N 1 5 0 6 0 2 5 0 6 0 3 5 0 6 0 Average 5 0 6 0 Resin adsorbed N Ilg N/5 cm217days Rep NH/-N N03'-N 1 5 3 2 8 1 0 3 0 0 4 0 1 20 5 3 2 0 6 0 60 7 0 5 0 8 0 0 9 0 2 8 1 0 0 8 Average 1 .6 29.9 Laboratory_ 2 M KCI-extractable N Ilg N/g soil After 7 days incubation Rep NH/-N N03'-N I 4.3 3 8 .3 2 3 . 2 39.3 3 5.3 4 1 Average 4.3 39,S Resin adsorbed N III N/5 cm2l7d1lYS Rep NH/-N N03'-N I 0 225 2 0 247 3 I 305 Average 0.33 259 Laborato!:y 2 M KCl-extractable N llg N/g soil After 7 days incubation Rep NH/-N N03'-N 1 6 0 2 9 0 3 5 0 Average 6,7 0 Resin adsorbed N Il� N/5 cm217days Rep NH4+-N N03'-N 1 0 1 0 2 8 20 3 0 1 5 AverCiKe 2.7 15 4.6 Modelling of in situ N adsorption to resin membrane spikes 4.6.1 Introduction 9 6 In previous sections, the potential of in situ use of ion exchange resin membranes for obtaining a sensitive measurement of the N supplying power of soils over time was identified. When using ion exchange resin membranes for the measurements of N dynamics in soil, it is necessary to understand the process of N adsorption to the resin, and the influence of other factors such as N diffusion rate, soil moisture content, and the resin burial period in the soil . As was noted earlier however, it is very difficult to design experiments to measure changes in resin-adsorbed N under different levels of a single factor. For example, increasing soil moisture is likely to increase the N diffusion rate, but also increase N availability by mineralisation. Thus, the final measurement of resin-adsorbed N is a result of both increased diffusion and increased N availability, and it is difficult to separate out these two effects. Hence, simulation of N adsorption to resin in soil offers an alternative way to identify the influence of experimental conditions on the adsorption process. This study involved modelling the experimental work described in Section 4.4. 1 .2 . Only N03'-N adsorption to resin spikes was modelled, as NH/-N adsorption was complicated by competition from other cations on the resin membrane and the soil exchange surfaces. Once constructed, the model was used to predict N adsorption to the resin spike in a given burial period, the N depletion zone in the soil when N is adsorbed by the resin spike, and the influence of moisture, initial N concentration and temperature on resin N adsorption from soil . 4.6.2 Basic equations for solute diffusion in soil In this section the basic partial differential equation for N03--N diffusion in soil is derived. To do this we consider a notional rectangular box of soil (Fig. 4 .9) with the dimensions of 6,x (L) �y (L) L (L). This notional box receives or loses N03--N by 97 molecular diffusion along the x and y coordinates. A constant concentration is assumed in the z coordinate and so there is no flux along it. The flux density (ML-2r 1 ) in the x and y coordinates is described by Fick's first law (Section 4.2.2) as where dC q = - B D -x s dx dC dy (4.2) (4.3) qx. = Amount ofN03--N moving across unit-cross-section per unit time in the x direction qy = Amount of N03--N moving across unit-cross-section per unit time in the y direction B = Volumetric water content of the soil (L3 L-3) Ds= Diffusion coefficient for N03--N in the soil (see Section 4 . 6 . 3 ) C = N03--N concentration in soil solution (M L-3) x = Coordinate in the x direction (L) y = Coordinate in the y direction (L) f..x f..y / / L / J{ / / Fig. 4 . 9 Notional box of soi l . 98 It is assumed that N03--N is conserved over the period of interest. i .e. there is no significant plant uptake, nitrification, denitrification, or immobilization. Thus, during any time interval 6:.t (T), the net N03--N movement in and out of the box equals the accumulation of N03--N in the box . Now Qin =[(qx. 6:.y. L) + (qy. tu. L)] 6:.t Qout =[(qx+ill<. 6:.y. L) + (qy+lly. tu. L) ] 6:.t where Qin = quantity of N03--N diffusing into the box (M) Qout = Quantity of N03--N diffusing out of the box (M) (4.4) (4. 5 ) Given there i s no adsorption, the accumulation o f N03--N i n the box over 6:.t, 6:.M i s given by 6:.M = e 6:.y tu L 6:.C (4.6) and as indicated above conservation of mass implies the (4.7) Substituting the equations (4.4), (4 . 5 ) and (4. 6 ) into equation (4.7) gives. e 6:.y tu L 6:.C = [[(qx 6:.y L) + (qy tu L )] 6:.t ] - [[(qX+llX 6:.y L) + (qy+lly tu L)]6:.t] (4. 8) Dividing both sides by (tu 6:.y L 6:.t) and a little arranging gives (4 .9) 99 This finite difference equation can be changed into a partial differential equation by taking limits as ill � 0 , �y � ° and �t � ° (i .e making the box infinitely small in the x and y coordinates and changing the time interval into an instant.) The resulting equation is () 8C _ 8qx 8qy - - - - - - 8x 8x ay (4. 1 0) Then by substituting the expressions for qx and qy from equation (4.2) and (4.3) into equation (4. 1 0) and dividing by () gives. (4. 1 1 ) This is one form of the equation known as Fick ' s second law of diffusion as applied to soil . As it stands, equation (4 . 1 1 ) cannot be solved. It tel ls how the solute (N03--N) concentration changes with time and distance in the x and y coordinates. To find the solute concentration at a particular point (x,y) and time it is necessary to specify the initial conditions and boundary conditions. In the box considered here, the initial condition in the soil i s C = C for x > 0, y > 0, t = ° (4. 1 2) where Ci is the uniform concentration of N03 --N in the soil solution. The non uniform boundary condition at the surface of the "resin spike" will be discussed later. Due to the complex geometry of the boundary condition, equation (4. 1 1 ) could not be solved analytically. Thus , a model was developed to solve this equation numerically. 4.6.3 Model development 1 00 The experimental design described in Section 4.4. 1 .2 was modelled. The effective soil volume for adsorption was considered as a 1 25 cm3 (5 x 5 x 5 cm) cube. This had 400 (20x20) cells or compartments, each 0.25 cm (�y) x 0.25 cm (�x) x 5 cm (F ig.4 . I O) . The resin membrane was in contact with the soil in cells (�y, �x) at ( 1 0,9), ( 1 0, 1 0), ( 1 0, 1 1 ) and ( 1 0, 1 2) . To ease computation, it was assumed that diffusion was symmetrical about a line through the mid point of the resin spike (see F ig. 4. 1 0). D iffusion was then calculated for half of the area (shaded area in F ig.4 . 1 0) . The resulting adsorption in cells ( 1 0,9) and ( 1 0, 1 0) was then multiplied by 2 to give the total adsorption to the spike. x 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 2 3 4 5 6 7 8 9 y 1 0 �e In 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 Fig. 4 . 1 0 Modelled experimental soil cube with resin spike inserted, viewed from above. The anion exchange resin membrane is facing cells ( 1 0,9), C l 0, 1 0), ( 1 0, 1 1 ), and ( 1 0, 1 2) and the cat ion exchange resin facing the cells ( 1 1 ,9), ( 1 1 , 1 0), ( 1 1 , 1 1 ) and ( 1 1 , 1 2) . 1 0 1 The model recognizes that the quantity of N adsorbed depends on the relation between N in soil solution and N on the exchange complex, and the rate of diffusion of N towards the resin surface. The relationship between N in soil solution and on the surface is assumed to be given b y a 1 : 1 adsorption isotherm. The 1 : 1 isotherm assumes that there i s n o selectivity i n N03- -N adsorption to resins (see Section 4 . 4) and that N03--N competes equally with other anions initially in the soil solution and on the resin surface for sorption sites. Thus where C X = Cmax X max C = N03 --N concentration of soil solution at the resin contact surface (llg/cm3 soil solution) (4. 1 3 ) Cmax = operational maximum N03--N concentration if N03- was the only amon existing in soil solution. (llglcm3 soil solution) X = N03--N adsorbed to resin (llglcm2 resin) Xmax = maximum N03--N adsorption capacity of resin ( llg/cm2 resin) The maximum anion adsorption capacity of the 5 cm2 of anion membrane was 2 1 79 Ilg (Section 4 . 4 ) . Thus, the maximum anion adsorption capacity of one cell is calculated as Xmax = 2 1 79/4 (4. 1 4) The operational maximum N 03--N concentration, if the soil solution contained only N03--N ions, (Cmax) was estimated using an indirect method. It was assumed in the model that the content of N03--N at the start of the period of N adsorption to the resin was constant at 1 92 Ilg N03--N per g soil . This was the measured 2 M KCI -extractable N03--N at 7 and 14 days after resin burial (Fig. 4 . 8 ) . 1 02 The electrical conductivity (EC) of different concentrations of NH4N03 solutions was measured and is plotted in Fig. 4 . 1 1 . At the same time, the electrical conductivity of the soil that was used in the experiment (4 . 4 . 1 .2) was measured in a 1 : 5 (soi l : water) suspensIOn. >. 0.7 .. ':; 0.6 :;:; u 0.5 ::J � E c 0.4 0 u U en 0.3 'i !. u 0.2 '� .. 0.1 u Cl W 0 0 0.001 0.002 0.003 0.004 0.005 0.006 Concentration of N�N03 ( moles/I) Fig. 4 . 1 1 : Electrical conductivity of differen t N�N03 concentrations at 2 0 ° C . The electrical conductivity of the soil that was used in the experiment was 0.46 m S/cm. The concentration of a N�N03 solution that has an EC of 0.46 mS/cm is 0 . 00 3 5 M (Fig. 4 . 1 1 ) . Thus the operational maximum N03--N concentration in the soil solution was found as (0.003 5)( 1 4) g N03--N/liter or 0 . 049 g N03--N/liter. In 5 mL of soil suspension this amounts to 245 Jlg N03--N per g soi l . A s the s o i l suspension was 1 : 5 (soi l : water), this would be the quantity of N03--N i n 1 g of soil if the existing soil solution contained only NH4N03 . In fact, the experimental soil contained 1 92 �g N03--N per g soil. Thus the operational maximum soil solution concentration of N03--N was greater than the initial soil solution concentration (Ci , defined below) by a factor of 245/ 1 92 or 1 .2 8 . A t the beginning o f the experiment the resin membrane does not have any adsorbed N. Thus, initial ly x= o (4. 1 5) 1 03 Before the simulation process started, the initial soil N03--N concentration in each cell (Ci, /lg N/cm3 soil solution) was calculated as follows : where G = Soil N03--N concentration ( /lg N03--N/g soil) Pb = Bulk density (g cm- 3) (4. 1 6) The amount of N03--N in each cell (M, /lg N03--N ) is calculated as follows (4. 1 7) The diffusion coefficient of the solute in soil (Ds) was calculated according to Tinker and Nye (2000). (4. 1 8) where the diffusion coefficient of the solute in free solution (Dl) was taken as 1 . 3 cm2/day at 2 5°C (Tinker and ye 2000). The impedance (tortuosity) factor if) for the relevant soil moisture content was obtained from the relationship in Fig. 4 . 2 in Section 4.2.2 . Consider a cell (y,x). The fluxes into the cell in the x and y direction are (y-1 ,x) 1 � �y .. ----. r (y,x-1 ) (y,x) � (Y,x+ 1 ) t (y+ 1 ,x) ( 4. 1 9) (4.20) 1 04 There are similar fluxes out the other side of the cel l . The net gain (or loss) of N03'-N by the cell over a time interval 6..t, 6..My,x is given by 6..M(y,x) = [ (Fy (y· l ,x) - Fy(y,x)) + ( Fx (y,x· l ) - Fx(y,x) ] 6..t (4 .2 1 ) This then changes the N content in each cell. (4.22) The model then uses the new M value to calculate a new N03'-N concentration value (C) for the soil solution in each cell. C = (M/6) (6..y) (�) (L) (4 . 2 3 ) A t the first time step the resin membrane i n contact with the soil i n compartments ( 1 0,9), ( 1 0, 1 0), ( 1 0 . 1 1 ) and ( 1 0, 1 2) adsorbs N03'-N from the soil surface. As noted earlier, adsorption is modelled for only half of the resin, as adsorption is assumed to be symmetrical. The interface between the soil and the resin surface in compartment ( 1 0,9) is designated as s I and the corresponding interface between the soil and the resin surface in compartment ( 1 0, 1 0) is designated as s2. The amount o f N03'-N adsorbed by the resin is equal to the amount of N diffusing towards the resin surface from compartments ( 1 0,9) and ( 1 0, 1 0). This flux of N (Fs l and Fs2) towards the resin is calculated using Fick's law as described previously. Fs t = 6 Ds � L ( C( I O,9). Cst )! (6..y/2) Fs2 = 6 Ds � L ( C( l O, IO)-Cs2)/( 6..y/2) C = N03'-N concentration ()lg N03'-N/cm3 soil solution) Cs t = N03'-N concentration at the resin membrane surface in cell ( 1 0,9) Cs2 = N03'-N concentration at the resin membrane surface in cell ( 1 0, 1 0) 6 = Volumetric moisture content Ds = Diffusion coefficient of the solute in soil (cm2/day) L = Length of resin membrane (5 cm) (4.24) (4. 2 5 ) I 1 05 The flux towards the surface calculated in each time step is then added to the quantity adsorbed on the resin Where Xl (t+L'.t) = Xl (t) + Fs I X2(t+L'.t ) = X2(t) + Fs2 Xl= N03--N adsorbed to resin from ( 1 0 . 9) compartment X2= N03--N adsorbed to resin from ( 1 0 . 1 0) compartment M = time step (4.26) (4.27) As the N03--N is adsorbed onto the resin membrane, the N03--N concentration in the soil solution adjacent to resin membrane surface changes. The new N03--N concentration in soil solution at the resin membrane surface is calcul ated using the 1 : 1 adsorption isotherm Csl (t+L'.t ) = (XI(tYXmax) (Cmax) Cs2(t+L'.t ) = (X2(tYXmax) (Cmax) (4.28) (4 .29) As the resin surface solution concentration (Csl and Cs2) has changed, a concentration gradient will develop in the cells o f the soil matrix near the resin membrane. The change of N03--N concentration in cells of the soil matrix leads to diffusion o f N03--N towards the resin membrane surface and N03--N adsorption to the resin. The model simulates these actions continuously at given time steps for the period that the resin spike is buried in the soil. 4.6.4 Model output 4.6.4. 1 The effect of time of burial on N adsorption by a resin Spi in soil Overall, the model predictions of resin adsorption of N03--N are reasonably close to the actual measured values (Fig. 4. 1 2) . Although the model forecast the final adsorption 1 06 reasonably accurately, it slightly overestimated the rate of adsorption in the early stages. The model estimation of N03--N adsorption to resin depends on N diffusion towards the resin. As pointed out in Section 4.6.2, the value for the diffusion coefficient (Ds) in the model was obtained from the literature. This may account for the overestimation of N adsorption in the early days after placement of the resin spike. It was found in Section 4.4 that the 5 cm2 resin membrane has a capacity to adsorb 2 1 79 Ilg N03--N. However, the measured and modelled N adsorption is less than the maximum capacity, reflecting competition from other anions. 1 800 � 1 600 .... 1 400 0 Z _ 1 200 � N • QI E 1 000 .= I,) .. It) 0 m 800 en � .:: ca 600 • Measured C Vi 400 -0- Modelled QI Cl:: 200 0 o 1 2 3 4 5 6 7 8 9 10 1 1 12 1 3 1 4 1 5 Days after resin spike burial Fig. 4 . 1 2 Measured and modelled N03--N adsorption to resin spikes with time. The modelled concentrations of N03--N in soil solution in each cell, 1 day after the resin spike has been placed in the soil, are presented in Fig. 4. 1 3 . It is apparent that there are marked concentration gradients in the direction of resin spike. These concentration gradients encourage diffusion of N03--N towards the resin. The data recalculated as the quantity of N03--N (Ilg) in each cell are presented in Fig. 4. 1 4 to visually depict the depletion zone By day 7 (Fig. 4. 1 5 and 4. 1 6) all soil compartments in the system have similar N levels and therefore the rate of diffusion approaches to zero. By this time, the resin membrane has slightly reduced the soil nitrate concentration throughout the 5 cm x 5 cm zone under consideration around the spike 1 07 cm 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 0.25 737 737 737 738 735 733 732 730 7211 728 0.50 737 737 736 735 733 732 7211 728 728 725 0.75 737 738 735 733 731 728 725 722 720 718 1.00 738 735 733 731 727 723 718 713 7011 707 1.25 735 734 731 728 722 718 708 700 8114 8110 1.50 734 732 7211 724 717 707 8115 113 872 118 1.75 733 731 727 720 711 8117 881 682 845 834 2.00 732 7211 725 717 705 ea 11 687 8311 810 5112 2.25 731 7211 724 718 703 ea4 857 618 567 537 2.50 731 7211 724 718 704 887 180 814 If. 41f R_ln 2.75 732 730 721 7111 710 8117 182 171 8111 IU 3.00 733 732 728 723 717 708 700 8111 702 701 3.25 734 733 731 727 723 718 713 711 713 714 3.50 735 735 733 731 721 725 722 721 721 721 3.75 731 738 735 733 732 730 721 721 727 727 4.00 737 737 738 735 734 733 732 732 732 732 4.25 731 737 737 737 738 735 735 735 734 734 4.50 731 731 731 737 737 737 736 731 736 738 4.75 738 731 731 731 731 737 737 737 737 737 5.00 731 731 731 738 731 731 731 731 738 738 F ig. 4 . 1 3 Modelled N03--N concentrat ion (Jlglcm3 soil solution) m each soil compartment (0.25 x 0 .25 x 5 cm) 1 day after placement of the resin spike in soil. The asymmetrical depletion pattern is caused by the placement of the anion resin strip on the side of the spike apparently closest to the top of page. (This diagram shows only half of the system). 1 8 a; 1 6 () z , ' (") 1 4 o Z g 1 2 1 0 2 X a · XIS (cm) _ 10 _ 12 _ 14 c::::::::J 1 6 _ 18 Fig. 4. 14 . Soil N03--N (Ilg/cell) d istribution in each soil compartment (0 .25x 0.25x l cm) 1 day after placement of the resin spike in soil. The res in spike is placed at 2-3 cm on the X axis and at 2. 5 cm on the Y axis. The asymmetrical depletion pattern is caused by the placement of the anion resin strip on the side of the spike apparently closest to the reader. 1 08 cm 0.25 O.SO 0.75 1 .00 1.25 1.50 1.75 2.00 2.25 2.50 0.25 ... ••• .. 7 ..7 •• 7 ... .. , , .. , .. "s 0.50 ... ... 197 197 .n .. , ... ••• US US 0.75 ... ... ... .. 7 ..7 • 11 ... , .. .95 .95 1.00 , .. ... ... ..7 ..7 .. 7 .. , ... .. S .95 1 .25 ••• ••• ••• U' .n • n ••• ••• ..S .95 1.50 ... ... ••• U' ... '97 .., ••• ..S US 1.75 700 , .. , .. ... ... '97 ..7 ••• '95 .95 2.00 700 700 700 ... ••• , .. • n • •• US ••• 2.25 701 701 701 700 700 ... ... .n US ... 2.50 702 702 702 701 701 700 ... U' IM 1.2 R •• ln 2.75 703 703 703 703 702 702 701 701 703 703 3.00 70. 704 704 704 704 703 703 703 704 704 3.25 70. 705 705 705 705 705 705 705 705 705 3.50 707 707 707 70. 70. 70. 70. 70. 70. 70. 3.75 70. 70. 70. 707 707 707 707 707 707 707 4.00 701 701 70. 70. 70. 70. 70. 70. 70. 70. 4.25 70. 701 701 70. 701 70. 709 701 701 701 4.50 710 710 710 710 710 710 710 710 710 710 4.75 711 711 710 710 710 710 710 710 710 710 5.00 711 711 711 711 711 710 710 710 710 710 F ig. 4 . 1 5 Modelled N03·-N concentration (/lg!cm3 soil solution) m each soil compartment (0.25 x 0.25 x 5 cm) 7 days after placement of the resin spike in soil (This diagram shows only half of the system) . Qi t) Z I I C') 0 Z (j) ::J.. 1 8 1 6 14 1 2 1 0 ... 2 Xa . XIS (crn) -,... ,.. _ 10 _ 12 _ 14 � 16 18 Fig. 4. 1 6. Soil N03--N (/lglce ll) distribution in each soil compartment (0 .25x 0 .25xl cm) 7 days after placement of the resin spike in soil . The resin spike is placed at 2-3 cm on the X axis and at 2 . 5 cm on the Y axis. 1 09 It can be seen that the concentration gradients in the cells on one side of the spike are much steeper than on the other side. This is because only one side of the resin spike has anion resin membrane attached. 4.6.4.2 Effect of soil moisture on N adsorption by resin spikes Adsorption of N03--N to resin with time was estimated by the model at different soil moisture levels and illustrated in Fig.4 . 1 7 . Under very dry conditions N adsorption by resin spikes in soil is predicted to be much slower than under moist conditions . However, in wetter soils ( W = 0.3 and 0 . 4) the effect o f changes i n moisture content on N03--N adsorption is not as great. Thus in wet soils, resin spikes would need to be buried in soil for shorter periods to achieve satisfactory comparisons while in dry soi ls longer burial periods may be more suitable. This is further illustrated by the effect of soil moisture on the soil N depletion zone after N adsorption to resin (Fig. 4 . 1 8). 1 800.-------------------------------------, 1600 Z .. 1400 � E 1: � 1 200 o - 1000 � � 800 � ,; c Z 600 Ui Cl Cl a: ::I. 400 200 o __ _r�--r_�_r_,--�_r�--r_�_r_,--�� o 1 2 3 4 5 6 7 8 9 10 1 1 1 2 1 3 14 1 5 Duration of resin spike burial (days) 1 ___ W= 0.1 -.-W=0.2 --.-W=0.3 -+-W=O.4 ! Fig. 4. 1 7 Modelled effect of soil moisture content on N03--N adsorption by resin spikes with time after burial. W = gravimetric moisture content. A) Soil moisture = 0.26 (W IW) 18 � z ;S 14 Z Cl � 2 Xax' IS (cm) B) Soil moisture = 0. 1 (w/w) 5 _ '0 _ ,2 _ ,4 = 16 _ ,8 _ ,0 _ 12 _ ,4 = '6 _ ,8 1 1 0 F ig . 4. 1 8 . Effect of soil moisture on so il N03--N ()lglcell) distribution 7 days after placement of the resin spike in soil. The resin spike is placed at 2-3 cm on the X axis and at 2 . 5 cm on the Y axis. The asymmetrical depletion pattern is caused by the placement of the resin strip on the side of the spike 'apparently closest' to the reader. 4.6.4.3 Effect of initial soil N concentration 1 1 1 The model was used to estimate N adsorption to resin membrane spikes with time at different initial soil N03--N concentrations (Fig 4. 1 9). It can be seen that as the soil N03--N concentration increases, the time taken to reach equilibrium decreases . 1800 1600 � :: z N 1400 " E � U 1200 J:Z � It) 0 Z 1000 en " . '" 800 III 0 . c Z 'iij 600 en � � a::: 400 200 0 o 1 2 3 4 5 6 7 8 9 10 1 1 12 1 3 14 1 5 Duration of resin spike burial (days) 1---'- 1 92 -+-- 1 00 -50 -l:r-25 -0- 10 -+-5 I Fig. 4. 1 9 Modelled effect of initial soil 03--N concentration (flg N03--N/g soil) on N03--N adsorption by resin spikes with time after burial. The gravimetric moisture content is 0.26 (w/w). At low soil N03 --N concentrations, resin spikes need to be buried for longer periods for better comparisons. At higher soil N03--N concentrations, resin spikes would need to be buried in soil for shorter periods to achieve satisfactory comparisons. Hence, in experiments with N fertilizer and urine application the resin spike would need to remain in the soil for only short periods. 4.6.4.4 Effect of temperature on N adsorption by resin spike in soil As discussed in Section 4.6.2, the diffusion coefficient of the solute in soil is calculated using the diffusion coefficient of the solute in free solution (DJ) and the impedance factor (j) . The diffusion coefficient of the solute in free solution (DJ) is generally affected by solution temperature through a change in solution viscosity. The diffusion 1 1 2 coefficient in solution (DJ) is directly proportional to temperature (K) and inversely proportional to viscosity (Scotter D.R. pers.comm). Hence, new values were calculated for D\ using the viscosity values of water at different temperatures. These values of D\ were 1 . 3 cm2/day (25°C), 0.98 cm2/day ( 1 S 0C) and 0 . 7 1 cm2/day (SOC). The N adsorption by the model (Fig. 4.20) was then estimated at these 3 temperatures. It can be seen that during the early period of the adsorption, under warm soil temperatures, the resin would adsorb more N than under cool conditions. After longer burial periods there seems to be no significant effect of temperature on N adsorption to re sin. The effect of temperature is not as large as the effect of soil moisture and initial soil N03--N concentrations on N adsorption to resin spikes . 1 800 ,-------------------------------------, 1600 Z N 1400 " E Q) u 1 200 .IJ It) � Z 1 000 " , IV 0 800 C Z .ijj 600 Q) 0) Ill: ::L 400 200 o +----r----�--�--�----�--_r----�� o 2 4 6 8 1 0 1 2 14 1 6 Days after resin spike buried I-A-- Temp = 5 C ___ Temp = 1 5 C -'- Temp= 25 C I Fig. 4.20. Modelled effect of temperature on N adsorption by resin spikes with time. The moisture content of the soil is 0.26 (w/w). 4.7 Recommended procedure Following the preliminary experiments described in the prevIOUS sections the recommended method for in situ measurement of soil mineral N using resin spikes is as outlined below. 1 1 3 a) Spike construction: Strips o f cation and anion membranes ( 1 x 6 cm) are taped t o both sides (one strip per side) of an acrylic spike (2 x 0 . 5 x 8 cm ) using water resistant tape, leaving 5 cm2 of membrane area for adsorption. 2 em O.5 em � ... ... 1.4� 8em 5em .... ..... 1�1r I.. l" V Fig. 4.2 1 . Schematic diagram of resin spike. b) Procedure: S Acrylic spike o il surface Resin membrane ( 5 cm2 Anion/Cation) Water resistant tape The spikes should be kept in a 2-3M KCI solution overnight for regeneration before use. To bury resin spikes, a thin vertical slot is made in the soil using a small steel bar of similar dimensions to the spike. The resin spike is placed into the slot carefully without disturbing the soil further. The resin spike is removed from the soil after an appropriate time, rinsed with deionized water, placed in a centrifuge tube containing 2 5 mL of 2 M KCl , and shaken for 1 hour. The NH/-N and N03--N adsorbed to the resin are displaced into the KCl eluent and their concentrations determined b y a Technicon Auto Analyser (Searle, 1 975 ; Blakemore et al., 1 987). 1 1 4 The length of time the resin spike is left in the soil depends on the soil N concentration and moisture content. Shorter burial periods are suitable under moist and high soil N situations while longer periods (7 or 1 4 days) are necessary for dry and low soil N situations. 4.8 Discussion In this study, it has b een demonstrated that the in situ measurement of mineral N availability using resin spikes may be an effective technique to adopt in field experiments. Initial experiments described in this chapter demonstrated that when cation and anion exchange membranes are directly buried in soil, N is adsorbed on to the membrane and this resin-adsorbed N b ears a relationship to soil N availability to pasture plants. It was evident that adsorption of N03--N was more efficient and uniform than adsorption of NH/-N. This appeared to be due in part to the greater competition for NH/-N adsorption from other cations in the soil/membrane system. Experiments conducted in the field demonstrated that in situ measurement of N availability by resin membrane spikes was able to detect measurable quantities of N even when the standard 2 M KCI extraction method could detect only negligible amounts of soil N. This probably reflects in part the greater volume of soil influenced by the resin strip. More importantly the resin procedure could discriminate soil fertility differences, both in intensively managed experimental plots and also among different hill country landscapes. The apparent advantages of the resin method can be explained as follows: The most common method of estimating soil mineral N availability in soil is the 2 M KCI extraction method (see Section 3 . 3 .4) . This measurement gives an estimate of the available N pool at the time the soil sample is collected. There are many processes that affect the available soil N pool as illustrated in Fig. 4.22. Additions to the pool through mineralisation are countered by losses, such as leaching and plant uptake. Thus, although considerable quantities of mineral N may be produced and consumed, the amount in the soil at any one time may be small . This applies equally to soils with high 1 1 5 and low fertility. It is for this reason that N extractable with 2 M KCI is not usually a good indicator of N availability to plants. NH3 Volatilisation Cation res m N�+ Mineralisation Plant uptake Available N pool Nitrification Irnrno bilisation Organic matter Denitrification Leaching Anion resin Fig. 4.22 Schematic diagram of processes controllin g the available soil N. It appears that in situ N measurement using resin membrane spikes may have the potential to at least partially overcome this problem. When resin spikes are embedded in the soil they can adsorb N over time, simultaneously with other processes. Therefore, the resins act as a net sink for mineral N and so give a better estimate of N availability over time. In this way, the resin data indicates a measure of the N flux through the available N pool. Resin-adsorbed N may be able to be used as a potential mineralisation index. The results in Section 4 . 5 . 2 give some evidence for thi s . When the soils were incubated in the l aboratary, larger amounts of resin-adsorbed N03--N were found in the campsite soils than in the soils from the steep slope. S imilarly, when resin spikes were directly buried in the field, the average resin-adsorbed N03--N from 1 0 resin spikes was twice as 1 1 6 high in campsites as in steep slopes. As discussed in the literature review, campsites receive large amounts of dung and urine, and have large amounts of easily decomposable organic material, whereas in steep slopes substantial amounts of resistant organic matter accumulates. Qian and Schoenau ( 1 99 5 ) used anion exchange resin membranes (AEM) in an incubation experiment to determine the influence of different tillage systems and landscape positions on mineralisable N. The N03--N accumulated on AEM strips buried in soil for 2 weeks showed the expected difference among soi ls due to tillage treatments and slope position. During the experiments described in this chapter, the resin spike ' s potential ability to detect spatial variability in soils was noted. Cain et al. ( 1 999) used resin membranes to study spatial and temporal variation in N availability in coastal dunes. Huang and Schoenau ( 1 996), also used resin membranes to study microsite assessment of forest soil nitrogen, phosphorus and potassium supply rates. It was noted in the literature review that recently scientists have been suggesting precise N application for hill country pasture using GIS and GPS systems (Gillingham et al., 1 999). Resin spikes may be useful in developing spatial N maps of the hill country landscapes. The computer model developed in this chapter based on the diffusion of N03--N closely simulated the actual experimental data. This has enabled an understanding of the N adsorption process and the influence of other factors on N adsorption by resin spikes. The model suggested that resin adsorption would continue until the rate of diffusion approached zero. As N adsorption to resin is diffusion dependent, soil moisture, soil temperature, and soil tortuosity will also be maj or factors influencing N adsorption. In warm, moist, relativel y high soil N conditions, N diffusion to resin is fast. However, in dry soils the diffusive resistance is high. Under such conditions, resin-adsorbed N would probab l y not correlate w e l l with soil N concentration. This is because the mobility of the N rather than its quantity, would control the resin-adsorption. However, resin-adsorbed N under these conditions might still provide a useful index of N flux and plant N availability in the soil. Even at high concentrations, soil N03--N is largely unavailable for plant uptake, when its movement in the soil i s restricted by low soil moisture . 1 1 7 Resin-adsorbed N may therefore be a better measurement of the actual plant availabil ity of mineral N under natural conditions. The experiment described in Section 4.4. 1 .2 and the diffusion model described in Section 4.6 could be used as an effective way of determining the diffusion coefficient of N03- ion in different soil conditions and soil types. Estimates of resin-adsorbed N are not expressed as the quantity of N per unit weight of soil . This is one disadvantage of the resin method. This can be overcome to some extent by estimating the N depletion zone when resin spikes are kept in the soil . It was demonstrated in S ection 4.6 .3 .5 that the depletion zone depends on other factors such as resin spike burial time and soil moisture. The possible potential of N measurement by resin spikes has been outlined in this Chapter. Resin spikes were then used in the field experiment described later in Chapter 5 and the incubation experiment described in Chapter 6 . As the N availability was also measured by the standard 2 M KCI method in these experiments, relationships between resin-adsorbed N and 2 M KCl -extractable N in different soils are discussed in detail in Chapters 5 and 6 . 1 1 8 CHAPTER S FIELD INVESTIGATION OF NITROGEN DYNAMICS UNDER URINE PATCHES IN NORTH ISLAND HILL COUNTRY PASTURES 5.1 Introduction The literature review revealed the importance of N cycling VIa animal excreta, especially in hill country where there are significant losses of N due to transfer of excreta from sloping land to flat camp areas . The preliminary experiment conducted at Waipawa, and described in Chapter 3, investigated the fate of urine N in steep and flat sites within a hill country pasture. In that experiment only about 20% of the added urine N was utilised by pasture and the rest appeared to be lost in some way from the mineral N pool. A concern with the conduct of the preliminary experiment at Waipawa was that the frequency of sampling was not sufficient to detect rapid changes in the mineral N pool . In particular, although little N03--N appeared t o form after urine application, the possibility could not be ruled out that nitrification was occurring but the N03--N so formed had then been leached from the profile before it could be detected. In Chapter 4, a method, using ion exchange resin membrane spikes, for in situ N measurement over time was described. These resin spikes can be used to effectively monitor mineral N changes in a urine patch and overcome in part the difficulty of monitoring continuous changes in soil mineral N by sampling at particular points in time. Also, it was suggested in Chapter 3 that urine N loss by NH3 volatilisation was substantial in hill country even during the cold and moist winter. There have been limited studies that have measured NH3 volatilisation in North Island hill country 1 1 9 pasture in New Zealand. Thus, it is important to obtain more information about NH3 volatilisation losses in hill country pasture. The obj ective of this study was to further monitor and understand the transformations and fate of urine N in different landscapes with contrasting topography in North Island hill country pasture. The use of resin spikes in this experiment would be useful to elaborate on the N transformations in urine patches and as well as providing further assessment of the technique. 5.2 Materials and methods 5.2.1 Site description This experiment was carried out at Ballantrae AgResearch hill country research station, 20 km north of P almerston North in the southern Ruahine range, during the mid-winter to early spring of 2000. The soils at this experimental site are Yellow Brown Earths and Intergrades to Yell ow Grey Earths (NZ genetic soil classification). In addition, these soils can be identified as Brown soils (NZ soil c l assification version 3) and Dystrochrepts (US soil Taxonomy), according to Hewitt ( 1 998). As many field studies have been carried out on this site in recent years, well-defined landscapes of contrasting fertility could be identified easily. Some soil properties of low and high fertility landscapes of the site were described in Sakadevan ( 1 99 1 ), Machado ( 1 994 ) and Carran et al. ( 1 995). 5.2.2 Field layout and soil sampling This experiment was carried out at two contrasting adj acent sites . The first was a flat site (F) that was a campsite for sheep and the second was a more steeply (20-25°) sloping site (S). The total C %, total N %, CIN ratio, Olsen P and soil pH of steep and flat sites were 5 . 5 , 5 .4; 0.4 ,0. 5 ; 1 4 , 1 2 ; 2 9 . 8 Ilg/g, 49.3 Ilg/g and 5 . 8 , 5 .4 respectively. In each experimental site, twelve 0.5 x 1 m plots were established and arranged into four blocks of three plots. Three treatments were assigned at random within each of four blocks to give a randomised complete block design. The plots were established with buffer zones around the edges to prevent interference from other treatments by run 1 2 0 off etc. Each 0 . 5 x l m plot was divided into two 0 . 5 x 0 . 5 m sections. One section was used to take soil samples and the other was used for resin spike burial and for measurement of pasture yield. 5.2.3 Treatments The treatments described in Chapter 3 were changed for this experiment based on measurements of available mineral N (NH/-N + N03--N) in newly-deposited urine patches at the site j ust before the experiment. The quantities of available soil mineral N in newly deposited urine patches were higher than the quantities of soil mineral N found in the 400 kg urine Nlha treated plots in the Waipawa experiment described in Chapter 3 , suggesting that the N application rate in the urine patches may have been higher than 400 kg urine Nlha. Other works have also suggested 500 kg Nlha as the equivalent application rate per urine patch (Haynes and Williams, 1 99 3 ; Sherlock and Goh, 1 984), although S akadevan et al. ( 1 993) used 280 kg N/ha. Hence, the urine application rates for this experiment were taken as, 0 N (control), 280 kg urine Nlha and 560 kg urine Nlha (Table 5 . 1 ) . Table 5 . 1 Treatments used in the experiment at Ballantrae AgResearch hill country research station investigating the fate of simulated urine N applied to hill country pasture. Rate Site N kg/ha Flat Steep 0 FO SO 280 F280 S280 560 F560 S560 * Applied as 2 litres of 0 . 7% N solutIOn * * Applied as 2 litres of 1 .4 %N solutIOn Table 5 . I A Constituents of the synthetic urine solution (PH=7 . 8) . Constituent 0.7% N solution (gll) 1 .4% N solution (gll) Urea 1 3 .7 27.4 Glycine 3 .78 7 .56 KCI 3 .64 7.28 K2S04 1 .4 2 .8 KHC03 1 0.5 2 1 .0 1 2 1 Synthetic urine solution was made using the same constituents used in the Waipawa experiment, described in Chapter 3 . As in that experiment, two litres of urine solution was added to the experimental plots. Thus, the quantities of constituents used to make the urine solution in the Waipawa experiment were changed to give the new urine N rates in the current experiment (Table S . lA). Two litres of 0.7 or 1 .4 % N urine solutions were added evenly using a small watering can to relevant plots at a rate which avoided surface runo ff. As discussed in Chapter 3 , it was decided that control plots should receive nothing- as opposed to 2 litres of water containing no N. 5.2.4 Soil sampling At 4, 1 2, 27, 45, 66 and 8 8 DAUA soil samples were taken from the 0- 1 00, 1 00-200 and 200-300 mm depths of each plot. Soil sampling was done as described in Chapter 3 . 5.2.5 Ammonia volatilisation NH3 samplers were placed and analysed as described in Chapter 3 . 5.2.6 Soil mineral and mineralisable nitrogen S oil NH/-N and N03--N were extracted with 2 M KCI and analysed as described in Chapter 3 . Soil samples from 3 DAUA were extracted with O.SM K2 S04 rather than 2 M KCI to use the extractions for analysis of mineral nitrogen, dissolved organic C and dissolved organic N. S oil samples from 4, 1 2 and 2 7 DAUA were analysed for mineralisable nitrogen after anaerobic incubation as described in Chapter 3 . 5.2.7 Dissolved organic carbon and dissolved organic nitrogen 1 22 At 3 DAUA, 4 g of soil was extracted using 30 mL of 0 . 5 M K2S04. The extracts were used to measure dissolved organic carbon (Sparling and West, 1 9 88) and total soluble N (Cabrera and Beare, 1 993). Mineral N content was deducted from the total soluble N to calculate the dissolved organic N . 5.2.8 Resin-adsorbed nitrogen Resin spikes were constructed as described in Chapter 4. Before commencement of the experiment, the reproducibility of the resin spikes was checked using an ammonium nitrate solution containing 1 0 Ilg NH/-N/mL and 1 0 Ilg N03--N/mL . It was found in Chapter 4 that shorter burial periods of resin spikes in soil would be more suitable for moist soil and for soils containing high amounts of N . As the current experiment was carried out in mid winter with moist soil conditions and high N concentrations due to urine addition, it was decided to keep the resin spikes in the soil for a short period of only 3 days. After urine solution was added to the soil, three KCl-saturated resin spikes were buried in random locations within the undisturbed section of each plot. After 3 days, the resin spikes were removed from the soil, washed with deionised water, shaken with 25 mL of 2 M KCI and the extracts were analysed for resin-adsorbed NH/-N and N03--N using a Technicon Auto Analyser as described in Chapter 4. When the three spikes were removed from the soil another three KCl-saturated resin spikes were buried for the next 3 days. The subsequent resin spikes were buried at different locations within the undisturbed section of each plot. This routine was continued until the first harvest at 3 3 DAUA. Thereafter resin spikes were buried for 3 days in every week until the last pasture harvest. 5.2.9 Plant dry matter production and pasture nitrogen uptake. Four pasture harvests were taken 3 3 , 66, 97 and 1 44 DAUA. At each harvest, pasture dry matter production and plant N uptake were measured as described in Chapter 3 . 1 23 5.2 . 10 Statistical analysis Analysis of variance was carried out using SAS for Windows verSIOn 8. Resin­ adsorbed NH/-N and N03--N, 2 M KCI -extractable NH/-N and N03--N in 0- 1 0 cm soil depth and total mineral N in 0-3 0 cm depth data were analysed using a repeated measures model in a mixed procedure (Littell et al., 1 998) to examine and compare response trends over time. Autoregressive co-variance s tructure was used in the model. Log transformed data were used in the statistical analysis. Log-transformed means were compared by a Least Mean Square method at a 5% significance level and based on this, the significance of observed means was expressed. Analysis of variance for NH3 volatilisation, herbage DM, and herbage N uptake data were analysed using a GLM procedure. Mean comparisons were done using the Duncan multiple range test at 5% significance. Log-transformed data were used in NH3 volatilisation statistical analysis. 5.3 Results 5.3.1 Climate The mean air temperatures and mean rainfall (mm/month) during the period of 1 970- 1 995 and mean wind speed krnJh for 1 994 at Ballantrae AgResearch hill country research station are presented in Table 5 . 2 . Average annual rainfall received at the experimental site for the last 25 years was 1 1 88 mm/year. The 2 5 -year average rainfall for the July to October period was 42 1 mm. Rainfall, air temperature and soil temperature data during the experimental period are presented in Fig. 5 . 1 and Fig. 5 . 2 . The total amount of rainfall received during the experimental period ( 1 417/2000 to 1 91 1 0/2000) was 3 1 6 mm. 124 Table 5 .2 Average meteorological conditions for 1 970- 1 995 at Ballantrae AgResearch Hill country research station (wind speed data was only available for 1 994). Month Mean air temp Max. air Min air temp Rainfall Wind speed (QC) temp (QC) (Q C) (mm) (km/h) Jan 16.6 1 8 .4 1 4.7 78 .8 1 5 .8 Feb 16.6 1 9.0 14 .4 82.5 1 2 . 1 Mar 15.5 1 7 .9 1 3 1 00.3 1 4.6 Apr 12.8 1 5 .2 1 0 . 1 90.9 14 .7 May 10.4 1 2 .0 7.9 1 03 .2 1 8 .6 JUll 8.6 10.7 8 .0 1 1 0.2 1 3 . 8 Jui 7.9 9.4 6.0 1 08 . 1 1 6 .3 Aug 8.5 1 0.9 7 .2 97.7 1 7 . 1 Sep 10.0 1 1 .8 8 .2 1 08. 1 1 8 .8 Oct 1 1 .5 1 2.6 9.9 1 06.0 1 5 . 1 Nov 13 .3 1 5.4 1 1 . 1 98.3 22.3 Dec 1 5 . 1 1 6 .8 1 3 .2 1 04.6 1 6 .8 30 25 :0. 20 1'1 C - iij .... 15 c iij Cl: E E 10 5 0 l� n� , n �n I� Inn n �.J�� dn nn In n 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 Days after urine application Fig. 5 . 1 Rainfall during the experimental period ( 1 417/2000- 1 9/ 1 0/2000). 125 20 1 8 1 6 1 4 U . 1 2 GI .. :s .. IV 10 Qj Q. E GI I- 6 -- Avg air 4 Temperature 2 Avg. Soil Temperature o 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Days after urine application Fig. 5 .2 Air and soil temperature (0- 1 0 cm) during the experimental period ( 1 4/7/2000- 1 91 1 0/2000). 5.3.2 Statistical interpretation As pointed out in Section 3 .4.2, the repeated measures model in a mixed procedure made it possible to compare the treatment means at a particular sampling time as well as treatment means between sampling times within a site. As in the Waipawa experiment described in Chapter 3, no comparisons were made between sites, as the site locations were not replicated appropriately. The least squares mean procedure for mean comparison supplied mean comparisons for all possible comparisons. This caused some difficulties in tabulating the statistical data. This detailed information is tabulated in Appendix 2. When necessary, any important statistical significant differences within and between samplings are identified and discussed in the text. 5.3.3 Mineral nitrogen Mineral N compnses NH/-N and N03--N. These two components are discussed separately in Sections 5 .3 .4 and 5 .3 .5 . This section provides some general comments on the combined quantities ofNH/-N and N03 --N extractable by 2 M KCI in the soils. 800 700 600 Z "i 500 ... !Cl GI .I: .: Z 400 � en JI: ·0 300 Cl) 200 1 00 - - - -0 _ _ - - - • ·0 . 0 - - - 0 _ .. -'-FO ___ F280 ___ F560 - t:J 'SO - 0 '5280 - 0 '5560 o 5 10 1 5 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 Days after urine application Fig. 5 .3 Effect of urine application on soil (0-30 cm) mineral N levels. 126 Urine application resulted in highly significant (Table 2 . 1 of Appendix 2) increases in . the levels of 2 M KCI -extractable mineral N in soil, to a depth of 30 cm, at both the flat site and steep site (Fig. 5 .3) . These increases in mineral N were greater at the flat site than the steep site and, not surprisingly, were greater when urine was applied at 560 kg Nlha than when it was applied at 280 kg N/ha. At the steep site, soil (0-30 cm) mineral N levels in the urine-treated plots tended to be close to their maxima 3 DAUA, and declined thereafter towards the control levels. In the S280 treatment, mineral N levels were not statistically different from the control values from 45 DAUA onwards. In the S560 treatment, mineral N levels were still significantly higher than the control at 66 DAUA, but not at 88 DAUA. In contrast to the steep site, soi l (0-30 cm) mineral N levels in the urine-treated plots at the flat site tended to increase further, beyond 3 DAUA. In the F560 treatment, this increase in mineral N levels continued until 27 DAUA, and this increase over time was statistically significant (Table 2 . 1 of Appendix 2). In the F280 treatment, the increase 1 27 continued only until 12 DAUA and was not statistically significant. In both treatments, mineral N levels then decreased towards the control values. Table 5 . 3 Net mineral N (extracted by 2 M KCI ) in the soil profile (0-30 cm) (i .e. treatment minus control values). Soil Depth 0- 10cm 1 0-20cm 20-30cm % Treatment Day NH4+-N NO)·-N NH/-N O)·-N N�+- O)·-N recovery in soi l kg N Iha F280 3 159 1 1 24 1 0 4 1 75 12 277 56 1 3 30 0 1 8 1 4 1 27 106 23 3 6 2 3 5 1 45 25 37 1 1 35 1 1 8 45 66 0 0 1 0 1 3 2 88 6 0 1 0 0 0 2 F560 3 208 29 1 5 1 2 7 4 49 1 2 3 1 6 7 5 60 49 0 1 2 9 1 27 334 1 53 5 1 76 2 1 7 1 1 3 45 9 1 1 57 29 93 9 38 74 66 2 5 7 24 5 38 1 5 88 0 0 0 0 0 0 0 S280 3 1 1 8 8 0 0 8 0 48 12 88 1 7 20 8 1 1 2 52 27 26 22 5 8 0 4 23 45 5 1 5 3 9 1 2 1 2 66 0 0 0 2 1 3 2 88 0 0 0 0 0 0 0 S560 3 279 14 1 8 9 4 4 59 1 2 259 44 20 26 2 7 64 27 209 76 1 2 29 0 7 60 45 77 97 6 5 1 2 2 1 45 66 10 10 8 47 5 38 21 88 1 0 1 0 0 0 0 128 In both the F280 and F560 treatments, mineral N levels were still significantly higher than control at 45 DAUA, but not at 66 DAUA. When soil (0-30 cm) mineral N levels were at their peak at the flat site, the differences in the quantities of mineral N between the treated and control plots were greater than the total amount of urine N added (Table 5 .3). In other words, the apparent recovery of urine N in the soil mineral form alone was greater than 1 00%. This was the case for both rates of urine addition, and suggests a possible priming effect. Although at the first sampling (3 DAUA), most of the mineral N resulting from urine addition was in the 0- 10 cm soil depth, there was clear evidence of some movement of urine N to the 1 0-20 cm and 20-30 cm soil depths (Table 5 .3) . As there had been no rainfall over this time (Fig. 5 . 1 ), this movement probably resulted from preferential flow through soil macro-pores during urine addition. At later samplings, greater quantities of urine N were detected in the 1 0-30 cm soil depths, presumably as a result of leaching. This is discussed in more detail in Section 5 .3 . 1 0 . 5.3.4 Ammonium Concentrations of 2 M KCI -extractable NH/-N (flg NH/-N/g soil in 0- 1 0 cm depth) and resin-adsorbed NH/-N (flg NH/-N/5 cm2 resinl3 days) over the duration of the trial are presented in Fig. 5 .4. The patterns for 2 M KCI -extractable NH/-N in the 0- 10 cm soil depth, are similar to those for soil mineral N, to a depth of 30 cm, (discussed in the previous section), particularly shortly after urine application. This was because at that time, NH4 + -N was the dominant component of mineral N in all the urine-added treatments (Fig. 5 .5), and most of the added nitrogen was in the 0- 1 0 cm soil depth. Thus, unne application resulted in highly significant (Table 2.2 of Appendix 2) increases in the levels of 2 M KCI -extractable NH/-N, to a depth of 1 0 cm, at both the flat site and steep site (Fig. 5 .4). As for mineral N, these increases in NH/-N were larger at the flat site than the steep site, and larger for the 560 kg/N/ha than for the 280 kg N/ha rate of urine application. 900,---------------------------, Flat site ---.-FO-2M KCI __ F280-2M KCI __ F560-2M KCI - t:>- FO-Resin - G- F280-Resin - 0- F560-Resln o 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 900,------------------------. Steep site ---.-SO-2M KCI __ S280-2M KCI __ S560-2M KCI - t:>- SO-Resin - G- S280-Resin - O· S560·Resin . • :: ; " .. g • ; . ;fl ; ; =8 o 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Days after urine application 129 Fig. 5 .4 Effect of urine application on soil NH/-N levels (0- 1 0 cm) as detennined by 2 M KCl extraction and resin adsorption methods. As was the case for mineral N, there was an apparent, continuing increase in 2 M KCI - extractable NH/-N after 3 DAUA, for both rates of urine application at the flat site. However, for NH/-N in the 0- 1 0 cm soil depth, these increases after 3 DAUA were not statistical ly significant-although for the F560 treatment the increase between 3 and 27 DAUA nearly reached statistical significance (P=0.08). As the trial progressed, the levels of NH/-N in the treated plots decreased towards those in the control plots. At both the flat site and steep site, 2 M KCI -extractable NH/ -N (0- 1 0 cm) levels in the 280 kg urine Nlha treatments were significantly higher 1 30 than in the control plots, up to 27 DAUA. In the 560 kg urine Nlha treatments, the significantly elevated levels of 2 M KCI -extractable NH/ -N (0- 1 0 cm) persisted until 45 DAUA. FO I--NH41 --<>-N03 40"'--------------, RI 30 .:: � 20 z 1 0 500 400 RI == 300 Cl "'" 200 z 1 00 0 500 400 RI == 300 Cl "'" 200 z 1 00 0 1 0 20 30 40 50 60 70 80 90 DAUA F280 o 10 20 30 40 50 60 70 80 90 DAUA F560 o 10 20 30 40 50 60 70 80 90 DAUA I--NH4 1 --<>-N03 so 40,----------------. RI 30 .:: � 20 z 1 0 1 0 20 30 40 5 0 60 7 0 8 0 90 DAUA 5280 I--NH41 --<>-N03 500 ,----------------, 400 RI == 300 Cl � 200 � ........ 1 00 I _ "'" o ����i"OF==;=:d o 1 0 20 30 40 50 60 70 80 90 DAUA 5560 500 ,----------------, 400 RI ::!: 300 Cl "'" 200 z 1 00 O �������� o 1 0 20 30 40 50 60 70 80 90 DAUA Fig. 5 . 5 Quantities of 2 M KCI extractable NH/-N and N03--N in the 0-30 cm soil depth during the experimental period. Note. change in scale between control and treated plots . Resin-adsorbed NH/-N followed similar trends to the 2 M KCI -extractable NH/-N in the 0- 1 0 cm soil depth, but there were some large variations between sampling times (Fig. 5 .4). Some of these variations were statistically significant, and there were some consistent patterns between treatments-providing further evidence that the variations 1 3 1 with time were not sole ly due to chance. For example, from 3 to 6 DAUA all urine­ added treatments showed a drop in resin-adsorbed NH/-N, which was then recovered by 9 DAUA. This could be due to high volatilisation losses in first few days and later recovery from a priming effect. One theoretical explanation of the variation in resin extractable NH/-N with time relates to the observation that NH/-N adsorption by resin spikes in the soil can vary with soil moisture content, as the NH/-N adsorption to resin is governed by diffusion. This was demonstrated in Section 4.6.4.2 in Chapter 4. In this instance however, the observed variations of resin-ads or bed NH/-N levels with time were not likely to be due to soil moisture differences because the soil moisture contents were usually close to field capacity due to the high rainfall (Fig. 5 .2) that occurred during the experiment. 5.3.5 Nitrate Concentrations of 2 M KCI -extractable N03- -N ()lg N03- -N/g soil in the 0- 1 0 cm depth) and resin-adsorbed N03--N ()lg N03- -N/5 cm2 resin 13 days) over the duration of the trial are presented in Fig. 5 .6 . Although urine application resulted in highly significant increases in 2 M KCI -extractable N03--N (Table 2.3 of Appendix 2), the pattern of increase was very different to that for NH/ -N, described in Section 5 .3 .4. At the first sampling (3 DAUA), there were no significant differences in 2 M KCI - extractable N03--N between the control and urine-treated plots at either site. Soil N03·­ N levels then built up with time in the urine-treated plots at both sites. These elevated levels of 2 M KCI - extractable N03--N were apparent throughout the 0-30 cm soil depth (Table 5 .3) . Presumably, N03--N was produced in the 0- 1 0 cm depth by nitrification of added urine N, and was then transported to the deeper soil depths by leaching. These processes are discussed in more detail in later sections. The 2 M KCl - extractable N03--N levels at both sites in the 0- 10 cm soil depth were at their maximum between 35-45 DAUA (Fig. 5 .6). Thereafter, N03--N levels in the surface soil decreased markedly, to reach the levels in the control plots by 66 DAUA. The elevated N03--N levels in the 1 0-20 and 20-30 cm soil depths persisted for 1 32 somewhat longer (Table 5 .3), but by 88 DAUA there was no difference in N03--N levels between the control and treated plots in any of the 3 soil depths at both sites. '0 VI .. � '" Cl .. ::I. " E '" ;;- u � C> l!2 e. Cl .: '" '" 0 0 z Z .. " :;; " .. LJ U <; :! VI ;; " 'l' ":' c: 13 Vi :.: .. � a:: ... 900 800 700 600 500 400 300 200 '00 .P . 9'0 o :. : ', : Flat site -.-FO-2M KCI __ F280-2M KCI __ F 560kg-2M KCI .. -6 . . FO-Resin . . O· . F280-Resin . . 0 · · F560-Resin 5 '0 '5 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 '00 900,---------------------------, o 0 : ' , . ' " : ", Cl '. . '. a"O . . O : ,..0 .. 0 ... : ·· .. : () . . . 0 Steep site :,0 -.-SO-2M KCI __ S280-2M KCI __ S560-2M KCI . . -6 . . SO-Resin . . 0· · S280-Resin . . 0· · S560-Resin -0 5 ' 0 '5 20 2 5 30 3 5 40 45 50 55 60 6 5 7 0 7 5 80 8 5 90 9 5 '00 Days after urine application Fig. 5 .6 Effect of urine application on soil N03--N levels (0- 1 0 cm) as determined by 2 M KCl -extraction and resin-absorption methods_ The quantities of 2 M KCl -extractable N03--N (0- 1 0 cm) appeared to be higher on the flat site than the steep site, but the trial was not designed to test these differences statistically. Resin-adsorbed N03--N fol lowed a similar pattern to 2 M KCl -extractable N03--N. However, N03--N extracted by the resin showed a much greater range than that by 2 M 1 33 KCI for both rates of urine addition at both sites. This is discussed in more detail in Section 5 .3 .6 . Although after 66 DAUA, no significant differences in 2 M KCI -extractable N03--N were observed between treated and control plots, there were still some statistically significant differences (Table 2 .5 of Appendix 2) observed in resin-adsorbed N03--N levels. This is also discussed in more detail in Section 5 .3 .6 . It was interesting to note that when the 2 M KCI -extractable N03--N levels were at their maximum at both sites, there was a more than three-fold difference in the amounts of N03--N between the 560 kg urine N/ha and the 280 kg Nlha urine treatments. In contrast, levels of resin-adsorbed N03--N were more in proportion to the two-fold difference in application rates . 5.3.6 Performance of resin spikes It is difficult to compare directly resin adsorbed N and N extracted by 2M KCI , as resin spikes do not measure from the same volume of soil as the 2 M KCI extraction method. Also, these two methods are measuring different properties. Data from the 2 M KCI extraction method report the pool size of available N at a particular time while resin data may reflect the N flux over time as well as the available N pool size at a given moment. A possible example of this is given below. After about 66 days, 2 M KCI extraction could detect very l ittle NH/-N or N03--N, but the resin method could stil l detect sufficient N03--N to show statistically significant treatment differences (Table 2 .5 of Appendix 2). To i llustrate this, Fig. 5 . 7 shows the data from 55 to 97 DAUA in Fig. 5 .6, redrawn with an expanded scale. At the steep site, the amounts of resin-adsorbed N03--N in the S560 plots were significantly (P a. 50 ... 0 55 55 60 65 70 75 80 85 90 95 100 60 Days after urine application 1-'- SO --- S280 --S560 1 Flat Site 65 70 75 80 85 90 95 100 Days after urine application 1-'- FO --- F280 --F560 1 Fig. 5 .7 Effect of urine treatments on resin-adsorbed N03--N at 5 5 to 97 DAUA. 5.3.6.1 Relationships between 2 M KCI -extractable N and resin-adsorbed N 1 35 Developing relationships between resin-adsorbed N and other soil parameters is complicated by the high variability in the resin-adsorbed N. For example, at 27 DAUA, resin-adsorbed NH/-N assessed by the twelve individual spikes in each of SO, S280 and S560 treatments varied from 0-34, 1 7- 1 43, and 54-25 1 )lg/5 cm2/3 days respectively, with corresponding co-efficients of variation of 88, 6 1 , and 55%. At the same sampling, N03--N varied among the twelve spikes from 5-34, 14-326, and 60-804 )lg/5 cm2/3 days, with co-efficients of variation of 60, 84, and 72% for the SO, S280 and S560 treatments respectively. As was indicated in Chapter 4, this variation reflects actual spatial variability of mineral N concentrations in the field, rather than errors within the technique itself. Resin­ adsorbed values are equivalent to analysing individual soil cores, whereas 2 M KCI - extractable NH/-N and N03--N are from bulked samples of five soil cores. In an attempt to overcome some of this variability the relationships between resin­ adsorbed N and 2 M KCI -extractable N illustrated in Fig. 5 . 8 were developed by comparing mean resin-adsorbed N values in each plot (average of 3 resin spikes in each plot) and 2 M KCI -extractable N (plot value obtained from a bulk sample of 5 soil cores) values in each p lot. Data were obtained from both sites at 3, 12, 27 and 45 DAUA. Thus, a total of 96 data points (3 treatments x 4 replicates x 4 sampling times x 2 sites) from each method were used for the relationships. For both N03--N and NH/-N, there was some evidence of linear relationships between resin-adsorbed and 2 M KCl -extractable N values. Although, the relationships were statistically significant they were weak. As the resin method and 2 M KCI methods are measuring different things (see Section 5 .3 .6), exact 1 : 1 relationships between resin-adsorbed N and 2 M KCl -extractable N are unlikely. Resin spikes tend to adsorb more N03--N than NH/-N at a particular N concentration as measured by the 2 M KCI extraction method (Fig. 5 . 8) . This was also observed in 1 36 Chapter 4 when resin spikes adsorbed higher amounts of N03--N th::m NH/-N from a NlL;N03 solution containing 1 0 Ilg N03--N/mL and 1 0 Ilg NH/-N/mL . To further minimise the variability observed in the resin-adsorbed N data, the linear relationships were re-developed (Fig. 5 .9) using the average of each treatment (average of four replicates) for both resin and 2 M KCI extraction methods. Relationships were also developed separately for the flat and steep sites. Thus, a total of 12 data points (3 treatments x 4 sampling times x 1 site) for each method were used for the relationship. 1200,-------------------------------------------. 1100 1000 900 Z u; 'tI >- 800 Cl C'iI of � 700 � NE 600 C'iI U c !!2 500 � � 400 a::: -- 300 200 o • • ".Y = 2.44x R2 = 0.21 " y = 1 .0 1 x R2 = 0 . 3 3 o 50 100 1 50 200 250 300 350 400 450 500 550 600 650 700 750 800 2M KCI-extractable N (kg N/ha) 0. Ammonium • Nitrate -Regression line-Nitrate -Regression line-Ammonium Fig. 5 . 8 Relationship between resin-adsorbed N and 2 M KCI extractable N when both steep and flat site data are used. Both the N03--N and NlL;+-N data then produced strong linear relationships (Fig. 5 .9). The resin spikes tended to adsorb higher quantities of N03--N at the steep site than at the flat site at a particular 2 M KCI -extractable N03--N concentration (Fig. 5 . 9A). Higher N03--N adsorption to resin can occur at higher soil moisture contents (see Section 4.6 .3 .2 in Chapter 4). However, it is not clear, why the wetter flat site had less resin-adsorbed N03--N than the steep site for a given value of 2 M KCI extractable N03--N. 1 37 In contrast, resin-adsorbed NH/-N was similar at both sites at a particular 2 M KCI - extractable N� + -N concentration. A) z . ... 0 Z ." I» .c � 0 I/) ." III . c 'iij I» a: B) � + -.r l: Z ." I» .c � 0 I/) ." III . c I/) I» a: 600 I/) 500 >-III ." 400 C") N- 300 E U 200 It') Cl � 100 0 0 y (Steep) = 4.59,p y (Flat) = 2.45x R2 = 0.68 50 100 1 50 200 250 300 350 400 2M KCI extractable N03-·N (kg/ha) o Flat • Steep -- Regression line-Flat - Regression line-Steep 600 I/) 500 >-III ." 400 � 300 N E u It) 200 Cl � 100 0 0 50 0 y (Flat) = 1 .23x 0 • 0 • y (Steep) = 1 .20x R2 = 0.74 1 00 1 50 200 250 300 2M KCI extractable NH4+·N (kg/ha) 350 400 o Flat • Steep -- Regression line-Flat - Regression line -Steep Fig. 5 .9 Relationships between (A) resin adsorbed N03--N and 2 M KCI -extractable N03--N. (B) Resin-adsorbed NH/-N and 2 M KCI adsorbed NH/-N. 5.3.7 Mineralisable N It was observed in Section 5 .3 .3 , that the apparent recovery of urine N in the mineral form was greater than 1 00% up to 27 and 66 DAUA in the F280 and F560 treatments 1 38 respectively, suggesting a priming effect after urine application. To investigate this, an attempt was made to see if there were any changes in soil mineralisable N during the period of the observed priming effect. Soil samples from 3, 1 2 and 27 DAUA, which had been stored at 4° C at field moisture content, were analysed for mineralisable N using the anaerobic incubation procedure described in Section 5 .2 . 5 . These incubations and analyses were conducted after the completion of the field trial . In the control soils at both the sites, the highest mineralisable N levels were found in the 0- 1 0 cm soil depth (Table 5 .4). Also at both sites, mineralisable N in the 1 0-20 cm and 20-30 cm soil depths was not affected by urine treatments. Therefore, only changes in mineralisable N in the 0- 10 cm soil depth are discussed (Fig. 5 . 1 0 and 5 . 1 1 ) . Higher mineralisable N levels were observed in the flat site control plots than in the steep site control plots (Table 5 .4). Table 5.4 Mineralisable N levels (kg/ha) in control soils at different depths . Values are the average of soils from sampling times at 3, 1 2 and 27 DAUA. Soil Depth Steep site Flat site 0- 1 0 cm 200 30 1 1 0-20 cm 1 1 5 1 3 1 20-30 cm 84 92 At both sites, mineralisable N levels in the control plots did not change significantly (P< 0.05) between any of the three sampling times. Although an increase in mineralisable N might normally be expected after unne application, the soil mineralisable N levels were significantly reduced compared to control . This decrease was first observed with the lower urine N application rate of 280 kg urine N/ha treatment at 3 DAUA at both sites. The mineralisable N in the S560 treatment was significantly lower than the control at 12 and 27 DAUA, while the F560 treatment had significantly lower mineralisable N values than the control plots at 27 DAUA only. 1 3 9 Mineralisable N in the F560, S280 and S560 treatments also decreased significantly with time from 3 to 27 DAUA. 400 350 Ii 300 � ;;, :!. 250 Z Cl :;; 200 " � 'i 150 � " :E 100 50 o 400 350 .. 300 � .. � 250 Z � 200 .. .!! 'i 1 50 � " i 100 so so 5280 5560 3 Days FO FZ80 F560 3 Days Steep site so 5280 5560 12 Days Flat site FO F280 F560 12 Days so 5280 5560 27 Days FO F280 F560 27 0ays Fig. 5 . 1 0 Effects of urine treatments on soil (0- 1 0 cm) mineralisable N levels at 3, 1 2, 27 days after urine application. Treatments with common upper case letters do not differ at P<0 .05 level within a sampling day. Treatments with common lower case letters do not differ at P<0.05 level between sampling days of the same treatment. When the levels of mineralisable N at the each site are plotted on the same graph as the level of mineral N (Fig. 5 . 1 1 ), there is an apparent relationship in some treatments. so 600 j I ca 500 :E 400 Cl 300 = ��� +-1 : ...,. , --: :-4 : : : , = : � : . ... : -......! o 5 1 0 1 5 20 25 30 DAUA 1-.-Mineral N -<>--Mineralisable N 1 5280 600 j ca 500 .r. 400 0, 300 .:.t. 200 -0 Z 1 0� 0:=:---,---;. ......:.. o 5 1 0 1 5 20 25 30 DAUA 1-.- Mineral N -<>--Mineralisable N I 5560 600 j I ca 500 .r. 400 n�! +-�S������-:�-,�--=�-.,...=�--=�-=-.. :� o 5 1 0 1 5 20 25 30 DAUA 1-.- Mineral N -<>--Mineralisable N I FO ca 500 .r. 400 600 J � ��! +-�:�,��.�::�:��:::::, .:� o 5 1 0 1 5 20 25 30 DAUA I-+-Mineral N -<>--Mineralisable N I 600 j ca 500 .r. 400 0, 300 .:.t. 200 Z 100 o F280 o 5 10 1 5 20 25 30 DAUA I-+-Mineral N -<>-- Mineralisable N I F560 600 1 ca 500 .r. 400 0, 300 .:.t. 200 Z 1 00 o o 5 1 0 1 5 20 25 30 DAUA I-+-Mineral N -<>--Mineralisable N I 140 Fig. 5 . 1 1 Effect of urine application on soil mineralisable and mineral N levels (0- 1 0 cm depth) at 3 , 1 2 and 27 days after urine application (DAUA). At the flat site in particular, it appears that increases in mineral N from 3 DAUA are closely paralleled by decreases in mineralisable N. This relationship between increases 1 4 1 i n mineral N and decreases in mineralisable N from 3 DAUA to 12 DAUA and from 1 2 DAUA to 2 7 DAUA across all sites is displayed in Fig. 5 . l 2 . It is apparent that a linear relationship (R2=0.6) exists with a slope of 1 . 1 . This close coincidence with the 1 : 1 line, suggests that the observed mineral N increase was possibly due to mineralisation of soil organic matter. When the relationships were redrawn separately for the two sites (data not presented) it was apparent that the linear relationship was much stronger on the flat site (R2=0. 8) than on the steep site (R2=0.3). QI III I'CI QI .. U .s o - 150 -200 y = 1 . 1x - 1 9. 1 R2 = 0.6 Decrease in mineralisable kg N/ha 2 0 Fig. 5 . 1 2 Relationship between increase in soil mineral N and decrease In soil mineralisable N from 3 to 1 2 and from 12 to 27 DAUA. 5.3.8 Pasture response Pasture yield cuts and associated herbage N analyses were conducted on four occasions covering the period 0- 1 44 DAUA ( 1 4/7/2000 to 71 1 2/2000). Soil measurements had ceased at 88 DAUA - corresponding to cut 3 - when no further residual effects of urine addition on soil N status could be detected. However one further pasture cut was taken 1 44 DAUA to check whether there was any observable residual effect of urine addition on either pasture growth rate or N concentration. 142 Dry matter (DM) accumulation during the experiment is il lustrated in F ig . 5 . 1 3 . When all the cuts are combined, over the whole experimental period, urine treatments s ignificantly increased pasture accumulation only at the flat site. When the cuts were considered individually (Fig. 5 . 1 3), it was apparent that most of the pasture response at the flat s ite occurred at the second and third cuts. And at the steep s ite, where there was no growth response overall, there was some evidence of a s ignificant DM response to urine application at the second harvest. 6 0 0 0 c 0 .. 5 0 0 0 ! " E 4 0 0 0 b " - u • u !!: 3 0 0 0 • III • � .. " 2 0 0 0 .. " • Go 1 0 0 0 :lE a 0 F O F 2 8 0 F 5 6 0 5 0 5 2 8 0 5 5 6 0 . C u t 1 ( 3 3 D a y s - 1 4 17 10 0 - 1 6 18 10 0 ) D C u t 2 ( 3 3 D a y 5 - 1 6 18 10 0 - 1 8 19 10 0 ) o C u t 3 (3 1 D a y 5 - 1 8 19 10 0 - 1 9 11 0 10 0 ) 0 C u t 4 ( 4 7 D a y s - 1 9 11 0 10 0 - 7 11 2 10 0 ) F ig. 5 . 1 3 Effect of urine treatments on pasture OM accumulat ion at the flat and steep s ites. Dry matter yields with common letters between treatments within same cut and same site do not differ at P 0 Nn (I) = Nn (I- I ) + Ln-I (I- I ) - Ln (I- I ) (5 .3) Final leaching was estimated by summing the daily leaching from the 20-30 cm layer during the experimental period. This was called Model 1 400 :i 350 la !. z 300 d' Z 250 • :a � 200 i , 1 50 ij � lE 1 00 ... 50 o 1 49 o 5 1 0 15 20 25 30 35 40 45 50 55 60 65 70 75 110 85 90 95 100 105 DAUA F ig. 5 . 1 6 Estimated and measured 2 M KCl -extractable N03--N in the 0- 1 0 cm soil depth during the experimental period. Arrows indicate measured 2 M KCI -extractable N03- N. Other marked data points ( e ) are estimated 2 M KCI -extractable N03--N values from the relationship with resin adsorbed N03--N illustrated in F ig . 5 . 8 . Table 5 .8 Example of leaching calculation by the model. Data extracted from leaching model of the F560 treatment. Day Drainage NI LI N , L, N3 L3 (mm) (g/m') (g/ mC) (g/mC) (g/ m') (g/m') (g/m') 6 0 5 0 3 0 2 0 7 1 0 5 5 x ( 10/-'8) 3 3 x (J 0'-'8 ) 2 2x (U)/-'8 ) = 1 .0 =0.7 =0.4 8 3 7 7x (3/-'8) (3 + 1- n.7) 3 x (3'-'8) (2+0.7- ClA) 2x (3'-'8) =0.4 = 3 .3 = 0.2 =23 = 0 . 1 3 As noted above, in Model 1 , measured values of 2 M KCI -extractable N03--N in the 1 0-20 and 20-30 cm soil depth at 3 DAUA were used to initialise the model and from then on daily N03--N concentrations in the 1 0-20 and 20-30 cm soil depths were calculated from inputs and outputs to each depth by leaching. I n fact, however, 2 M KCI -extractable N03--N was measured at these depths on several occasions throughout 1 50 the trial. This therefore gave an opportunity to check the predictions of Model 1 against measured values. First, Model 1 was run as described above using the inputs of daily N03--N levels in 0- 1 0 cm over the whole experimental period and soil N03--N levels in 10-20 and 20-30 cm soil depths only at 3 DAUA. Then the predicted values of the N03--N concentrations in the 10-20 and 20-30 cm soil depths at 3 , 1 2, 27, 45, 66 and 88 DAUA were replaced in the model by the measured 2 M KCI -extractable N03--N values on these days. Then the model was rerun. This second version of the model was termed Model 2. Finally, the outputs of the two models were compared. 5.3 .10.2 Leaching model output The simple model described in Section 5 .3 . 1 0 . l suggested that substantial leaching of urine N as N03--N during the experiment was likely. Predicted leaching during the experimental period using Model 1 (Table 5 .9) was large in all treatments. The predicted leaching losses from urine-treated plots ranged from 1 30 to 320 kg !ha, which is 38-48% of added urine N. In addition, the predicted N leaching from the flat site controls was 50 kg N!ha. These predicted loses were large and when the predicted soil N03·-N concentrations in the 1 0-20 and 20-30 cm depths at 1 2, 27, 45, 66 and 88 DAUA were compared with the measured values of 2 M KCI -extractable N03--N (Fig. 5 . 1 7), it was apparent that Model 1 was overestimating the quantities of N03--N at depth-particularly towards the end of the trial. This might then have led to overestimation of leaching levels by Model 1 (Table 5 . 9 and Fig. 5 . l 8) The overestimation of N03-- levels at depth by the model probably resulted from the assumption that changes in the N03--N concentrations in the 1 0-20 and 20-30 cm depth intervals could only occur through leaching. In reality, N03--N could also have been lost from these depths by plant uptake, denitrification and immobilisation. 1 5 1 As described in section 5 .3 . 1 0. 1 , Model 2 made use of measured values of soil N03--N at the 0- 1 0, 1 0-20, 20-30 cm soil depths where data were available. When the Model 2 was run the predicted leaching was markedly decreased (Table 5 .9). Table 5 . 9 Estimated leaching losses from urine treatments by the two models. Leaching % of added urine Leaching % of added urine Treatment kg N03·-N /ha (Treatment-control)/added N) kg N03--N /ha (Treatment-control) ladded N) Model I Model 2 FO 5 1 - 22 - F280 1 57 38 83 2 1 F560 320 48 1 88 30 SO 25 - 22 - S280 1 37 40 48 09 S560 267 43 207 33 Thus, in Model 2, updating of the N03--N concentration by measured values at several times, may account for the influence of some of the other N transformations noted above. Because of this, Model 2 outputs are likely to be the c loser to the actual amount of leaching. Although the model presented here makes many simplifying assumptions, it is clear that the potential for nitrate leaching is very high in these hill pasture urine patches . The predicted leaching losses from the urine treated plots by Model 2 ranged from 48 to 207 kg Nlha which is 9 to 33% of added urine N. 1 52 F280 (10-20 cm depth) F280 (20-30 cm depth) 25 25 ! 20 � 1 5 --Model 2 - Model 1 --Model 2 N ! 20 - Model 1 � 1 5 � 10 � 10 0 5 Z 0 0 5 Z .......--. 0 I-'" -- l M N .., M .., .., .., M M M M ..... ..... N M 'It III '" t- '" DAUA DAUA F560 (10-20 cm depth) F560 (20-30 cm depth) --Model 2 25 - Model 1 25 --Model 2 N E 20 § 1 5 ! 20 � 15 -Model 1 � 10 � 10 0 Z 5 0 5 Z 0 0 M N ..... 0 Cl> Cl) t-..... N .., .., .. III M N ..... ..... N DAUA DAUA 5280 (1 0-20 cm depth) 5280 (20-30 cm depth) --Model 2 25 -Model 1 25 20 --Model 2 ! 20 ! - Model 1 � 1 5 � 15 Z 10 � 10 0 0 Z 5 Z 5 ./V '\r-"" "\. :).. ../" '---0 0 M N ..... 0 Cl '" t- '" 1/1 .. M N ..... 0 Cl Cl) t- '" 1/1 Cl) ..... N M M .. 1/1 '" t- '" ..... N .., M .. 1/1 '" t- DAUA DAUA 5560 (10-20 cm depth) 5560 (20-30 cm depth) --Model 2 --ModeI2 25 25 -Model 1 -Model 1 N E 20 NE 20 § 1 5 § 1 5 � 10 � 10 0 0 Z 5 Z 5 r-v- '"1 0 0 M N ..... 0 Cl> Cl) t- '" 1/1 'It M N ..... 0 Cl Cl) t- '" III '" ..... N .., M .. 1/1 '" t- '" ..... N .., .., .. III '" t- DAUA DAUA Fig. 5 . 1 7 Estimated quantities of soil N03-·N (g/m2) in the 1 0·20 cm and 20-30 cm depths from the two models during the experimental period. F280 (10-20 cm depth) ... 40 0 35 11> N - e: E 30 :c ! 25 " 20 .. � .!! 1 5 E 0 1 0 Z " 5 Co> 0 --Model 2 -Model 1 "'�i DAUA F560 (1 0-20 cm depth) 40 '0 35 11> ;' 30 .E ! � 11> 25 " - 20 co � .!! 15 E 0 1 0 " Z 5 U 0 .., co .., co .., co ... .., ., IQ t- DAUA 5280 (10-20 cm depth) 40 --Model 2 '0 35 11> ;' 30 -Model 1 e: E :c m 25 " -co � 20 .!! 0 15 E 1 0 " z U 5 0 .., IQ 11> N ., co ... ... N ., ., IQ co DAUA 5560 (10-20 cm depth) 40 --Model 2 '0 35 -Model 1 11> N - 30 e: E :c Ch 25 " -co � 20 .!! 1 5 E 0 10 z " 5 U 0 .., co .., co .., co ... .., ., '" t- DAUA '0 11> N - e: E :E ! " la � .!! E 0 " Z U ... 0 II> N e: E :c m " -.. � .!! E 0 Z " U '0 11> N - e: E :E -a, " _ . co � .!! E 0 z " U ... o 11> N - e: E :c ! " co � .!! E 0 " Z U F280 (20-30 cm depth) 40 35 --Model 2 30 --Model 1 25 20 15 10 5 0 im!!I'I!!Iiiiiiiiiiiiiiiiiiiiiiii� DAUA F560 (20-30 cm depth) --Model 2 40 35 -Model 1 30 25 20 15 1 0 5 0 .., DAUA 5280 (20-30 cm depth) 40 35 -Model 2 30 -Model 1 25 20 15 10 5 0 .., IQ '" ... N DAUA 5560 (20-30 cm depth) -Model 2 35 -Model 1 40 � 30 l! L� M � � m � M � � � N M � � � = DAUA 1 5 3 Fig. 5 . 1 8 Estimated cumulative leaching of N03--N (glm2) from 1 0-20 cm and 20-30 cm soil depths from urine treatments from the two models during the experimental period. 1 54 5.3.1 1 Nitrification In the Waipawa experiment described in Chapter 3, it appeared as if there was relatively little nitrification after urine application (Fig. 3 . 8). In contrast, in the current experiment there was active nitrification after urine application (Fig. 5 .5 and 5 .6). The gross nitrification rate is the net rate of N03--N accumulation plus the rate ofN03-­ N consumption. However, while the net rate of N03--N accumulation in soils can be measured reasonably easily, it is very difficult to measure the rate of N03--N consumption as it can be taken by plants and microbes or lost through denitrification and leaching. An attempt was made to obtain an approximate estimate for the nitrification rate in the current experiment by assuming that leaching was the major contributor to N03--N loss/consumption. This leaching component was then estimated from the simple model described in the previous section. Nitrification was calculated as fol lows: Day t a.m ( N (t) Soil nitrate (D-iDcm) at day t kg/ha t th Day + Nitrification kg/ha - Leaching (L (t) kg/ha Day t+1 a.m (N (t+1) Soil nitrate (D-iDcm) at day t+1 kg/ha Fig. 5 . 1 9 Schematic diagram to i l lustrate nitrification rate calculation. Daily N03--N leaching values (Let)) and daily soil N03--N (Net)) values in the 0- 1 0 cm depth increment were obtained from the leaching model data. Daily nitrification rate was then estimated as fol lows: 1 55 Nitrification rate on t th day (kg N03 -- N/ha/day) = (N(/+ 1)) -N(/) + L(I) Table 5 . 1 0 Example of nitrification calculation. Calculations up to day 8 are presented for the F560 treatment plots. N03- -N Daily leaching Daily Cumulative Day (0- 1 0cm) Difference (0- 1 0cm) nitrification nitrification kglha kg/ha kglha/day kglha 3 5 1 .2 -0.6 0 -0.6 -0 .6 4 50.6 -0.9 0 -0.9 - 1 . 5 5 49.7 -0.6 0 -0 .6 -2 . l 6 49. 1 2 .3 0 2.3 0.2 7 5 1 .4 2 1 . 5 10 3 l .5 3 l .7 8 72.9 1 0 . 8 5 1 5 .8 47.5 9 83 .7 Cumulative daily nitrification plotted for a l l treatments is shown in Fig. 5 .20. Very low cumulative nitrification was observed in the controls compared to urine-treated plots (Fig 5 .20A). This is not unexpected, as after urine addition, and subsequent urea hydrolysis, large amounts ofNH/-N would have been present to provide a substrate for nitrifiers. The controls at the flat site had higher cumulative nitrification (6 .8 kg N03 - - N/ha) than control plots at the steep site (0.4 kg N03--N/ha ). Similarly, higher and faster nitrification was observed in the flat site than the steep site in the 560 kg urine N/ha treatments (Fig. 5 .20C). In the F560 treatment, cumulative nitrification at the end of the experiment was 32 1 kg N03·-Nlha. This is 56% of added urine N. In the S560 treatment, cumulative nitrification by the end of the experiment was 267 kg N03--Nlha, which was 48% of added urine N . When urine was applied at 280 kg N/ha, nitrification rates were similar at both the flat site and the steep sites, with 1 1 6 and 1 24 kg N03--N/ha respectively being nitrified. This equated to 41 and 44% of added urine N respectively. 1 56 A) Controls 50 40 ca 30 .c Z 20 d , '" 1 0 0 -50 Z 0 Cl -10 .lI: 1 0 20 30 40 50 60 70 80 -20 DAUA B) 280 kg urine N/ha treatments 600 ca 500 .c � 400 -5280 '" 300 0 Z 200 -o-F280 Cl 1 00 .lI: 0 o 10 20 30 40 50 60 70 80 90 DAUA C) 560 kg urine N/ha treatments 600 ca 500 .c Z 400 , -o-F560 '" 300 0 Z 200 -5560 Cl 100 .lI: 0 o 1 0 20 30 40 50 60 70 80 gO DAUA Fig. 5 .20 Cumulative daily nitrification during the experimental period. 1 57 It was observed in Section 5 .3 . 5 that when the levels of 2 M KCI - extractable N03--N were at their maximum at both sites, they were more than three times as high in the 560 kg urine N/ha treatments as in the 280 kg N/ha urine treatment. A similar trend can be seen in the cumulative nitrification values at the flat site. The cumulative nitrification in the F560 treatment was 2 .8 times higher than in the F280 treatment. In the S560 treatment, cumulative nitrification was 2.2 times higher than in the S280 treatment. It can be seen in Fig. 5 .20 that the apparent cumulative nitrification decreased with time on some occasions. This was not surprising as the nitrification calculation assumed that N03--N could only be lost by leaching. In reality however, N03- -N can be consumed by plants, microbes, and denitrification. Losses of N03--N by these processes produced apparently negative nitrification rates . The variation in daily nitrification rates during the experiment is illustrated by the frequency diagrams shown in Fig. 5 .2 1 . Only the nitrification rates up to 45 DAUA were considered for the histograms. This is because, following 45 DAUA, very low NH/ -N levels were observed in the soil (Fig. 5 .4) and as a result nitrification rates would have been constrained by lack of substrate. On most days, the daily nitrification was <5 kg N03--N/halday in control s at both sites (Fig. 5 . 2 1A and 5 .2 IB) . In the 280 kg urine N/ha treatments, daily nitrification values ranged up to 45 kg N03--N/halday at both sites (Fig. 5 .2 1 B and 5 .2 1C). The most prominent differences between the two sites occurred in the daily nitrification values in the 560 kg urine N/ha treatments. For the S560 treatment, there were only 7 days nitrification rates were >20 kg N03--Nlhalday, compared with l 7 days for the F560 treatment. ,., u ;; c . ,., , • 0 120 u Cl ... Z 1 00 Cl 80 c � � ::I 60 � 0 • 40 : • 20 0 o 1 0 20 30 40 50 60 70 80 90 Days after urine application I � F280 -0- F560 --+- 5280 --+-- 5560 - - - Added urine N I Fig. 5 .23 Urine N recoveries (%) during the experimental period, estimated as the sum of soil mineral N, NH3 volatilisation, plant uptake and leaching ofN03--N. High unne N recovenes were observed in the F280, F560 and S560 treatments, particularly in the early stages (Fig. 5 .23). Low urine N recoveries were observed in the S280 treatment throughout the experiment. In addition, lower recoveries were observed in the S560 treatment in the latter part of the experiment when compared to the F560 treatment. It was assumed that urine N not recovered by soil mineral N, ammonia volatilisation, leaching and plant uptake during the experiment was immobilized to complex organic matter in the soil and converted to non-mobile N. Thus, according to Fig. 5 .22, the 1 6 1 estimated immobilisation was larger and occurred earlier i n the steep site than in the flat site. Also, immobilisation in the 280 kg urine Nlha treatment was higher than in the 560 kg urine Nlha treatment at both sites (Fig. 5 .22). By the end of the trial at 88 DAUA the estimated percentages of added urine N that had been immobilised into soil organic matter were 24, 8, 57 and 2 1 for the F280, F560, S280 and S560 treatments respectively. 5.4.2 Mineral N As with the experiment at Waipawa, described in Chapter 3 , this experiment again demonstrated that increased soil mineral N levels would remain for the first 2 months after urine application. For example, by 66 DAUA, 8 1 and 1 1 8 kg urine Nlha were sti l l present in soil in the mineral form, which was 2 1 and 29% of the added urine N in the S560 and F560 treatments respectively. However by the end of the experiment (88 DAUA), urine N recovery in the mineral form was very small, ranging from 0 to 3%. During the first month after urine application, NH4 + was the dominant form of mineral N and during the second month N03- was the dominant form. In contrast to the current experiment, N03--N was not present in large quantities throughout the experiment described in Chapter 3, highlighting differences in soil nitrification activity between the soils ofWaipawa (see Chapter 3) and Ballantrae. A pasture response to urine N was observed only in the flat site, suggesting pasture growth was limited more by N at the flat site than the steep site. However, this was not reflected in the soil mineral N levels in the flat and steep site control plots. The resin adsorbed N levels during the experimental period in flat and steep site control plots (Fig. 5 .24) indicated that soil mineral N levels were similar. Similarly, no differences in mineral N levels between control plots at flat and steep sites were observed in the Waipawa experiment described in Chapter 3 . A) NH/-N z 120�------------------------------------� , + -I � 1 00 z ClS "0 "0 80 � � (5 Ne 60 m () � � 40 c: :1. 20 'iij Q) � O+---��r---��--�--�--�---r--�--� o 10 20 30 40 50 60 70 80 90 100 Days after urine application 120 z , 1 00 '" 0 m Z >-ClS 80 "0 "0 Q) � .0 ... N 60 0 e m () "0 It) 40 cp 0, c: :1. 'iij 20 Q) � 0 0 10 20 30 40 50 60 70 80 90 100 Days after urine appl ication �Flat site -0- Steep site 1 62 Fig. 5 .24 Resin adsorbed NH/-N (A) and N03--N CB) levels during the experimental period in control treatments at both sites . 5.4.3 Priming effect As noted in Section 5.4. 1 , urine N recoveries at the flat site were greater than 100% suggesting a priming effect after urine application. At the time of the maximum apparent priming effect ( 1 2 DAUA for F280 and 27 DAUA for F560), 237 and 379 kg Nlha in excess of that added in urine were recovered from the F280 and F560 treatments. These quantities of N are equivalent to 85 and 68 % of added urine N for 1 63 the F280 and F560 treatments respectively. Priming effects could be observed up to 27 and 45 DAUA for the F280 and F560 treatments respectively (Fig. 5 .22). It is interesting to note that the urine N recovery by soil mineral N alone could detect the priming effect. In the F280 treatment, at 12 DAUA, an additional 1 1 4 kg N more than the added urine N/ha was measured as soil mineral N. In the F560 treatment, at 27 DAUA, an additional 73 kg N more than the added urine N/ha was measured as soil mineral N This suggests that the addition of urine induced soil N mineralisation. This was reinforced by the subsequent measurement of mineralisable N (Fig. 5 . 1 0 & 5 . 1 1 ) . Mineralisable N levels in the urine treated plots were significantly decreased compared to controls . For example, at 27 DAUA, the mineralisable N in the F560 treatment had decreased by 1 24 kg N/ha from the levels in the control plots As illustrated in Fig. 5 . 1 0 and 5 . 1 1 , the mineralisable N levels at 3 DAUA in the 560 kg urine N/ha treatments at both sites had decreased significantly by 27 DAUA. In the F560 treatment, this decrease was 204 kg N/ha. This release of mineralisable N appeared to be reflected in an increase in soil mineral N from 3 to 27 DAUA (Fig. 5 . 1 1 ) . In the F560 treatment the mineral N levels increased by 236 kg N/ha from 3-27 DAUA. The correspondence of decreases in mineralisable N and increases in mineral N are illustrated in Fig. 5 . l 2 . Priming effects were not so prominent at the steep site, with an apparent excess recovery of only 39 kg N/ha over and above that added in urine to the S560 treatment. The greater priming effect on the flat sites indicates that there was a difference flat and steep sites in their mineralisation/immobilisation balance. Norton ( 1 999) suggested that indicators of mineralisation/immobilisation turnover include the availability of substrate for decomposition and the CIN ratio of the substrate. Sakadevan ( 1 99 1 ) observed higher mineralisation of N from soil samples taken from easy slopes than from medium s lopes at a site adjacent to this experiment. This result was associated with the narrower CIN ratio in the soils from easy slopes. Further, Sakadevan ( 1 99 1 ) commented that the narrower C/N ratio of soils taken from low slopes mainly results from the greater deposition of dung and urine on the low slopes. 1 64 However, the CIN ratios of the soils at the two sites in the current experiment do not differ very much. They are 1 2 and 14 for the 0-30 cm soil depth for the flat and steep sites respectively. Several other studies (discussed in Section 3 .2 .3 .6) have pointed out that stimulation of N mineralisation appears to be caused by added salt. After urine application, as urea hydrolysis proceeds, there is a rapid rise in soil pH in the urine patch. This could increase solubility of soil organic matter as well as increase NH3 volatilisation. A large amount of NH3 volatilisation after urine application appeared to occur in this experiment. Free NH3 in the soil could perhaps have had a lethal effect on soil flora and fauna. Mineralisation of these dead materials could have increased the mineral N level in soi l . Further clarification of the reasons for the differences in priming effect between the two soils after urine application would be useful in future experiments . 5.4.4 Ammonia volatilisation The predicted large ammonia volatilisation losses clearly indicated the importance of NH3 volatil isation when studying N balances in urine patches, irrespective of the season. In an N balance study at Ballantrae in which volatilisation was not actually measured, Lambert et al. ( l 982a) used an estimate of 5% of excretal N for volatilisation losses. He argued that this low value was probably appropriate because the conditions at the site were unlikely to be conducive to volatilisation. However, results in this study have shown that there is a high potential for volatilisation even during mid winter. Ball and Keeney ( 1 983) concluded that NH3 volatilisation losses were 5% in winter, 1 6% in spring and 66% summer. However, Holland and During ( 1 977) suggested that within the temperature range of 8- 1 8° C, temperature does not have a large effect on volatilisation of ammonia. The estimated volatilisation losses in the Waipawa experiment described in Chapter 3, were also relatively large compared to the quantity of the added urine N. Thus, both experiments consistently support the view that there is a high potential for volatilisation in urine patches in hill country even during mid winter. 5.4.5 Nitrification 1 65 Unlike the experiment described in Chapter 3 , nitrification was active at both sites in this experiment. Carran et al. ( 1 982), Haynes and Williams ( 1 992) and Williams and Haynes (2000) have also reported active nitrification after urine application. The daily nitrification rate calculation using the leaching model data enabled a comparison of soil nitrification activities between sites. The soil at the flat site had a faster nitrification rate than the soil at the steep site. As nitrification is an important N transformation that governs N losses through leaching and denitrification, observed soil nitrification differences between the flat and steep sites need further clarification in future studies. Steel et al. ( 1 980) reported mean rates of nitrification under field conditions in Kiripaka silt loam of 0 .053 Ilg N/g soil/h which is equivalent to l .2 kg /ha/day (assuming a bulk density of 1 000 kg/m3 and soil depth of 10 cm). The daily nitrification values estimated for controls in the current experiment were reasonably in agreement with this value. However, the daily nitrification rates in urine­ treated plots were much higher than this on most days. This reflects the high levels of NH/-N in soil. 5.4.6 Leaching The simple model used in this experiment suggested that significant leaching losses of N03--N beyond 30 cm might have occurred - up to 3 0% of added urine N in some treatments. These predicted urine N losses by leaching were large compared to some previous studies. Whitehead and Bristow ( 1 990) measured a loss of 1 6% of added N below 30 cm from urine patch areas of pasture during a spring period in which exceptionally heavy rainfall was recorded. Williams and Haynes ( 1 994) observed 1 1 % of the 1 5 N from cattle urine patches leached below the 30 cm soil depth. Monaghan et al. ( 1 989) estimated the N losses from urine patches after varying amounts of drainage through undisturbed soil monolith lysimeters. They observed that after 1 pore volume of drainage ( 1 36 mm) about 1 4 % of the applied urine N was leached below a depth of 340 mm. 1 66 Sakadevan et al. ( 1 993) did not observe leaching ofN03--N after adding 280 kg Nlha as urine N in an experiment conducted on a Iow-slope site, described as Iow fertility, at Ballantrae AgResearch hill country research station. This site had received 1 56 kg single super phosphate (SSP)lhalyr for 7 years and then 1 25 kg SSP /halyr for 8 years, but no Olsen P values were available to compare with the current study. In contrast to the findings of Sakadevan et aI. ( 1 993), estimates of leaching in the current study, at the equivalent application rate of 280 kg urine Nlha, were 2 1 % and 9% of added urine N at the F280 and S280 sites respectively. These sites were classified as high fertility and had received 500 kg SSPlhalyear from 1 973 to 1 980 and 375 kg SSPlha per year from 198 1 to 1 988 (Sakadevan and Hedley, 1 993) . These results suggest that the potential for leaching in hil l country is variable between sites. In large part, this variation of leaching loss appears to be related to nitrification activity, which in turn may be higher at high fertility sites . In a typical hil l country sheep farm, (stocking rate= 14 su ha-l , production = 63 kg wool and 1 80 kg meat/halyr) approximately 1 78 kg Nlha are recycled annually in urine (Haynes and Williams, 1 993) . As pointed out in the literature review, most of this N is accumulated in flat campsite areas due to sheep camping behaviour. Sakadeven et. al. ( 1 993b) estimated, using the data of GiIIingham et al. ( 1 980), that 55% of urine was returned to low slope areas (0- 1 2°). The leaching estimate for the F560 treatment in the current experiment was 30% of added urine N. If this figure was applied to all urine N deposited on low slope areas, approximately 29 kg N/ha ( ( I 78x 55%) x 30%) could be leached annually from the flat sheep campsites in hill country. Realistically however, extrapolation of the data from this study based on one urine application will be approximate at best since leaching losses of N from urine patches vary considerably throughout the year being highest in winter and practical ly non-exist in summer (Sherwood and Fanning, 1 989). Nevertheless, this calculation suggests that leaching losses of N from hill country may be significant in terms of N l ikely to be added by N fixation. In addition, the leaching model used in this experiment demonstrated the importance of measuring climatic data (e.g. rainfall, temperature, evapotranspiration, etc) when 1 67 studying N transformations. When such data are available simple models such as that used in this experiment may provide a cost effective way of assessing the likely significance of leaching losses in experiments of this type where it is difficult to measure leaching losses due to the technical difficulties in setting up lysimeters . The resin technique was valuable in acquiring the daily soil N03--N data in the 0- 1 0 cm soil depth for the leaching model . Studies using resin spikes at different soil depths would also be useful to input daily soil N03--N in 1 0-20 and 20-30 cm soil depths directly to the model 5.4.7 Pasture response Over the whole experimental period, significant pasture responses to urine addition, were observed only in the flat site (Fig. 5 . 1 3 and 5 . 1 4). During the first pasture growth period no response to urine was observed in either site. The winter conditions that were not conducive to rapid pasture growth may have restricted pasture responses. The warmer spring conditions and N availability in urine treated plots would then have created suitable conditions for pasture response during the second and third harvest. In the current experiment the pasture response was present up to 3 months from urine addition. Haynes and Williams ( 1 993) also indicated that pasture response to added urine N normally lasts for 2 to 3 months. However, Sakadeven et al. ( 1 993a) observed a statistically s ignificant response (P< 0 .0 1 ) for up to one-year (June 1 990 to June 1 99 1 ) after adding 280 kg urine N/ha to a low fertility low slope site at Ballantrae. Urine N recovery by herbage in this experiment was low ( 1 - 1 4%) compared to the experiment described in Chapter 3 and some previous studies. Sakadeven et al. ( 1 993a) recovered 1 9% of urine N added at 280 kg urine N/ha. Ball et al. ( 1 979) recovered 22% and 37% of urine N applied at 300 and 600 kg urine N/ha respectively, in herbage within 53 days of application. Williams and Haynes ( 1 994) recovered 1 9% of 1 5N_ urea labelled urine applied at 290 kg urine N/ha within 5 months after urine application. The low urine N recovery by herbage in the current experiment was probably due to the high urine N losses by volatil isation and leaching. Sakadevan et al. ( 1 993) did not 1 68 observe any leaching and consequently they measured relatively high urine N recovery in herbage for a longer period. 5.4.8 Denitrification No direct observations were made to quantify denitrification. Although, theoretically, the potential for denitri fication losses from grazed pastures i s high (Ball and Ryden, 1 984) due to the high levels of readily available C in surface soil , and the high concentrations of nitrate present in soil under urine patches, s everal studies in New Zealand have suggested that denitrification losses may be low. Luo e t al. (2000) measured low denitrification losses (4-6 kg N/halyr) from a poorly drained soil on a dairy farm near Palmerston North. Carran et al. ( 1 995) also observed very low N20 emissions (0 . 5 kg N20-N/yr) from the same site as the current experiment and noted that nitrification or other transformations of urine-derived N did not appear to contribute to overall N20 emissions in an important way at that site . 5.4.9 Immobilisation As noted previous ly in S ection 5 .4. 1 , urine N recovery at the steep site was higher than at the flat site. The soil CIN ratio is normally used as an indicator of potential immobilisation. However in this case the soil CIN ratios were reasonably similar in both sites (see section 5 . 4.3). Deluca and Keeney ( 1 993) suggested that total soil CIN ratios do not reflect the status of the internal N cycle and observed that soluble C to inorganic N ratios better reflected the condition of the N cycle. The ratios of soluble organic C to inorganic N in the 0- 1 0 cm depth increment in the control plots of the current experiment were 6 and 1 5 for the flat and steep sites respectively. This suggests that at the steep site, there is a surplus of soluble organic C leading to a high potential for immobilisation of urine N. Sakadevan et al. ( 1 993 ) also suggested that the majority of added urine N might have been immobilised in soil, as the proportion of observed leaching was very low from an experiment conducted at a low fertility, low slope farmlet at same site. Williams and Haynes ( 1 992, 1 994) reported that significant proportions of urine-derived N are rapidly immobilised to soil organic forms by the active microbial biomass in pastoral soils. 5.5 Conclusions 1 69 Urine application markedly increased the soil mineral N availability, and a priming effect after urine application in flat campsites increased the soil mineral N availability stil l further. The increased mineral N availability generally resulted in increased pasture growth however, the observed response was small and did not persist for a long period, indicating a high potential for N losses from the available pool in urine patches in hill country. Ammonia volatilisation after urine application is a major N loss mechanism associated with hill country urine patches. Also, substantial N losses through leaching can occur in some areas in hill country pastures, especially in flat sheep campsites. The differential leaching potentials in different sites in hill country seem to be related to soil nitrification activity Immobilisation of urine N was high in steep sites in hill country pasture soils. The influence of lower nitrification activity in steep sites might lead to low leaching, and increasing potential for immobilisation. In addition, high amounts of soluble C in steep hill soils could provide a high potential for immobilisation. The resin spikes showed great potential for use as a N measuring technique. Their potential ability to monitor mineral N changes over time, ability to detect N fluxes in soil, and the ability to estimate daily N levels as inputs for leaching models, were highlighted in this experiment. CHAPTER 6 LABORATORY INCUBATION STUDY OF NITROGEN TRANSFORMATIONS IN HILL COUNTRY AND LOWLAND PASTURE SOILS AFTER APPLICATION OF URINE 6.1 Introduction 1 70 It was clear from the experiments described in Chapters 3 and 5 that the N transformations in urine applied to hill country pasture soils are highly variable across the different sites and different topographic units . Nitrification after urine application was much higher in the Ballantrae experiment than in the Waipawa experiment. In addition, in the Ballantrae experiment, nitrification was higher in the campsites than in the steep sites . These variations led to differences in N leaching after urine application. The other important N transformation observed during the previous experiments was a priming effect after urine application. The observed priming effect was higher in the Ballantrae experiment than in the Waipawa experiment. In the Ballantrae experiment, the priming effect was more prominent in the flat campsite than in the steep site. It is important to understand why these variations occur III order to improve our understanding of the N cycle in hill country pastoral soils. The previous two experiments were conducted at two different sites under different environmental conditions. Laboratory incubation provides a means to study N transformations in different soils under the same environmental conditions. This makes it easier to identify soil factors responsible for these variations . The obj ective of this experiment was to study the N transformations after unne application to different soils collected from hill country. To provide greater contrast in soil properties, which might enable critical soil properties to be identified, a number of soils collected from highly fertile, lowland sites were also included in the study. 6.2 Materials and methods 6.2.1 Soils used for the incubation 1 7 1 Seven different soils were used in this experiment. Hil l country soils were taken from the sites of the previous experiments conducted at Waipawa (Chapter 3) and Ballantrae (Chapter 5) . From each of the sites, soil s were collected from flat campsites and steep slopes. Two lowland soils (Kairanga Silt loam and Karapoti silt loam (both Fluvial Recent Soils)) were collected from ongoing experiments at AgResearch (Grasslands) Palmerston North. For each soil type, samples were taken from experimental plots that had received 1 000 kg urine N/ha/year from November 1 996 to August 2000. A further lowland soil (Manawatu sandy loam (Fluvial Recent Soil)) was collected from a paddock at Massey University's No. l Dairy Farm. A summary of the soils used in the experiment is as fol lows. 1 . Summer dry hill country (Waipawa). Flat site (WF) 2. Summer dry hill country (Waipawa). Steep site (WS) 3 . Summer wet hill country (Ballantrae). Flat site (BF) 4. Summer wet hil l country (Ballantrae). Steep site (BS) 5 . Lowland soil with history of urine application (Kairanga silt loam) (KAI) 6. Lowland soil with history of urine application (Karapoti silt loam) (KAR) 7. Lowland soil from dairy farm (Manawatu sandy loam) (MD) 6.2.2 Experimental procedure Bulk soil samples ( 1 0 kg) were collected from the 0- 1 0 cm soil depth and brought from each site to the laboratory. Then herbage, stones, roots and other debris were removed. Soils were crushed by hand and mixed. A sub sample of each soil was used to determine soil moisture content by drying at 1 05°C for 1 6 hours. A weight of field­ moist soil, equivalent to 1 00 g of dry soil, was put in each of 3 6 small plastic cups, for each soil type. Al l soil-filled cups were covered by a p lastic sheet to prevent moisture loss. These cups were then aerobically incubated at room temperature for 14 days. After 1 4 days, 6 mL of cattle urine was added to 1 8 cups of each soil type. This urine was collected from four cows during milking two weeks before the experiment and 1 72 frozen until required. For urine addition, soil was removed from the cup, spread in a small tray and mixed with the 6 mL of urine uniformly. Then the soil was replaced in the cup.The rest of the 1 8 cups were used as controls. No solution was added to the control cups. However, soil in the control cups was also removed from the cup, spread in a small tray, mixed and the cups refilled. The application rate of 6 mL of urinel l 00 g dry soil (40 mg urine Ni l 00 g dry soil) was chosen based on the approximate N content of the 560 kg urine N/ha plots 4 DAUA in the Ballantrae experiment (Chapter 5) and the N concentration (0.7%) of the cattle urine used in this experiment Immediately after unne application, 6 cups (3 unne treated and 3 control) were randomly identified from each soil type for the first sampling to take place 3 DAUA. A KCI -saturated resin spike (see Chapter 4) was buried in each of these cups. These resin spikes remained in the soils for 3 days as described in Section 5 .2 .7 . In addition, immediately after urine addition cylindrical polycarbonate vials to measure ammonia volatilisation (see Section 3 .3 .5) were placed in each of the cups identified for the first sampling. Sampling days during the experiment were 3, 9, 1 5 , 2 1 , 28 and 45 DAUA. Three days before each sampling day KCl-saturated resin spikes were buried in the cups identified for the next sampling At the first sampling (3 DAUA), ammonia samplers were removed, 1 0 mL of deionised water was added and NH3 volatilised was measured as described in Chapter 3 . At each sampling day, 3 urine-treated and 3 control soil cups were sampled. From each cup • resin spikes were removed from soil, shaken with 25 mL of 2 M KC] and extracts were analysed for resin adsorbed NH/-N and N03--N (Section 6.2.3 . 1 ). • 3 5 g of soil was weighed into a 200 mL centrifuge bottle, 1 00 mL of 0.5 M K2S04 was added, shaken for 30 minutes, filtered through "Whatrnan 42" filter paper and the extracts analysed for NH/-N, N03--N (Section 6.2.3 . 1 ), total dissolved nitrogen and dissolved organic carbon (Section 6 .2 .3 .2 and 3) 1 73 • another 3Sg of soil was weighed into a SO mL beaker and fumigated with CHCb as described in Vance et al. ( 1 987). Then the soil was transferred to a 200 mL centrifuge bottle and 1 00 mL of O.S M K2S04 was added, shaken for 30 minutes and the extracts were analysed for total dissolved nitrogen and dissolved organic carbon (as above). • Sg of soil was weighed into a SO mL centrifuge tube, 20 mL of distilled water was added and the soil anaerobically incubated at 30° C for 2 weeks, to measure the mineralisable N content as described in Section 3 .2 .4 . • a sub sample of soil was used for moisture determination by drying at 10SoC for 1 6 hours. 6.2.3 Chemical analysis 6.2.3. 1 Mineral nitrogen (NH4 + -N and N03--N) The N H/-N and N03·-N contents of 2 M KCI and O.SM K2S04 extracts were analysed colorimetrically using a Technicon auto analyser (Searle, 1 975 ; Blakemore et al. , 1987) 6.2.3.2 Total dissolved nitrogen Total dissolved nitrogen was determined as described in Section S .2.6 from the 0.5M K2S04 extracts of fumigated and non-fumigated soils . 6.2.3.3 Dissolved organic carbon Dissolved organic carbon was determined from the O.SM K2S04 extracts of fumigated and non-fumigated soils as described in Section 5 .2 .6 . In this experiment, instead of titrating with ferrous ammonium sulphate, a colorimetric procedure was used. After chromic-sulfuric acid digestion, the dissolved C was determined by the intensity of green colour in the solution at 600 nm wavelength (Heanes, 1 984). Glucose solutions were used as C standards. 6.2.3.4 Microbial carbon Microbial biomass carbon was estimated using the equation ofVance et al. ( 1 987) Biomass C = 2.64 Ec 1 74 where Ec = the difference in dissolved organic carbon extracted by O.5M K2S04 from fumigated and non-fumigated soil. 6.2.3.5 Microbial nitrogen Microbial biomass nitrogen was estimated using the equation of Brookes et al. ( 1 985) Biomass N = 1 .85 EN where EN = the difference in total dissolved nitrogen extracted by O.5M K2S04 from fumigated and non-fumigated soils. 6.2.3.6 Total carbon and total nitrogen in soil Total C and N contents of control soils were determined by combustion using a Leco FP-2000 CNS analyser. 6.2.3.7 Hot water soluble carbon Field moist samples of control soils were sent to AgResearch, Ruakura Research Centre, Hamilton to measure the hot water soluble carbon content (Ghani et al. , 1 999). 6.2.3.8 Clay fixed nitrogen Clay fixed N was determined only on Karapoti soils sampled at 45 DAUA, to check the extent of NH/-N fixation to clay minerals after urine application. Air-dried, finely l 75 ground « 1 50 /lm) Ig soil samples were taken for the analysis from three urine treated cups and three control cups. The analytical procedure followed was that of Silva and Brernner ( 1 966). Analysis includes treatment with alkaline KOBr followed by washing with KCI and decomposition of minerals containing fixed NH4 + ions with HF-HCI solution. The NH4 + ions released were determined by collection and titration after steam distillation of the soil acid mixture. 6.2.4 Statistical Analysis Analysis of variance was carried out using SAS for Windows version 8. Data were analysed using a repeated measures model in mixed procedure (Littell, 1 998) to examine and compare response trends over time. Autoregressive co-variance structure was used in the model. A Least Mean Square method at the 5% significant level was used to compare the means. 6.3 Results 6.3. 1 Organic matter quality of tested soils The major objective of this experiment was to further study the nitrification and priming effects observed after urine application in the previous experiments described in Chapters 3 and 5 . The results observed in the previous chapters seemed to have a relationship with soil fertility and organic matter quality. Sparling and Schipper ( 1 999) identified total carbon (TC), total nitrogen (TN) , soil microbial biomass (SMB), and anaerobically mineralisable N (AMN) as useful parameters to characterise soil quality. Haynes (2000) also suggested that labile organic matter, such as dissolved organic C (DOC) and soil microbial biomass were good indicators of soil organic matter quality. The soil organic matter properties measured in this experiment are i llustrated in Fig. 6. 1 , 6.2 and 6.3 . Relationships between different soil quality parameters are il lustrated in Fig. 6.4. The values for DOC, 5MB-C, 5MB-N, total mineral N (TMN), and AMN were obtained from the grand mean of all sampling days for control soils. These 1 76 properties in controls were relatively constant throughout the experiment (data will be presented in following sections). The values for TC, TN, and hot water carbon (HWC) were analysed only once, from control soils collected before the experiment. The TC levels of tested soils were arranged in descending order from the left in Fig. 6 . 1A. To facilitate the comparison of trends in soil properties the same order of soi ls was continued in al l the other graphs illustrated in this section. Total C levels in the soils collected from hill country (BS, WF, BF, WS) were higher than the soils collected from flat land (MD, KAR, KAI) (Fig. 6. 1A) . This is not a surprise, as other studies have also indicated the presence of a large pool of partially decomposed organic residues in hill soils (Ball et al. , 1 982). In developed flatland, most of the organic material is decomposed rapidly by active microrganisms. The higher ratio of 5MB-C to TC in flat soils than in hil l soils (Fig. 6. 1 E) supports this view. The order of soils according to the TC levels was generally fol lowed by the various measures of labile organic C (Fig. 6. 1 B, 6 . 1 C, 6. 1 D). HWC level in the WS soil was an exception to this trend It is interesting to note that the Ballantrae soils (BF and BS) have the highest levels of DOC, 5MB-C and HWC (Fig. 6. 1 ) . Within Ballantrae, these labile organic matter levels were higher in flat campsites (BF) compared to the steep slopes (BS). This i s mainly due to the high dung and urine return to the flat campsites . The KAI and KAR soils had low concentrations of labile organic matter (DOC, 5MB-C, HWC). The KAR and KAI soils were collected from experimental plots that had received 1 000 kg urine Nlha for each of the last five years. As observed in previous chapters (3 and 5), a priming effect after urine application might have resulted in the low levels of easily decomposable organic matter in KAR and KAI soi ls . The labile organic fraction is the main energy source for soil microorganisms and i t is a primary source of mineralisable N (Haynes, 2000) . The KAR and KAI soils had lower 5MB-C (Fig. 6. 1 C) and mineralisable N levels (Fig. 6 .2D) compared to the other soils used in the experiment. A) Total C (TC) BS WF BF WS MD KAR KAI Soil C) Soi l Microbial Biomass C (SMB-C) ·0 800 .,-- -----------, t � l l l , I , 1 I I, l l "- BS WF SF WS MD KAR KAI Soil E) (SMB-C I TC) % I I I • I i 1 , 1 BS WF BF WS MD KAR KAI Soil G) (DOC/TC) % BS WF SF WS MD KAR KAI Soil 1 77 B) Dissolved Organic C (DOC) i � I I I , I , I ,I . ,. 1 BS WF SF WS MD KAR KAI Soil D) Hot water organic C (HWC) 1 � 11 ,. ,I ,. ,I , • . • 1"- BS WF BF WS MD KAR KAI Soil F) (HWCI TC) % 1 , 1 • SS WF SF WS MD KAR KAI Soil H) (SMB-C/HWC) % WF BF WS MD KAR KAI Soil Fig. 6. 1 Soil carbon related organic matter properties studied during the experiment. Ballantrae hill country soils (BF, BS) contained higher HWC than the soil collected from Massey University' s No. I Dairy Farm (MD). Ghani et al. (2000) also observed a lower level of HWC in intensively managed dairy soils compared with extensively managed sheep grazed pastures . Further, they commented that long-term N application reduced the HWC significantly. As noted above the KAR and KAI soils that were collected from experimental plots that had received large amounts of urine N over the last five years had the lowest HWC levels compared to the other soils . 178 When the ratio of labile organic C to TC is considered, the order of soils was opposite to the order of absolute levels of labile organic C (Fig. 6. 1 E, 6. 1 F, 6. 1G). A higher proportion of the TC pool was present as labi le organic C in soils collected from flatland, than in the soils collected from hil l country. As noted above, this appears to be related to higher microbial activity in flatland soils than in hill soils. A) Total N (TN) � � 1111 1111 I BS WF BF WS MD KAR KAI Soil C) Soil microbial biomass N (SMB-N) BS WF SF WS MD KAR KAI Soil B) Total mineral N (NH/ + NOl') (TMN) 75 ,-----------------------. .� 50 � � 25 � o SS WF BF WS MD KAR KAI Soil D) Anaerobically mineralisable N (AMN) 200 r---------------------� :g 150 en � 100 z � 50 o BS WF SF WS MD KAR KAI Soil Fig. 6.2 Nitrogen related organic matter properties studied during the experiment. The order of soils according to the TN levels (Fig. 6 .2A) shows that the soils collected from hill country, except WS, contained higher levels of TN than the soils collected from flatland. The TN level of WS was similar to the TN levels of soil s collected from flatlands. It might be expected that the flatland soils would have high TN levels as these soils were collected from a shelter belt area of a dairy paddock or experimental plots that had received urine N regularly over a number of years . Substantial N losses occurring through ammonia volatilisation, leaching, denitrification and plant uptake might have caused the lower levels of TN in these flatland soils . Although these losses were also observed in hill soil s in previous chapters, the TN levels remain higher than the flatland soils - supporting the view, that N is conserved in hill soils more effectively (Sakadevan et al., 1 993). 1 79 Among the soils collected from hill country, mineral N levels were higher in flat campsites than steep sites. This was not evident in the previous field experiments described in Chapters 3 and 5 , although the soils were taken from similar sites. As noted in the literature, higher soil N levels are due to N transfer to camp areas through urine and dune. Surprisingly, the 5MB-N was higher in soils collected from steep hill country (BS, WS) than from flat campsites (BF, WF) and flatland soils. The ratio between carbon and nitrogen (CIN) has widely been used to study N availability in different soil systems. Soils collected from hill country had higher CIN ratios ( 1 2- 14) than the flatland soil s ( 1 0) . Although the soils collected from hill country had higher TN than soils collected from flatland, the large pool of partially decomposed organic residues resulted in higher CIN ratios (Fig. 6 . 1 ) in hill soils than flat soils . Ball et al. ( 1 982) reported that many top soils in the hill country of the lower North Island contain very large quantities of nitrogen, rendered largely unavailable by the relatively wide CIN ratio of the soi l organic matter present. It was clear from the literature review that hill country steep soils and hill country flat campsites are fundamentally different in how they cycle N. However, the observed soil CIN ratios do not reflect these differences . As noted in Section 5 .4 .9, Deluca and Keeney ( 1 993) suggested that total soil CIN ratios do not properly reflect the status of the internal N cycle, and pointed out that the ratio of soluble C to inorganic N is a better option. Hill country steep soils (BS and WS) had markedly higher ratios of dissolved organic C to mineral N than other soi ls (Fig. 6 . 1 ) . This indicated that there was a surplus of DOC, well above the levels of inorganic N in these soils . CIN SS WF SF WS MD KAR KAI Soil DOCfTMN ss WF SF ws MD KAR KAI Soil Fig. 6 .3 Soil carbon to nitrogen ratios in experimental soi l . 1 80 In order to identify the relationships between the various organic matter characteristics regression relationships were developed using the data in Figs. 6 . 1 , 6 .2 and 6.3 . (Fig. 6 .4) . The labile organic C, assessed by DOC, 5MB-C and HWC was positively related to the total C (R2 = 0 .8 , 0 .6, 0.4. for DOC, MBC, and HWC respectively) . This suggests that the total C content was a major determinant of the quantity of labile organic matter present. There was a strong "negative relationship between the ratio of 5MB-C/TC and the C/N ratio (Fig 6.4 J). These two parameters could both be considered indicators of organic matter quality. The HWC was strongly related to DOC, 5MB-C and AMN (Fig. 6.4 D-F with R2 values of approximately 0.8 in each case). Ghani et al. (2000) also observed strong relationships between HWC and both microbial C and mineralisable N. Mineralisable N was positively related to the labile organic matter fractions of DOC, HWC and 5MB-C (Fig. 6.4 G, F and L). Soil microbial biomass plays a dual role in the soil, as an agent of decomposition and release ofN from fresh organic residues and soil organic matter and secondly, as a pool of labile soil N. 5MB-C and AMN showed a strong (R2=0.7) positive relationship (Fig. 6 .4 L) supporting the view that the microbial population plays an important role in decomposition and release of N from soil organic matter. Surprisingly, the 5MB-N and AMN were negatively related, which questions the view of microbial biomass N as a labile pool of soil N. I t is interesting to note that a strong negative relationship was observed between total mineral N (TMN) and soil microbial biomass N (SMB-N) (Fig. 6.4 H). 3000 2500 u = 2000 � o .!! .. � '" � - 1500 - '" o :>. ", - 1000 u 180 160 .2 140 ; =- 120 � � 100 "0 � 80 " '" � :L 60 : - 40 o 20 500 A) y • 0.03x • 3&.4.42 Rt - 0.39 • • /. 20000 40000 60000 Total C (1'91 9 soli) 0) y • 0.05 . .. 40.01 /. .. o +-----__ -----,----� � = ::: 1 .. 0 1 Q ': 1 000 2000 Hot Water C 1'91 9 soli) G) y • 0.65 • • 30.46 R� • 0.44 3000 � � 100 I i 5: +---r--'---r--� 2.5 .. 2 U -;;; 1.5 "0 ... � III lE 0.5 !!l. 50 100 150 Oissolved Organic C (1'9 1 9 soli) J) 200 � y -= -a.,Ix .. 3.25 Rl . 0.7' . 10 1 5 CIN 200 u � _ 150 • = "' 0 � :. 100 � o. � .: 50 c 1000 u = 800 e = � � 600 III '" 400 200 S) • Y " O?-OOX + 21.21 Ff - 0.80 . 20000 40000 Total C (�g 1 9 soli) E) y . O.18x + 252.18 R" 0.84 1 000 2000 Hot Water C (1'9i 9 soil) H) 60000 3000 40 ,--------------------, z 35 E =- 30 E 0 25 o .. � � 20 � � 15 � .: 10 � 200 y • ·O.69x + 36.97 W - 0.97 20 40 Total mineral N K) y . ·1.75x " 136.00 � - 0.36 60 10 20 30 40 Microbial biomass N ()Jg f 9 soil) 800 u 700 " � 600 .. -g � 500 in ; 400 � D 300 .D :>. g - 200 � 100 C) y . O.Ob + 243.11 Rl · 0.63 1 8 1 o +---�--�-� 200 z 150 ., = D O .. .. .!! CD 100 -;;; - � '" � 2: � 50 u 800 � 700 fIJ =- 600 g � 500 ii5 Cl 400 � � 300 .g ..:: 200 t 100 20000 40000 60000 Total C (I'Qi 9 soi� F) y · 0.05. · 27,51 RJ • 0.83 1000 2000 Hot water C (1'9 1 9 soil) I) y - 3.25 • • 158.63 3000 � o +---__ --�--__ --, 200 � � 150 -:; .. � � 100 � z " '" i .?: 50 50 100 150 Oissolved Organic C (1'91 9 soil) L) y • 0.2" • • 23.62 � .o/ . 200 o +,----__ ----�--__ --� 200 400 600 Microbial Biomass C (jJgIg soil 800 Fig. 6.4 Relationships between organic matter quality parameters. 6.3.2 Mineral N 1 82 As noted in previous chapters, mineral N comprises NH/-N and N03--N. The changes in the relative quantities ofNH/-N and N03--N after urine application are discussed in this section. Mineral N changes during the experimental period in urine treated and control pots are i llustrated in Fig. 6 .5 and the recovery of added urine N at the beginning and end of the experiment is i llustrated in Fig. 6.6. Not surprisingly, as seen in previous chapters, urine application significantly (P< 0.05) increased the mineral N levels in all soils (Fig. 6 .5 & 6.6). However, the persistence of the elevated mineral N levels in urine treated soils with time was very different to the patterns observed in the field experiments described in previous chapters . In the field experiments, close to 1 00% of added urine N could be recovered as mineral N immediately after urine application. Thereafter, mineral N levels decreased rapidly with time. In contrast, in the current incubation experiment, at the first sampling mineral N could only account for only 64%-8 1 % of added urine N. Urine N present as soil mineral N at 3 DAUA was slightly higher in hil l soils, ranging from 68% to 8 1 % than in the lowland soils which ranged from 64% to 74%. The mineral N levels of urine treated soi ls collected from hill country remained relatively constant from 3 DAUA until the end of the experiment at 45 DAUA (Fig. 6.5) . In contrast the mineral N levels of urine treated soils collected from flatland increased significantly (P<0.05) from 3 to 45 DAUA (Fig. 6.5) . At the last sampling (45 DAUA), mineral N levels in urine treated lowland soils had increased by 65 to 84 Ilg/g soil from the levels at 3 DAUA. These quantities are equivalent to 24% to 30% of the mineral N present at 3 DAUA. '50 '00 350 300 � 250 i 200 150 100 50 Lowland soils Karapoti (KAR) Kalranga (KAI) Manawatu (MO) Z '50 Z '50 � '00 � il 350 � 300 � g � '00 � 350 U 300 250 , '" 200 O D ", " 150 ,z 100 � 50 � 0 � a 250 , '" 200 g � 150 ,z 100 � 50 0--<>-" -0 C 0 10 20 30 '0 50 10 20 30 '0 50 DAUA DAUA 10 20 30 DAUA -.--Urine -0--Control -+-Urine � Control .......... Unne -<>-Control z � U i g , '" g � ,z � 0 Hill country campsite soils Ba!lantrae-Flat (SF) '50 Z '50 J1 '00 '00 '.i: 350 --------350 ------- U 300 300 � g 250 250 , '" 200 200 g � 150 150 100 ,z 100 � SO �rl)----50 � -<> 0 0 0 0 10 20 30 '0 50 0 10 20 30 '0 50 DAUA DAUA -+-unne -<>-Control .......... Unne -0-Control Hill country steep soils Ballanlrae-Sleep (8S) 10 20 30 40 so DAUA .......... Urine -0-Control Walpawa·Steep (WS) '50 �-------� '00 350 100 50 0 ����� ____ -4� 10 20 30 40 50 DAUA .......... Urine -0-Control '0 1 83 50 Fig. 6.5 Effect of urine application on O . 5 M K2 S04-extractable soil mineral N. (The statistical analysis of the data in this figure is included in Appendix 3 ) . These trends were reinforced when the data were presented a s recovery o f added urine N (Fig. 6 . 6). At the end of the experiment (45 DAUA), the urine N recovery as soil mineral N had increased in all soils except the BF soil (Fig. 6.6) . These increases were markedly higher in the lowland soils. 1 00 Z 90 '0 80 Cl) '0 70 '0 ca 60 - -o � 50 � � 40 Cl) > 30 0 u 20 Cl) 0:: 1 0 0 KAR KAI MD SS Soil SF o 3DAUA . 45DAUA WF WS 1 84 Fig. 6 .6 Percentage of urine N recovered as soil mineral N at the beginning (3 DAUA) and end (45 DAUA) of the experiment. The increase in unne N recovery as mineral N can be explained by one of two possibilities. The first possibility is the remineralisation of urine N, which was initially immobilised by microbes. The second, as discussed in previous chapters, is the release ofN from resident organic matter as a result of urine addition. 6.3.3 Ammonium Concentrations of O.5M K2S04-extractable NH/-N (/lg NH/-N/g soil) and resin­ adsorbed NH/-N (Ilg NH/-N/5 cm213 days) over the duration of the experiment are presented in Fig. 6.7 and 6 .8 . Urine application significantly increased the soil NH/ -N levels in all urine treated soils at the first sampling, 3 DAUA (Fig. 6.7 and 6.8) . "" JOO "" '§ 200 i ,so "Xl 50 0 0 Lowland soils Karapoh (KAR) Kalranga (KAI) "" JOO \ "" � '§ 200 'i tSO "Xl 50 0 10 " 30 '" 50 0 10 20 30 " 50 CAUA CAu;. �Unne �ControI ........ Urine � Control Hill country campsite soils >; ':t. z Ballanltae--Flal (BF) "" �--------, JOO 10 20 30 110 so CAU' ......-Urine -<>-Control Walpawa--Flal (WF) 10 20 30 ,f,() so Hill country steep soils Ballantrae- Steep (B5) L � � o � '! l lSO � lOO .z � 50 o o����--�� o 10 � � � so OAUA Walpawa-Sleep (WS) 10 20 JO so DAUA .......... Unne -<>-Control 1 85 Manawatu (MD) " 30 50 CAU' -+-Unne -<>-Control Fig. 6.7 Effect of urine application on soil 0 .5M K2S04-extractable NH/ -N. (The statistical analysis of data in this figure is included in Appendix 3). By 3 DAUA, 245-3 1 8 Ilglg soil of urine N were present as 0.5M K2S04-extractable soil NH/-N. These quantities are 61 % to 80% of added urine N, indicating rapid urea hydrolysis after urine application. However, in lowland soils, relatively lower quantities (245-259 Ilglg soil) of urine N were present as 0 .5M K2S04-extractable NH/-N at 3 DAUA compared to the hill country soils (272-3 1 8 Ilg N/g soil). As the trial progressed, the elevated levels of 0 .5M K2S04-extractable soil NH/-N in urine treated soils decreased towards the levels of the controls in all soils except in the WS soils. 1 86 In the urine treated, lowland soils the elevated O.5M K2S04-extractable soil NH/-N levels declined rapidly with time and reached the levels of the controls by 1 5-2 1 days. In the urine treated hill country soils, the O.5M K2S04-extractable soil NH/-N levels declined more slowly. The O .5M K2S04-extractable soil NH4+-N in the urine treated BF soils had reached control levels by 45 DAUA but in the BS and WF soils O .5M K2S04- extractable soil NRt +-N levels were still significantly higher than controls at 45DAUA. The O .5M K2S04-extractable soil NH/-N in the urine treated WS soil, remained at similar levels to those at 3 DAUA throughout the experiment. Karapoli (KAA) 10 20 30 40 50 OAUA ___ Urine -0-Control Bailantrae-Flal (SF) Lowland soils Kauanga (KAI) ��D 300 2SO 200 ,so '00 so o o 10 20 30 "0 50 OAUA ........ Urine -0-Control Hill country campsite soils W3lpawa-Flat (WF) Manawalu (MD) �[J '50 '00 50 o 10 20 30 <&0 50 CAUA �Unne �Control � �[[J � �E;J ':L • ':L • z i;- i l I "� >l � "0 >l � .� � '00 . .", 50 .� :1 0: 0 0: 0 '0 20 30 '" 50 0 10 20 30 '0 '" OAUA OAUA --+- Unne -0-Control .......... Unne -<>-Control Hill country steep soils Wa1pawa·$leep (WS) o ,0 20 30 40 50 OAUA ........ Urine -o-Control 8allantr�Steep (8S) 400 350 300 �E V---'00 so o o 10 20 30 40 50 CAUA ......... Urine -0-Control Fig. 6.8 Effect of urine application on resin-adsorbed NH/-N. (The statistical analysis of data in this figure is included in Appendix 3) . 1 87 At the end of the experiment, in lowland soils, < 1 % of applied urine N could be accounted for as 0.5M K2S04-extractable soil NH/-N. In hil l country soils, 0.5M K2S04-extractable soil NH4 +-N accounted for 4% (BF) to 25% (WF) of applied urine in flat campsite soils and 40% (BS) to 66%(WS) in steep soils at 45 DAUA. In most soils, resin-adsorbed NH/-N (Fig. 6.8) followed similar trends to the 0.5M K2S04-extractable NH/-N. In hill country, urine treated soils there was a consistent pattern; resin-adsorbed NH/-N at 1 0- 1 5 DAUA dropped significantly and then recovered by 20 DAUA. The exact cause for this is not clear. 6.3.4 Nitrate Concentrations of 0.5 M K2S04-extractable N03·- T ( Ilg N03--N/g soil) and resin­ adsorbed N03--N ( Ilg N03--N/5 cm2 resinl3 days) over the duration of the experiment are presented in Fig. 6.9 and 6. 1 0 . At the first sampling (3 DAUA), there were no significant differences in 0 .5M K2S04 - extractable N03--N between the control and urine-treated soils. Soil N03--N levels then built up with time in the urine-treated soi ls due to nitrification. However, the increase of soil N03--N with time varied greatly between the tested soils , ranging from 38 to 40 1 1lg N03--N Ig soil at the end of the experiment. The 0 .5 M K2S04-extractable N03--N levels in all the unne treated lowland soi ls increased rapidly up to 1 5-20 DAUA and thereafter the rate of increase of soil N03--N dropped. The changes in 0 .5 M K2S04-extractable N03--N levels over the duration of the experiment in hill country soils varied between Waipawa and Ballantrae sites and between soils from flat and steep areas. Ballantrae soils had faster soil N03--N accumulation than their Waipawa counterparts and soils from flat areas had faster nitrification than the steepland soils. 'SO "'" 3SO 300 � "" � 200 � lSO 100 SO 0 0 Lowland Soils Karapoti (KAR) Kairanga (KAt) Manawatu (MD) '50 '50 '00 "'" I 3SO C � :: "' 200 � 150 100 50 0 3SO L 300 � "" "' 200 � '50 100 SO 0 - , 10 20 30 .. SO 0 10 20 30 .. 50 0 10 20 30 OAUA OAUA QAUA -Urine -Control -Urine -Control -Unne -Control Hill country campsite soils Baltanttae-Flat (SF) 50 L.e::::::::::====:...J 10 20 30 40 50 OAUA -UrIne -Control Waipawa-Flat (WF) Z .50 .-------------� a 400 ! 350 � 300 1/ '§ 250 � i 200 g 150 .z 100 a 50 ci ��� ______ �� 10 20 30 40 so DAUA -Urine -Control Hill country steep soils Ballantrae-Steep (8S) z .50 r-------------� ci' "'" ! 350 1i 300 1i �"" � i200 � ISO � 100 a 50 ci ��----_===�� 10 20 30 40 so OAUA -Unne -Control z Waipawa-Steep (WS) '50 r----------------, '00 10 20 30 40 50 OAUA -Urine -Control 1 88 .. 50 Fig. 6.9 Effect of urine application on 0.5 M K2S04 -extractable soil N03--N. (The statistical analysis of data in this figure is included in Appendix 3) . At the end of the experiment, in lowland soils, 81 %-87% of added urine N was present as 0 .5M K2S04-extractable soil N03--N. In hill country soils, 0 .5M K2S04-extractable soil N03--N accounted for 55% (WF) - 68% (BF) applied urine N in flat campsite soils and 9% (WS) - 44% (BS) in steepland soils . Resin-adsorbed N03--N in al l soils (Fig. 6. 1 0) fol lowed the same patterns as the 0 .5 M K2S04 -extractable N03--N. However, resin-adsorbed N03--N in KAR, KAI, MD and BF soils dropped somewhat in the latter part of the experiment. This could be due to loss ofN03--N from the soil by denitrification. KatapoII (KAR) '500 ,--------, 10 20 30 40 SO OAUA � Unne �Conttol '500 ! '200 ,<2 900 � "'" g 300 Balantrae·Flat (BF) 'SOD ,---------, 10 20 30 40 50 CAUA ........... Unne N �Control Ballanlrae-Steep (85) 1500 � '200 � 900 ,- � 600 g 300 � 0 ... 0 10 20 30 40 50 OAUA -+-Unne �Control 1 89 Lowland soils Kalranga (KAI) Manawatu (MO) '500 ,-------, 10 20 30 4() 50 10 20 30 40 50 DAUA OAU" ......... Unne � Control ......... Urn! � Conlrol Hill country campsite soils Wa!paWa·FLat (WF) 'SOD r-----------, z 6' � 1200 z • � El 900 � � "'" � '" � :1 300 a: 0 U;::.o.:��� o 10 20 � 40 50 OAUA -+- Unne -c>-Control Hill country steep soils Willpawa-Steep (WS) 'SOD r----------, 10 20 30 40 50 OAUA --+-Unne -c>-Control Fig. 6. 1 0 Effect of unne application on resin-adsorbed N03 --N over time. (The statistical analysis of data in this figure is included in Appendix 3 .) 6.3.5 Mineralisable N The effect of urine application on readily available mineralisable N in the experimental soils is i llustrated in Fig. 6. 1 1 . Among the unne treated, lowland soils, mineralisable N levels were significantly increased compared to the control throughout most of the experiment in the KAR and Manawatu MD soils. In the KAI soil, urine application only increased mineralisable N levels at 3 DAUA. 1 90 Lowland soils Karapoti (KARJ Kairanga (KAl) Manawatu (MO) 300 ,---------, 300 ,--------..., 300 ,----------, 250 >SO 200 � oo ,SO � ': so o ��-�� ___ � 10 20 30 40 50 10 20 30 40 50 o '0 20 JO 40 50 OAUA OAUA OAUA -+- Urine -0-Control � Urine -0-Control � Urine -0-ConlroJ Hill country campsite soils Banantrae-Flat (SF) Waipawa-Flal (WF) 300 ,-----------, 300 ,----------, so 10 20 30 40 so 10 20 30 40 50 CAUA OAUA -+- Urine � Control Hill country steep soils Ballantrae-Steep (8S) Waipawa-Steep (WS) 300 ,---------, 300 .---------, 250 so so 0 10 20 30 40 50 10 20 30 40 50 CAUA OAUA -+-Urine -0-Control Fig. 6. 1 1 Effect of urine application on soil mineralisable N with time. (The statistical analysis of data in this figure is included in Appendix 3). The mineralisable N levels in urine treated KAR soils were higher than control by 1 7-40 )lg/g soil from 3-28 DAUA. These quantities are equivalent to 4 - 10% of added urine N. From 3-28 DAUA, mineralisable N in urine treated MD soils increased by 34-45 )lglg soil and these quantities were equivalent to 9- 1 1 % of added urine N . At the beginning, 3 - 15 DAUA, mineralisable N levels significantly increased compared to control in all urine treated hill country soils. The increase in mineralisable N was higher in Ballantrae soils than Waipawa soils. 1 9 1 At 9 DAUA, mineralisable N levels were 97, 95, 47 and 53 I-lg/ g soil higher than the controls in BF, BS, WF, and WS soils respectively. These differences were statistically significant and were equivalent to 24%, 24%, 1 2% and 1 3% of added urine N respectively. By 2 1 DAUA, the mineralisable N levels in urine treated hill soils had declined to the levels of the controls. Interestingly, at 45 DAUA the mineralisable N levels in urine treated, hill country steep soils (BS and WS) were significantly lower than the controls . 6.3.6 Dissolved organic carbon (DOe) Previous studies have reported that solubilization of soil orgamc carbon occurs following urine application (Monaghan and Barraclough, 1 993). It was pointed out in Section 6.3 . 1 that DOe is strongly related to soil microbial biomass and mineralisable N. Therefore, observing the changes in DOe are vital in the urine N cycle . The effect of urine application on soil DOC levels is i l lustrated in Fig. 6 . 1 2 . The DOC levels after urine application in lowland soils generally increased compared to control, suggesting solubilization of organic matter and release e to soil solution. These increases were statistically significant in the MD soils . In contrast, the DOe levels in urine treated hill soils did not show any significant differences compared to the control . 6.3.7 Soil microbial biomass (SMB) Haynes and Williams ( 1 999) reported that soil microbial biomass and microbial activity are high in stock camp areas in hill pastures. This is mainly due to the transfer of nutrients and organic matter to the camp areas via dung and urine. However, in the current experiment there was no significant difference observed in 5MB-C in urine treated soils compared to control soils for any of the soils used (data not presented). Ka,.pot, (KAR) 250 ,----------, 200 g � 150 ° i '00 �_� 50 · 10 20 30 40 50 OAUA -+-Unne ....0-Control 250 200 (.) '§ 150 8 � 100 50 0 0 Lowland soils Kalranga (KAI) Manawatu (MO) 250 r-------� 200 ,0 20 30 40 50 10 20 30 40 .so DAUA CAUA ..-.-Urine -0-Control -+-Unne -0-Control Hill country campsite soils BaJlantrae--Flat (SF) 2� r-------, 200 U � 1SO 0 0> ° � ,oo 10 20 30 40 50 DAUA �Urlne -o-Control 250 200 u '§ 1SO 8 i 100 50 0 Walpawa·Flat (WF) � =: 0 '0 20 30 40 DAUA -+-Urine -0-Control Hill country steep soils Ballantrae-Steep (85) Waipawa-Steep (W S) 250 250 200 200 u "£ 150 � 8 ! 1SO 8 i ,oo � ,oo 50 50 0 '0 20 30 40 50 0 '0 20 30 40 DAUA DAUA -- Unne -o-Control __ Urine -0-Control 50 50 1 92 Fig. 6 . 1 2 Effect of urine application on soil DOe levels. (The statistical analysis of data in this figure is included in the Appendix 3 ) It was evident from the previous sections of this chapter that microbial activities, l ike nitrification, were greatly increased by urine application. This suggests that microbial activity was more affected by urine application than was microbial biomass size. Lovel l and Jarvis ( 1 996) also did not observe an effect of urine addition on biomass size but did observe increased activity. They commented that substantial qualitative changes within the 5MB could take place quickly and independently of any change in biomass SIze. 1 93 In contrast to 5MB-C, 5MB-N was increased after urine application (Fig. 6 . 1 3) . Most of the significant increases occurred from 9-2 1 DAUA. The effect of urine application on 5MB-N levels is illustrated in Fig. 6 . 1 3 . Lowland soils Karapoti (KAR) Kairanga (KAI) Manawatu (MD) 160 .-------, 160 .....--------, = 120 � � � � 120 � .!1' 80 "' � 40 CD '" :E � 80 ", 0> =>. 40 10 20 30 40 so DAUA -+-Urine -<>-ContrOl 10 20 30 40 SO DAUA -+-Urine -0-ContrOl 10 20 30 40 50 OAUA -+- Urine -0-Control Hill country campsite soils 160 z = 120 cD g ::E .!1' 80 ", 0> =>. 40 � g 120 CD 0> 80 ::E ­", 0> =>. 40 Sallantrae·Flat (SF) Waipawa-Flat (WF) 160 /--- z = 120 f\v cD g ::E .!1' 80 "' � 40 10 20 30 40 50 10 20 30 40 SO OAUA DAUA � Urine -0- Control -+- Urine � Control Hill country steep soils Sallantrae-Steep (SS) Waipawa-Steep (WS) 160 .....--------, = 120 � .!1' 80 � g � Cl) � 40 10 20 30 40 SO 10 20 30 40 SO DAUA OAUA ........ Urine -0- Control -+-- Urine �Control Fig. 6 . 1 3 Effect of urine application on soil microbial biomass N with time. (The statistical analysis of data in this figure is included in Appendix 3) At 1 5 DAUA, all of the soils had significantly higher (P<0.05) microbial N values in urine treated pots compared to control. In lowland soils, this increase was 40 - 45 )lglg soil which was 1 0%- 1 1 % of added urine N. Interestingly the increased microbial N in hill country soils was approximately twice that in the lowland soils, ranging from 7 1 -87 )lg 0 soi l . These quantities were 1 8%-22% of added urine N. 6.4 Discussion 1 94 As noted in Section 6.3 .2, the proportion of added urine N recovered as soil mineral N at the beginning of the experiment was different in the current incubation experiment than in the field experiments described in Chapters 3 and 5 . In the Waipawa experiment at 1 and 6 DAUA, urine N recovery as soil (0- 1 5 cm) mineral N was close to or > 1 00% (Table 3 .4). The urine N recovery as soil (0-30 cm) mineral N at 3 DAUA in the Ballantrae experiment was 48% - 75%. The urine N recovery as mineral N in the current incubation experiment at 3 DAUA was 68% - 73% in Waipawa soils, 78%- 8 1 % in the Ballantrae soils and 64 %-74% in the lowland soils . The unne N recovered at 3 DAUA is presented in Table 6. 1 . The total unne N accounted for at 3 DAUA was calculated using urine N recovered as soil NH/- , N03-­ N, resin N and anaerobically mineralisable N (AMN). Table 6 . 1 Apparent fate of urine N at 3 DAUA. * (rate of urine N was 400 Ilg N/g soil) i .e . values are treatment minus control). Fate of urine N at 3 DAUA ( Ilg N/g soil) Soil NH/-N N03·-N AMN 5MB- N Resin N Quantity of N unaccounted for (A) (B) (C) (D) (E) (400*- (A+B+C+E)) (% of added urine N) KAR 245 1 2 1 7 5 1 4 1 22 (3 1%) KAI 250 1 5 47 29 3 85 (2 1 %) M D 259 35 43 7 5 5 8 ( 1 5%) B S 3 1 1 3 0 48 3 83 (2 1 %) BF 3 1 8 4 45 1 5 4 29 (7%) WS 272 I 32 69 3 92 (23%) WF 294 0 49 28 3 S4 ( 1 4%) In this experiment, reSin N was added to the other measures of mineral N when calculating recovery of urine N because the resin strips had been inserted in the pots for 3 days prior to analysis. They would therefore have adsorbed NH/-N and N03--N that would otherwise have been included in the mineral N. The N adsorbed to resin was from the 1 00 g of soil in the cup. Thus resin adsorbed N was also expressed as Ilg/g soil . 1 95 The percentage of urine N unaccounted for at 3 DAUA was reasonably large, ranging from 7%-3 1%. Some of the unaccounted for urine N could have been lost through ammonia volatilisation although the quantity of NH3-N trapped by the volatilisation samplers was very small , ranging from 1 -9 Jlg N/sampler. This was not surprising as the cattle urine was well mixed with soil before placement into the incubation cups. However, there was stil l a possibility that some of the added urine N was volatilised and not detected by the samplers. Therefore, a simple experiment was carried out to check on the possibility of volatilisation when the same amount (6 mL ) of urine was mixed with the same amount of soil ( 1 00 g). Only KAR and WS soils were used. Urine (6 mL ) was well mixed with 1 00 g soil and placed in the same incubation cups. The soil­ filled cup and a small beaker with 20 mL of 0.2 M H2S04 were placed in a 1 5x 1 5x 6 cm plastic container. The container was then sealed and allowed to stand for 3 days. The NH3-N trapped in the 0.2 M H2S04 was analysed. The results confirmed that the volati lisation losses were minimal, ranging from 1 2- 1 5 Ilg Ni l 00 g soil. Fixation of NH/-N to clay minerals is another possible s ink for the unaccounted for urine N. Carran et al. ( 1 982) accounted for 1 0% of added urine N as clay fixed N and Crush and Evans ( 1 988) observed that fixed 1\TH/-N rose sharply and declined more slowly in plots treated with urine. Therefore, an attempt was made to check on the possibility of urine N loss through NH/-N fixation by clay minerals. However, the method commonly found in the literature, described in Section 6.2.3 .8 (Silva and Brernner, 1 966) was time consuming and involved the use of toxic chemicals such as concentrated hydrofiuoric acid (HF). Therefore, a preliminary analysis was carried out using only the KAR soil before deciding whether or not to analyse the other soils. The KAR soil was selected for two reasons. Firstly, it had the highest amount of unaccounted for urine N and secondly, Crush and Evans ( 1 988) had reported fixed N values in Karapoti soils with and without urine addition. The decision to undertake analysis for clay fixed N was taken at the end of the experiment. Thus this analysis was done on soils from the sampling 45 DAUA. The results suggested that the urine application had no effect on clay fixed N in the Karapoti soil . The amount of fixed NH/-N in the three urine treated and three controls was approximately 260 mg N/kg soil . This value is similar to the values reported by 1 96 Crush and Evans ( 1988) for clay fixed NH/-N in Karapoti soils sampled throughout a year, - ranging from 1 87-244 mg N/kg soil . Thus, the possibility that the unne N not accounted for had been fixed into clay minerals was minimal as the differences in clay fixed N between control and urine treated pots were much less than the amounts of N not recovered in the KAR soil ( 1 22 Ilg/g soil) . The 5MB-N was not included when accounting for urine N (Table 6 . 1 ) as it could have been double counted through the inclusion of AMN. However, the regression relationship developed between 5MB-N and AMN (Fig. 6.2 K) was negative, suggesting microbial biomass N may not be included in AMN. If 5MB-N was also included when accounting for urine N in Table 6 . 1 , the unaccounted for urine N dropped to only 1 3% - 1 8% of that added in soils collected from lowland and 4 % - 9% in soils collected from hill sites. As mentioned in Section 6.3 .2, after the first sampling (3 DAUA), soil mineral N levels in urine treated lowland soils increased somewhat with time and soil mineral N levels in urine treated, hill soils remained at the same levels as at 3 DAUA throughout the experiment. The elevated NH/-N levels in urine treated lowland soils declined rapidly (Fig. 6 . 14). At the same time the level ofN03--N increased rapidly. By 1 5 DAUA the N03--N levels in urine treated lowland soil had increased to a level approximately similar to the levels of soil NH/-N at 3 DAUA (Fig. 6. 1 4). This suggests that all the NH/-N that was present at 3 DAUA may have nitrified by 1 5 DAUA. However, the soil N03--N in urine treated lowland soils continued to increase still further indicating that some NH4 +­ N continued to be added to that present in the soil 3 DAUA. In contrast, the N03--N levels in urine treated hill soils did not increase to the same extent, with only the BF site approaching complete conversion ofNH/-N to N03--N by the end of the experiment. 1 97 Lowland Soils Karapoti (KAR) Kairanga (KAI) Manawatu (MD) 500 .,---------� 500 .------------, 500 .---------� = 400 o '" 300 � Z 200 Cl ::I. 100 15 400 III Cl 300 Z 200 g 100 O �������-.� = 400 o '" 300 !?} Z 200 Cl ::I. 100 o 10 20 30 40 50 10 20 30 40 50 o 10 20 30 40 50 DAUA DAUA DAUA ......- Ammonium ......-Nitrate -+-Ammonium ......-Nitrate �Ammonium -+-Nitrate Hill country campsite soils Baliantrae-Flat (BF) 500 .,--------­ = 400 o '" 300 Cl Z 200 g 100 O ��-_-_��� o 10 20 30 40 50 DAUA -+-Ammonium -..-- Nitrate Waipawa-Flat (WF) 500 .,----------, _ 400 "0 '" 300 Cl Z 200 Cl ::I. 100 O ����-��-� o 10 20 30 40 50 DAUA -+- Ammonium � Nitrate Hill country steep soils Baliantrae-Steep (BS) 500 ,..-------� = 400 o '" 300 !?} Z 200 g 100 O ���_-,.._�-� o 10 20 30 40 50 DAUA ___ Ammonium --r- Nitrate Waipawa-Steep (WS) 500 .,----------, = 400 o ; 300 � Z 200 Cl :::I. 100 10 20 30 40 50 DAUA -+-Ammonium -+- Nitrate Fig. 6. 1 4 The distribution of NH/-N and N03--N in urine treated soils with time after urine application. These results demonstrate that nitrification is occumng at different rates in soils collected from lowland and hill sites. Further, there was considerable variation in the rate of nitrification between soils collected from flat campsites and steep slopes in hill country. This is discussed in more detail in Section 6 .4. 1 . 1 98 Table 6.2 Quantitative comparison of NH/-N decrease and N03--N increase in urine treated soils from 3 -45 DAUA. (All values are in �g N/g soil). Soil Decrease ofNH4 + -N from Increase ofN03·-N from Quantity ofNH/-N unaccounted 3-45 DAUA 3-45 DAUA for as NO)--N KAR 245 329 -84 KAI 255 32 1 -66 MD 261 339 -78 BF 309 303 6 WF 223 252 -29 BS 1 5 5 1 69 - 14 WS 1 1 35 -24 Table 6.2 demonstrates that the decrease in NH4 + -N from 3-45 DAUA was completely accounted for by the increase in soil N03--N. In fact in all soils except BF, increases in N03--N in urine treated soils were greater than the decreases in NH/-N. The overall increase in mineral N from 3-45 DAUA was much higher in soils collected from lowland. 6.4.1 Nitrification Nitrification is a p articularly important process in grazed grassland since this is the rate limiting step in the transfer from the pool of relatively immobi l e NH/-N, derived from animal excreta and mineralisation of soil organic N, through to the vulnerable N03--N form. As pointed out III S ection 3 .2 .3 _3 , nitrification is mainly carried out b y chemoautotrophic (organisms utilize the CO2 a s their sole carbon source and gain energy from oxidation of inorganic compounds) bacteria. The major factors governing chemoautotrophic nitrification are considered to be NH4 + substrate supply, soil temperature, soil moisture and soil pH (Section 3 .2 .3 .3) . As discussed earlier, nitrification occurred at variable rates in the seven experimental soils after urine application. The experimental soils were ranked according to the rate of nitrification in Table 6.3 _ In lowland soils, nitrification was rapid up to 1 5 DAUA, and 1 99 then slowed as supplies of NH/-N became limiting (Fig. 6 . 1 1 ). Therefore, soil N03--N accumulation to 1 5 DAUA was used to calculate the daily nitrification rates in the experimental soils. As the aim was to assess the nitrification rate after urine application, the soil N03--N concentrations in the controls were deducted from the N03- -N levels in the urine treated soils. Soils collected from lowland sites had much higher nitrifi cation rates than the soils collected from hill country sites. Interestingly, daily nitrification rates o f soils collected from hill country showed considerable variability. BF soil had a greater nitrification rate than BS soil. S imilar results were observed in the field experiment described in Chapter 5. Both soils collected from Waipawa had very low nitrification rates, supporting the results observed in the field experiment described in Chapter 3 . Table 6 . 3 Nitrification rates during first 1 5 DAUA i n the experimental soils. Soil Nitrification rate ± SD (�g N03--N/g soil/day) KAI 1 8 . 3 ± 1 . 5 1 KAR 1 7. 1 ± 1 .4 5 MD 1 6 . 6 ± 0 . 8 4 BF 1 1 . 3 ± 0 . 94 BS 2 . 1 ± 0.26 WF 0 . 6 ± 0 . 8 WS 0 . 3 ± 0.0 1 As discussed in the literature review (Section 3 . 2 .3 .3), high variability of nitrification rates has previously been observed in New Zealand soils ( S arathchandra, 1 978; Steel et al., 1 980). However, the exact cause for this variation has not been clearly explained. Watson and Mills ( 1 998) conducted a similar incubation experiment by applying the same rate of NH4N03 to different soils that had a history of fertil iser inputs of 1 00, 200, 300, 400 and 500 kg Nlha/yr for the past 7 years. In that experiment, the rate of nitrification was cl early influenced by previous fertiliser history and management. The nitrification rate was greater in soils with a history of high N inputs. 200 In the current incubation experiment, the soils collected from lowland sites had received high N inputs in previous years. The KAR and KAI soils were collected from the AgResearch experimental plots that had received large amounts of urine N for the last five years. The MD soil was collected from a sheltered area of a dairy fann where large amounts of dung and urine had accumulated. Thus, soils that had a history of high N inputs had the highest nitrification rates in this experiment, which agrees with the findings of Watson and Mills ( 1 998). S imilarly, the soils coll ected from sheep campsites in hill country, where l arge amounts of N are transferred through urine and dune, had higher nitrification rates than soils collected from steep sites . It is interesting to note that Jarvis and Barraclough ( 1 99 1 ) showed evidence of "memory" effects of previous fertiliser treatments on the NH/-N : N03--N ratio in soils when a common rate of fertiliser was subsequently applied. Therefore, it can be assumed that previous N inputs to the current experimental soils could have resulted in the variation of nitrification rates during the incubation. 20 >; 1 8 Q) ra 1 6 - ;g ra 14 ... 0 s::::: I/) 12 0 :;:; C) 1 0 -ra Z () I 8 ;;: M . .: 0 6 =: z 4 z C) :::1. 2 - 0 < 0:: C u. Cl) u. Cl) ::.::: < := cc cc � � � Urine N nitrification Control N nitrification Soil Fig. 6 . l 5 Comparison of nitrification of urine N and control soil N. When nitrification rates were calculated as described above for the control soils (Fig. 6 . 1 5), a similar pattern of nitrification rate was observed. In general, the KAI, KAR, MD, BF and WF soils, which had received high N inputs, had higher nitrification rates 2 0 1 than the low N input B S and W S hill country steep soils. However, the order o f the soils based on soil nitrification rates in control pots was not exactly the same as the order after urine had been applied Interestingly the nitrification rate in the hill country steep soils was not affected much by urine addition. As these soils had been receiving very limited mineral N inputs, mainly through organic matter mineralisation, autotrophic nitrifiers might not have been able to compete with plant and N immobilising microbes. Indeed, nitrifiers are generally poor competitors with the heterotrophic biomass for NH/-N (Jansson, 1 95 8 ; Jones and Richards, 1 977). Hence, the autotrophic nitrifier population and in situ nitrifier activities might be limited. The role of heterotrophic nitrification has been discussed by several researchers (Belser, 1 979; Haynes, 1 986; Bramley, 1 989; Wragee et al., 200 1 ) . These nitrifiers use organic carbon as a C source and for energy. Kester et al. ( 1 997) reported that fungi play an important role in heterotrophic nitrification in soils with low pH. As these hil l soils contain high soluble C (Fig. 6. 1 ), heterotrophic nitrification could be more active than autotrophic nitrification. However, no studies have obtained values for the actual amounts of heterotrophic nitrification in situ, and most have merely confirmed that it is occurring in their soils (Belser, 1 979). Results from a series of experiments conducted by Bardgett et al. ( 1 997, 1 999, and 200 1 ) provide evidence that soil microbial communities of heavily grazed sites are dominated by bacteria, where�s in systems that are less intensively grazed, or completely unmanaged, fungi have a proportionately greater role. Therefore, the role of heterotrophic nitrification in hill country soils would be a useful area for future research. Jarvis et al. ( 1 989) commented that the greater CIN ratio of decaying plant material and excreta in grass/clover and grass swards with low fertiliser inputs may limit nitrification. 202 The C/N ratio can be used as a guide to the state of decomposition, or likely ease of decomposition and mineralisation of N from organic residues (McLaren and Cameron, 1 996). Soils with high C/N ratios normally contain resistant organic materials that retard decomposition and mineralisation. Previous work in hill country suggested that soils from campsites have low C/N ratios and resulting higher mineralisation of organic N than the steep soils, which normally have high C/N ratios (Ledgard et al., 1 982; Sakadevan, 1 99 1 ). In the present study, nitrification rates and soil C/N ratios showed a strong (R2=0.9) negative relationship (Fig. 6 . 1 6) . A similar relationship was obtained from the data of Steel et al. ( 1 980) in Section 3 . 2 . 3 . 3 . -;:; 25 Q) RI - "0 RI ::: 20 ... "0 y = -4.60x + 64.00 c: I/) R2 = 0.90 0 Cl 1 5 .. -RI � 10 • (J ;: .., .;: 0 5 !:: Z Z Cl • 2: 0 0 5 10 1 5 C/N ratio Fig. 6 . 1 6 Relationship between C/N ratio and nitrifi c ation rate. Continuous high N inputs to lowland sites and hill campsites could lead to a narrowing of the soil C/N ratio. This l eads to greater microbial activity and organic matter mineralisation and in turn will increase the availability of NH/-N substrate for autotrophic nitrifiers and readily available C for heterotrophic nitrifiers. In the hill country steep soils soil N is depleted through animal transfer (Gillingham and During, 1 973). Radcliffe ( 1 982) pointed out that herbage on flatter areas tended to decay faster than on steep slopes. Therefore, more dead plant shoot and root material accumulates in hill country, steep soils. Thus, low N inputs and high C accumulation results in higher C/N ratios in steep soils. These conditions lead to low quantities of NH/-N substrate for autotrophic nitrifiers and a l o w nitrification rate was observed. 203 25 25 25 y = 1 36.13. · 9.25 � 20 RZ"' O.61 "C • '" I 0 15 " � Z .� 10 0 Z '" .20 y . 4.94x · 9.81 Y = 21 .60. · 20.25 � 20 R' = 0.58 � 20 R'. 0.60 • � • � 0 1 5 '0 15 .. .. � � � Z Cs 10 CS 10 Z Z a> '" .20 .20 5 • . 0.1 0.2 0.3 (HWC/TC)'/, (SMB.cJTC)% (DOC/TC)% Fig. 6 . 1 7 Relationship between nitrification rate and the ratio of labile organic C to TC. It was demonstrated in Section 6 .3 . 1 that ratios of labile organic C to total C (SMB­ C/TC, HWC/TC and DOC/TC) are higher in soils collected from lowland sites than hill country sites indicating perhaps that they are more biologically active. Therefore, greater organic matter mineralisation would be expected in these soils. The ratios of 5MB-C/TC, HWC/TC and DOC/TC, expressed as a percentage of TC, were negatively related to C:N ratios (Fig. 6.2) and therefore these ratios had positive linear relationships with nitrification rate (Fig. 6 . 1 7) . Among the other factors governing nitrification, soil p H plays a major role a s microbial popUlation and enzymatic reactions are heavily dependent on soil pH. The common belief is that the nitrification activity is retarded at lower pH (Haynes, 1 986a; Paul and C lark, 1 988). Aluminium toxicity is suspected to be the maj or factor limiting nitrifier activity at low soil pH (Haynes, 1 9 86a) . However, some studies have revealed that nitrifying organisms tend to adapt to the existing soil pH (Bramley and White, 1 98 9 ; Islam e t al., 2000). Walker and Wickramasinghe ( 1 979) presented evidence of Nitrosospira mediated nitrification in situ in a soil with a pH as low as 4 . 1 . Also the role of heterotrophic nitrification at lower pH cannot be ruled out (Haynes, 1 986a). The relationship between nitrification rate and soil pH in the experimental control soils is i l lustrated in Fig. 6. 1 8 . 204 - � 20 Cl) RI • .. 'tI •• RI ::: .. '0 15 c rn 0 en • ;; - 10 RI Z U I � .., ''::: 0 5 � Z • Z en • ::::L 0 - 4 4.5 5 5.5 6 6.5 7 Soil pH Fig. 6 . 1 8 Relationship between nitrification rate and soil pH of the experimental soils. It can be seen that soil pH and nitrification rates are not linearly related when all the data are considered. The difference between the nitrification rates for a comparable pH was large. In Fig. 6 . 1 8 , when BF, MD, KAI and KAR soils are considered alone, there is some evidence of a linear relationship. This relationship is in agreement with the low nitrification in low pH soils. This trend is not showed by soils collected from hill country sites . The BF soil had the highest nitrification rate among the soils collected from hill country sites even though it had the lowest soil pH. 6.4.2 Relationships between resin-adsorbed N and O.5M K2S04 - extractable N As in the field experiment described in Chapter 5 , soil N was measured as resm­ adsorbed N and O . S M K2S 04-extractable N in the current incubation experiment. This enabled a check on the consistency of relationships developed between resin-adsorbed N and 2 M KCI -extractable N. 205 All the measured resin-adsorbed N and 0.5M K2S04-extractable N values during the experiment were used to develop the relationships illustrated in Fig. 6. 1 9. Thus, a total of 252 data points (7 soils x 2 treatments x 3 replicates x 6 samplings) from each method were used for the relationships. 1 800 ,--------------------------------------------, 1600 Z �1400 ] � 1 200 • y = 3.0Sx R2 = 0.78 • .a .. o Cl) "Cl ca , .5 Cl) G.I Q:: CO') N- 1000 E U 800 It) Z 600 � 400 200 o o 100 • • •• • •• •• • 200 • • 300 . , • • y = 0.62x R2 = 0.66 400 O.SM K2S04-extractable N (..,g/g soil) 500 t:. Ammonium • Nitrate -Regression line-Ammonium -Regression line-Nitrate Fig. 6 . 1 9 Relationship between resin-adsorbed N and O.5M K2S04-extractable N. As observed in Chapter 5 , both NH/-N and N03--N showed linear relationships between resin-adsorbed N and 0.5 M K2S04-extractabl e N. The relationships in this experiment (Fig. 6. 1 9) were stronger than the field experiment (Fig. 5 . 9). This could be due to both resin-adsorbed N and 0.5M K2S04-extractable N being measured from the same volume of soil in this experiment. In the field experiment, N was not measured from the same soil volume by the two methods, although both measurements were taken from the same experimental plot. In field experiment soil sampling site was spatially separated from resin sampling site to minimise soil disturbance around the resin. In addition, in the current experiment there were only two treatments and thus at some sampling times the data points were distributed at the two extreme ends of the concentration range. 1800 1600 z _ 1"00 i � 1200 � .. !3 1000 ." E 800 � � : � 600 "' - 400 200 a. Ammonium Karapoti (KAR) 100 200 300 O.W KzSO ... xtracbbl& N II>gIg 1011) • Nitrate 400 500 Kairanga (KAJ) 1000 ,------------------------------ 1600 .. 1400 � 1200 !:! 1000 "e aDO � 600 400 200 J.e-=;:.=--��--.-:.:.:.�---' 100 6 Ammonium 200 300 O.SM t(zSO.-extractable N Wo 0011) • Nitrate 400 500 -Regression Ilne-Ammonlum -Reg,e ... on IIne-Nltrate -Regrualon IIn.-Ammonlum -R.gr.S5lon IIne-Nltrata Ballantrae-Flat (BF) Ballantrae-St •• p (BS) '� r----------------------------' 1100 ,-----------------------------� '600 Z _ '400 Z � 1200 � � ,ooo � e lOO l � 600 i .: 400 200 6 Ammonium '00 y =- 4.2x Rl ", 0.9 •• • 200 300 y = 1.Ox Rl = 0.9 ... O.SM �SO ... .,:trutabl. N !I9Ig 0011) • Nitrate SOD 1600 Z .. 1400 ." ... � � 1200 � !! 1000 " e aDO l � 1II ::::a. 600 y = o.S)( y =- 1.7x R' . 0.' :. - 400 --J... ....2 200 �.;:"""""--�� �::;:���==::�� .. 8----J 100 6 Ammonium 200 300 O.SM �SO •• xtractable N (l'Oio 10111 • Nitrate 400 500 -R.gr.alon Iln.-Ammonlum -R.grealon Iln.-Nltrate -R.gresaon IIne-Ammonlum -Regr.alon Ilne�ltrat. 6 Ammonium Waipawa-Flat (WF) O.SM KzSO ... xtractable N wglg 0011) • Nltrata Walpawa-Ste.p (WS) 1100 .---------------------------� 1600 Z _ 1400 ." . .e � 1200 � � 1000 -; � lOO � � 600 Y = 0.5)( : _ 400 Y '" 2.5)( R' = 0.6 200 ��R. ' �.=Or·3�==�=.��� �--__ --__J '00 0. Ammonium 200 300 O.SM KzSO .... xtractabl. N (l'Oig oolll • Nltrat. 400 500 -Regression IIne-Ammonlum -R.gr .... on Iln.-Nltrate -Regr.SAIon IIne-Ammonlum -Regrealon IIne�ltrate Manawotu (MD) 1800 ,----------------------------, 1600 Z _ 1400 � � 1200 � � 1000 ., 5 100 � � 600 :. '= 400 y = 2.2x • • RZ = O.s. 200 J,.!��u:;=:.--���-__l 0. Ammonium 100 200 300 O.SM KzSO .... xtr.letabl. N (""'0 0011) • Nltrat. 400 SOD -R.gresslon IIne-Ammonlum -Regr.sslon IIne-Nltrata 206 Fig. 6.20 Relationships between resin-adsorbed N and O.SM K2S 04-extractable N III different soils. 207 The slopes of the regression lines were reasonably similar for both NlLt +-N and N03--N to the relationships developed in the field experiment described in Chapter 5 . The slopes were 0 . 6 for NH/-N and 3 . 1 for N03--N in the current experiment compared to slopes of 1 .0 1 for NH/-N and 2 . 4 for N03--N in the field experiment described in Chapter 5 . Similar trends were observed when the relationships were developed separately for each soil (Fig. 6.20). The relationships between resin-adsorbed NH/-N and 0 . 5 M K2S 04- extractable NlLt + -N were reasonably similar between the soils with the slope of the regression line varying only from 0 . 5 - 1 .2 . However, the relationship between resin­ adsorbed N03--N and 0 . 5 M K2S04-extractable N03--N showed greater variabi lity between soils . The slope of the regression lines between resin-adsorbed N03--N and 0 . 5 M K2S 04-extractable N03--N varied from 1 .7 - 4.2. There was no obvious explanation for this variation. 6.5 Conclusion Unlike in the field experiments, the mineral N recovery after urine application was less than 1 00%. Thus, there was no direct evidence of a priming effect after urine application. However, the quantity of mineral N in lowland soils increased with time, presumably due to a priming effect. This effect was not observed in hill country soils. The AMN and 5MB-N increased after urine application indicating immobilisation of urine N. This increase was higher in hill country soils. The mineral N changes with time after urine application varied between the soils. This variation could be explained by the variation of nitrification rates. S o i ls collected from lowland sites showed greater nitrification rates than soils collected fro m hill country. In hill country, soils collected from campsites had higher nitrification rates than the soils collected from steep slopes . The history of N application, CIN ratio of the soil and soil organic matter quality are likely factors influencing this variation in nitrification. Resin adsorbed N demonstrated similar patterns of mineral N change to those indicated by 0 . 5 M K2 S 04-extractable N. In addition, the relationships developed between resin adsorbed N and 0 . 5M K2 S04-extractable N in the current experiment were in reasonable 208 agreement with the relationships developed in the field experiment described in Chapter 5 . However, the relationships between resin adsorbed N03--N and 0 . 5 M K2 S04- extractable N03--N showed considerable variability between soils. CHAPTER 7 MODELLING THE NITROGEN CYCLE IN SHEEP-GRAZED NORTH ISLAND HILL COUNTRY PASTURE 7.1 Introduction 209 The aim of this chapter was to apply the maj or research findings from the preceding chapters, and information available in the Review of Literature (Chapter 2), to study the N cycle in sheep grazed, North Island, hill country pastures. The N cycles within landscape units of contrasting slope and aspect in hill country pasture were developed using a simple model developed in Microsoft Excel. The model was developed using the assumptions for the above ground N balance described i n Section 2 . 3 , and using the assumptions discussed in the later sections o f this Chapter for the below ground components of the N cycle. The model was constructed on yearl y time step. The maj or purpose of the model was to develop a N balance for a hill country paddock that contained steep slopes (s), easy slopes (e) and flat campsites (c). The N cycle within each of these topographic units was developed and then these individual cycles were combined to provide an overview of the N cycle in the whole paddock. 7 .2 Model inputs and development Initially this model was developed for the notional 1 ha hill country paddocks described in Section 2 . 3 . These notional north- and south-facing paddocks were assumed to have the same proportion of flat (campsite), easy and steep slopes as one of the paddocks investigated by Gillingham ( 1 978) in his detailed study of P cycling in hill country. The relative areas (A) of these 3 slope categories in the paddocks were: campsites (Ac, 1 2 . 2%), easy slopes (Ae, 45 . 5 %) and steep slopes (As, 42 .3%). 2 10 Annual pasture production (DM, kglhafyr) on each slope and aspect was as measured by Blennerhassett (2002) at the Waipawa site (Section 2 .2 . 1 and Table 7 . 1 ). S imilarly, the percentage (%) of clover in the sward ( C) was as measured by B lennerhassett (2002). N fixation (kg Nlha/yr) by legumes (NFL) was assumed to be proportional to clover growth, although the proportionality constant varied between campsites and sloping sites (Ledgard et al., 1 987) (Section 2.3 .2). Thus for each slope category, (NFL)c= ((DM)c) ((C )c/ 1 00) (0.03) (NFL)e= ((DM)e) ((C)eI l OO) (0.04) (NFL)s= ((DM)s) ((C)JI 00) (0.04) (7. 1 ) (7.2) (7.3) where the constants (0.03 and 0.04) are as suggested by Ledgard et al. ( 1 987). The amount of N taken up (kg Nlhafyr) by pasture (NP) was calculated in the model from annual pasture dry matter production (DM) and herbage N concentration (%) (RN). These data for DM production and herbage N concentration (RN) were obtained from the experimental results of Blennerhassett (2002) who had measured pasture production in summer-dry hill country at the Waipawa site. Thus for campsites, (NP)c = (DM)c ((HN)c /1 00) (7.4) S imilar calculations were done for steep and easy slopes. The measured DM production consisted of both grass and clover. Hence, to estimate the N uptake (kg N/ha/yr) from soil (NS) the amount of legume-fixed N was deducted from the plant N uptake. As an example, for campsites, (NS)c = (NP)c - (NFL)c (7.5) The percentage pasture utilisation (PU) (Section 2.3 .4) for each slope was as in Gillingham ( 1 978) and was used to estimate the amount of pasture N eaten (kg N/hafyr) by the animals (NEA). The plant N not utilised b y animals (kg N/hafyr) was considered to be added to the soil organic matter through the litter (LN). Thus for campsites, 2 1 1 (NEA)c== (NP)c ((PU)c 1 1 00) (7.6) and, (LN)c== (NP)c - (NEA)c (7.7) As mentioned in Section 2.3 .6 it was assumed that 1 0% of the N eaten by animals (NEA) was retained in animal products (NAP, kg Nlhalyr) and the rest was excreted (NEX, kg Nlha /yr) . For campsites, (NAP)c == (NEA)c ( 1 01 1 00) (NEX)c == (NEA)c - (NAP)c S imilar calculations were done for steep and easy slopes. (7 .8) (7.9) It was pointed out in Section 2 .3 .7 that the dung and urine distribution is uneven in hill country pasture paddocks. An example of estimating the dung and urine N deposited on each site was presented in Section 2.3 .7 . The same principle was used in the model to estimate dung and urine received by each site . This is explained below. The total amount of excretal N added to the paddock (TEN, kg Nlhalyr) is given by TEN == [(NEX)c (A)cIlOO]+[ (NEX)e (A)J I 00] +[ (NEX)s (A)/1 00] (7. 1 0) Dung & Urine (DU) return to a particular site (e.g. kg Nlha of campsite/yr) from the total excretal N returned to the paddock (TEN) was calculated using the percentage excreta return to each site and the percentage land area (A). As pointed out in Section 2.3 .7, Gillingham ( 1 978) measured the dung P distribution in two hill country paddocks (Section 2.3 .4). His data were used to estimate the proportion of the excreta deposited on the whole paddock that were deposited on each slope category (E). (DU)c == [((E)c /1 00) (TEN)]/ ((A)c /1 00) (DU)e == [((E)e /1 00) (TEN)]/ ((A)e /1 00) (DU)s == [((E)s /1 00) (TEN)]/ ((A)s /1 00) (7. 1 1 ) (7 . 1 2) (7. 1 3) 2 1 2 It was assumed in this model that 7 1 % of excretal N is deposited as urine N ( UN) and 29% as dung N (DN). Lambert et al. ( 1 992) reported excretal N partitioning of 65%, 71 % and 7 8 % in urine for unimproved, low P ferti liser, and high P fertiliser hill country pastures respectively. The average of these values, 7 1 % was used in these N balances. The bulk of dung N is in organic form. (Haynes & Williams, 1 99 3 ) . Thus, all N in dung was assumed to be incorporated into soil organic matter and released to soil slowly. The dung and urine N added to each site (kg N/ha/yr) was calculated as ( UN )c = (DU)c (7 1 1 1 00) (DN)c = (D U)c (291 1 00) with similar equations for the other slope categories. (7. 1 4) (7. 1 5 ) It was evident in Chapter 3 and 5 that ammonia volatilisation was a major pathway of N loss from urine patches in hill country pastures. The losses ranged from 2 1 % to 5 1 % of added urine N in the two experiments . Thus, it was assumed in the model that 3 3 % (mean o f the volatilisation losses from two experiments) o f added urine N i s lost by ammonia volatilisation from hill pasture. It is interesting to note that these substantial losses were observed from the experiments conducted in moist winter conditions. Thus , the volatilisation estimation might be conservative, a s higher volatil isation would occur m summer. The N lost through ammonia volatilisation (VN, kg Nlhalyr) was calculated as ( VN)c = ( UN)c (3311 00) (7. 1 6) with similar equations for the other slope categories. Generally, leaching hasn 't been considered as a major N loss mechanism in hill country pastures ( Sakadevan et al., 1 993). This was supported by the results of Chapter 3, which indicated very low nitrification and little potential for loss of N by leaching. However, the experiment described in Chapter 5 revealed that leaching could be a major N loss mechanism in some hill country sites, especially in stock campsites . The incubation experiment described in Chapter 6 demonstrated that nitrification is variable across the tested hil l country sites. Nitrification tended to be high in flat campsites. Soils coll ected 2 1 3 from Waipawa stock campsites for the incubation experiment showed nitrification after an initial lag period. Soils collected from the Waipawa steep slopes for the incubation experiment had very low nitrification rates. The soils collected from Ballantrae steep slopes however had reasonable nitrification rates. An attempt was made to discuss this variation in nitrification in Chapter 6 . However, finding the exact cause for this variation is a maj or research area for future work. B ased on the leaching model results discussed in Chapter 5, it was assumed that 30% of added urine N to hill country campsites is leached. No leaching was considered to occur from sloping sites in hill country, based on the low nitrification rates observed in Chapter 6. Although, it was evident in Chapter 5 that some steep slopes in hill country may lose urine N through leaching, steep sites actual l y receive little urine and thus, even if leaching is occurring the absolute amounts would be small. The amount of N lost through leaching from campsites (LN, kg N/ha/yr) was calculated as (LN)c = ( UN)c (301 1 00) (7. 1 7) To calcul ate net mineralisation in the model it was assumed that although the size of the mineral N pool may fluctuate widely from day to day, as a result of urine addition and various loss mechanisms, on an annual basis the pool size would be low and constant. In other words, inputs would equal outputs when summed over a year. With this assumption, net mineralisation N (NM, kg/ha/yr) could be calculated as the difference between the other inputs to the mineral N pool (urine, atmospheric deposition) and the losses from that pool (plant uptake, volatilisation and leaching) according to the equation. (NM)c = (NS)c - [ [( UN)c + (NAD)cJ - [( VN)c + (LN)cJ J (7. 1 8) where NAD is N added b y atmospheric deposition (kg N/ha/yr) 7.3 Model outputs 2 1 4 First, the model was used to predict the N balances for each slope category under the same conditions and assumptions described in Section 2 .3 . The input parameters are summarised in Table 7. 1 . The complete N cycles are illustrated in Fig 7 . 1 and 7 . 2 . Table 7 . 1 Data used to evaluate the model. G= measured data o f Gillingham ( 1 978), B= measured data of B lennerhassett (2002), T= Findings from this thesis. A) North aspect Input Campsites Easy slopes Steep slopes Land area (%) U 1 2 .2 45.5 42.3 P asture DM Production (kglha/yr) IS 5543 667 1 2406 N concentration in herbage (%N) IS 2.7 2 .73 2.2 Clover in herbage (% by weight of total DM) tI 5 1 1 0 P asture utilisation (%)U 79.2 82.8 76.2 Proportion of total excretal N deposited on slope category (%)" 66.6 28.6 4.8 Proportion of urine N l eached (%) J 30 0 0 Proportion of urine N volatilised (%) 33 33 33 B) South aspect Input Campsites Easy slopes Steep slopes Land area (%)" 1 2 .2 45.5 42.3 Pasture DM Production (kglha/yr) tI 5543 4574 2467 N concentration in herbage (%N) IS 2.7 2.5 2.3 C lover in herbage (% by weight of total DM) l> 5 2 2 Pasture utilisation (%)" 79.2 82.8 76.2 Proportion of total excretal N deposition on slope category (%)U 66.6 28.6 4.8 Proportion of urine N leached (%) J 30 0 0 Proportion of urine N volatilised (%) 33 33 33 In addition, the following inputs were unchanged in all model analyses described in this chapter. Non symbiotic N fixation Atmospheric N deposition Percentage of excretal N deposited in dung Percentage of excretal N deposited in urine Percentage of ingested N retained in animal products = 1 3 kg N/ha/yr = 3 kg NI ha/yr = 29% = 7 1 % = 1 0% Aim. Legume N fixation 8.3 t Plant Campsite Animal Products 1 1 .9 i deposition 3.0 149.7 _ - Dung & urine pool 106.7 � � Non.Sym. =:�� ·�":: 'J ':�: ' ·�� · · ): A�:': · · � Soil solutIon � I Or9anlc matter I Atm. deposition Leaching 104.8 legume N fixation 29.4 t Net mineralisation 9.1 Easy slopes Animal products 15.1 t 3 0 1 A�� 3�� .. .. ���.� ... ......... �::I��.�:�� .. . � Soli solutIon + ____ _ Leaching 0.0 Legume N fixation 0.0 t Leaching 0.0 Net mineralisation 122.8 Steep slopes Net mineralisation 45.1 2 1 5 Fig. 7. 1 Modelled N cycle in a hill country paddock with northerly aspect ( 1 2.2% campsite, 4 5 . 5 % easy slope, 42 .3% steep slope). All values kg Nlhalyr. Aim. deposition 3.0 Legume N fixation 8.3 � Leaching 79.4 Legume N fixation 3.7 Campsite Animal Products 1 1 .9 i - Net mineralisation 40.5 Easy slopes Animal products 9.5 Dung & urine pool 106.7 t Dung and unne pool to the site 1 AnImal 1-- 85 2 -- 42.9 A �?�: .. �:·�;'�:: ':::L?:� � . �'� � :�":' �� /······ · · · ·· · SOIL Sod solutIon Organic matter Net mineralisation 87.3 Leaching 0.0 Steep slopes legume N Animal fixation products 2.0 4.3 • t Aim. 56 7 - - 38 9 - lo the Slle �Iant Dung and urine pool =::' ��l �� �� ==�',�:� ISoil solutIon 1 4 Organoc matter SOIL Leaching 0.0 Net mineralisation 48.1 2 1 6 Fig. 7 . 2 Modelled N cycle in a hill country paddock with southerly aspect ( 1 2 .2% campsite, 4 5 . 5 % easy slope, 42.3% steep slope) . All values are kg Nlha/yr. 2 1 7 Table 7.2 Modelled N balances for individual slope categories and for the overall paddocks taking into account that campsites, easy slopes and steep slopes occupy 1 2 .2%, 45 . 5 % and 42.3% of the paddock area respectively. A) North aspect INPUT ( kg N/ha/yr) Campsite Easy slopes Steep slopes Legume N fixation 8.3 29.4 0.0 Non.Symb.fixation 1 3 .0 1 3 .0 1 3 .0 Atm.deposition 3 .0 3 .0 3 .0 Total 24.3 45 .4 1 6 .0 OUTPUT (kg N/ha/yr) Animal products 1 1 .9 1 5 . 1 4 .0 Animal transfer -385.3 79. 1 26. 1 Ammonia volatilisation 1 1 5 .3 1 3. 3 2.4 Leaching 1 04.8 0.0 0.0 Total - 1 53.4 1 07 .4 32.5 N surplus (Input-Output) kg N/ha/yr 1 77.7 -62. 1 - 1 6 .5 N Balance for overall I ha I!addock INPUT kg N/yr OUTPUT kg N/yr Campsite 3.0 Animal products 1 0.0 Easy slope 20.6 Ammonia volatilisation 2 1 . 1 Steep slope 6.8 Leaching 1 2 . 8 Total 30.4 Total 43.9 IN BALANCE �kg N/ha!�r) - 13 .51 B) S outh aspect INPUT ( kg N/ha/yr) Campsite Easy slopes Steep slopes Legume N fixation 8.3 3 .7 2 .0 Non.Symb.fixation 1 3 .0 1 3 .0 1 3 .0 Atm.deposition 3 .0 3 .0 3 .0 Total 24.3 1 9. 7 1 8 .0 OUTPUT (kg N/ha/yr) Animal products 1 1 .9 9 .5 4 .3 Animal transfer -265.9 42.3 3 1 .2 Ammonia volatilisation 87.3 1 0. 1 1 . 8 Leaching 79.4 0.0 0.0 Total -87.4 6 1 . 8 37 .3 N surplus (Input-Output) kg N/ha/yr 1 1 1 .7 -42.2 - 1 9.3 N Balance for overall 1 ha I!addock INPUT kg N/yr OUTPUT kg N/yr Campsite 3 .0 Animal products 7.6 Easy slope 8 .9 Ammonia volatilisation 1 6 .0 Steep slope 7.6 Leaching 9.7 Total 19.5 Total 33.3 IN BALANCE �kg N/hall:r) -1 3.71 2 1 8 The model suggests that for the notional 1 ha paddocks on both north and south aspects, there is an annual negative N balance of approximately 1 4 kg Nlha over the whole paddock (Table 7.2) . In other words, the N losses due to retention in animal products, leaching and volatilisation were slightly greater than the N inputs by legume N fixation, non-symbiotic fixation and atmospheric deposition. The main driving force behind these losses is N transfer from sloped areas to campsites and then high losses of N from the campsites through ammonia volatilisation and leaching. The fact that N is transferred from sloping areas to flat campsites in hill country is well documented in the literature (Gillingham and During, 1 97 3 ; Gillingham, 1 97 8 ; Ledgard et al. , 1 987; Ledgard, 200 1 ) . Nevertheless, Ledgard (200 1 ) pointed out that these significant losses of N due to transfer of excreta are ignored in many published estimates and models of N flows and balances in grazed pastoral systems. Ledgard (200 1 ) tabulated a N balance for a hill country paddock that had 45% of the area as steep slopes (>20°), 40% of the area as easy slopes ( 1 0-20°) and 1 5 % of the area as campsites. He reported an annual N input (legume + non symbiotic N fixation) of 4 3 k g N/ha and an annual output (animal products + transfer t o farm l anes and yards) o f 2 0 k g N/ha, resulting i n a N surplus o f +23 k g Nlha/yr. However, h e did not consider the N losses from campsites where large amounts of urine N are deposited. S imilar N surpluses were predicted in Chapter 2 when only the above ground components were considered. It can be seen that the model predicted marked differences in N balance between slope categories in the notional paddocks (Table 7 .2). In the steep slopes of north and south aspects, the model predicted annual negative N balances of 1 7 and 1 9 kg Nlha respectively. In the easy slopes of both aspects, the annual N balance was more negative than for the steep slopes. In the easy slopes, the predicted annual N balances were - 62 and - 42 kg N/ha for north and south aspects respectively. The model predicted annual N balances for campsites that were positive, indicating net N gain. The annual N balances for campsites were 1 78 and 1 1 2 kg Nlha for north and south aspects respectively. 2 1 9 The magnitude of any N deficit results from an interaction between N inputs via fixation etc, N uptake by pasture, the extent of pasture utilisation by animals and the redistribution of excreta between different slope categories. Thus in the modelled notional paddock, the greater N losses from easy slopes than steep slopes results from the greater pasture production on easy slopes as measured by B lennerhassett (2002) at the Waipawa site and the higher pasture utilisation on easy slopes as measured by Gillingham ( 1 978). Although N fixation as estimated from the data of Blennerhassett (2002) was greater on the easy s lopes, N losses due to animal transfer and volatilisation were proportionately higher on these slopes Other works have come to similar conclusions. Ledgard et al. ( 1 987) predicted that on steep slopes accumulation of N would be very slow or could be negative under high pasture utilisation. They stated that this was because on sloped sites, N was removed by grazing animals with little return in animal excreta (Fig.7. 1 and 7.2). In contrast, on campsites an annual net N gain was observed due to N accumulation through animal transfer. Gillingham and During ( 1 97 3 ) used dung distribution measurements to estimate that campsites received a net gain of 2 1 7 kg Nlha /yr, which was transferred from steeper sites. Finally, as mentioned in Section 2 . 3 . 8 , Ledgard (200 1 ) reported a similar N balance for hill country soils. This, gave an opportunity to compare the current model predictions for sloping sites with his estimations. The negative N balances for steep sites in the current model are 1 7 and 1 9 kg Nlha/yr for north and south aspects respectively. These prediction are in close agreement with Ledgard' s estimation (-25 kg Nlha/yr) for steep sites. However, Ledgard (200 1 ) reported a surplus of + 1 5 kg Nlha/yr for easy slopes while the current model is predicting a deficit of - 62 kg Nlha/yr for north easy slopes. This difference between the two models -needs to be explored further. If the N concentration in herbage was assumed to be 2 . 5 %, the annual plant uptake of N for the reported pasture DM production of Ledgard (200 1 ) would be 1 75 kg Nlha/yr for easy sites. This is similar to the current model ' s input of 1 82 kg Nlha/yr of plant N 220 uptake for easy slopes on the north aspect. However, there was a difference in N inputs (legume N fixation + non symbiotic fixation + atmospheric deposition) between the current model for north easy slopes (45 kg Nlha/yr) and the N input (65 kg Nlha/yr) reported by Ledgard (200 1 ) . The N removal in animal produce was 1 5 kg Nlha/yr in both Ledgard ' s estimates and in the current model. The volatilisation losses were not considered in Ledgard ' s balance while the current model estimated 1 3 kg Nlha/yr was lost through ammonia volatilisation from urine deposition on north easy slopes . The other maj or difference is the amount of N transfer. The current model ' s prediction for N loss from easy slopes by animal transfer (79 kg Nlha/yr) is twice as high as Ledgard ' s estimation (35 kg Nlha/yr). N losses through animal transfer are mainl y dependent o n the excretal distribution and pasture utilisation o n different slope categories. However, there was not suffi cient information about these components i n the model o f Ledgard (200 1 ) to identify where the differences in predicted N transfer between the 2 models occurred. Comparison of the current model and that of Ledgard (200 1 ) demonstrates that the exact level of N surplus or deficit on a particular slope category will depend on the levels of N input, pasture growth rate, pasture utilisation and excretal transfer that are assumed. What is clear from both models however, is that there are likely to be large areas of hill country where virtually all of the N added to the system by N fixation and other processes is lost and it is even likely that some areas may experience an annual N deficit. Clearly however such a situation could not persist for an extended period as soi l reserves would become depleted and D M production would decrease until losses again matched inputs. The implications of this to the sustainability of hill country farming are discussed further in Chapter 8 . 7.4 Sensitivity of the model to different conditions The results summarised i n Fig. 7 . 1 and 7.2 and Table 7 . 2 appl y to the inputs and assumptions used in the model. It would be helpful however to check the model ' s sensitivity to different combinations of model inputs. 2 2 1 Gillingham ( 1 978) described the P cycle for a second paddock with another combination of slope categories in hill country. This paddock included campsites (20. 1 %), easy slopes ( 5 5 . 7%) and steep slopes (24.2%). The measured pasture utilisations were 77.2%, 8 6 . 5 % and 8 0 . 8 % for campsites, easy slopes, and steep slopes respectively. The percentages of excreta deposited to the s ites were 7 5 % , 24% and 1 % for campsites, easy slopes, and steep slopes respectively. When these combinations were used for the model with the same model inputs for the north aspect, there wasn ' t any maj or difference in the outcome with regard to the paddock' s overall N balance (Fig. 7.3). However, within the contrasting slope categories of the paddock the model indicated that when the campsite area is increased in the paddock, the annual net N surplus in campsites (expressed as kg Nlha/yr) was decreased. S imilar results were reported b y Gillingham ( 1 978) for the P balance under the same two contrasting land area combinations. He observed a 1 1 9 kg P/ha/yr gain when the campsite area was 1 2.2% of the total area and a gain of 60.7 kg P/ha/yr in campsite areas when they made up 20. 1 % of the area. 7.4.1 Impact of excretal distribution As mentioned before, the negative N balances predicted o n sloping sites and in the paddock as a whole are mainly due to N transfer by animals . This is mainly governed by the proportion of the total excreta return to the paddock that is assumed to be deposited on each slope category. In the model to date, these proportions were as measured by Gillingham ( 1 978) in his two experimental paddocks. Thus, in the first paddock considered, 66.6% of the total excreta deposited on the paddock was assumed to be deposited on flat campsites, 2 8 .6% on easy slopes and 4.8% on steep slopes. To check the sensitivity of the model to the pattern of excretal deposition, the model was re-run with a contrasting pattern of 1 0% of the total excreta deposited on flat campsites, 5 0% on easy slopes and 40% on steep slopes . It is evident from Fig. 7.4 that the assumed pattern of excretal distribution plays a major role in deciding the negative or positive N balance on each slope, as well as in the whole paddock. 222 When the proportion of excreta deposited on campsites is very low, the N balance has changed to negative on campsites. On easy sites, when the proportion of excreta deposited was increased the negative balance was decreased by half, while on steep sites the negative balance changed to positive. When the whole paddock was considered, the negative N balance changed from - 1 4 to -3 kg N/ha/yr. Therefore, the model clearly demonstrates that any attempt to quantify the N balance in hill country paddocks should consider the excretal distribution pattern on the paddock or block concerned. 7.4.2 Impact of pasture utilisation The pasture utilisation values reported by Gillingham ( 1 978) were much higher than the pasture utilisation of 50% reported by Chapman et al. ( 1 984). Ledgard et al. ( 1 987) also assumed 50% pasture utilisation to estimate soil nitrogen accumulation on steep sites. Thus, it is interesting to check the model outcomes when the pasture utilisation is assumed to be at a 50% level. The model was also run to check the outcome under very low (25%) pasture utilisation. It can be seen from Fig. 7 . 5 , that the assumed value of pasture utilisation had a major effect on N balances in hill country. As mentioned before, when the measured (Gillingham, 1 978) high pasture uti lisation was used in the model, the overall paddock had a negative N balance. When a conservative pasture utilisation (50%) was assumed, the model predicted a slight accumulation of N over the whole paddock. A more substantial accumulation of N over the whole paddock occurred if a very low pasture utilisation (25%) was assumed. These results highlight the importance of pasture utilisation on the fate of N in hill country pasture, as considerable N accumulation can only occur when pasture utilisation is very low. As more pasture N cycles through the animal, there are higher losses through transfer, leaching and volatilisation. Campsites 12.2% Input Output AP=12 AV= 1 1 5 F=24 AT=385 L=105 • t N Balance Excretal N 66.6% + 1 78 P.u. = 79.2% Easy 4S.S% Input Output F=45 AP=15 AT=79 Excretal N 28.6% P.U. = 82.8% N Balance - 62 Steep 42.3% Input F= 1 6 Excretal N 4.8% P.U. = 76.2% Output AP=4 AT=26 AV=2 Overall paddock balance -14 Campsites Input F=24 AT=303 + Easy Input F=45 Excretal N 24% P.U. = 86.S% Steep Input F= 1 6 Excretal N 1 % P.U. = 80.8% 20.1% Output AP= 1 2 AV=96 L=87 t N Balance + 1 34 SS.7% Output AP= 1 6 AT=95 24.2% AV= 1 1 1 N Balance - 76 Output AP=4 AT=34 AV=1 Overall paddock balance -2 1 223 Fig. 7 . 3 Modelled N balance for two hill country paddocks with contrasting proportions of steep, easy and flat land. All values are kg N/ha/yr. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P.u. = Pasture utilisation, Excretal N = Percentage of excretal N deposited on each slope category i n that paddock. Campsites 1 2.2% Output Input AP=1 2 AV=1 1 5 F=24 AT=385 L=105 • t N Balance Excretal N 66.6% + 1 78 P.U. = 79.2% Easy 45.5% Input Output F=45 AP=1 5 AT=79 Excretal N 28.6% P.U. = 82.8% N Balance - 62 Steep 42.3% Input F= 1 6 Output � AP=4 AT=34 AV=1 t Excretal N 4.8% P.U.= 76.2% Overall paddock balance -14 Campsites 1 2.2% Input Output F=24 AT=33 AP=12 � AV= 1 7 L=16 t Excretal N 10% N Balance P.U. = 79.2% -53 Easy 45.5% Input Output F =45 AP= 1 5 AT=37 AV=23 J Excretal ' 50% P.U. = 82.8% N Balance -30 42.3% Output AP=4 AV=20 i Excretal N 40% P.U. = 76.2% Overall paddock balance -3 224 Fig. 7.4. N Balances for hill country paddocks with different excretal distributions. Values are kg Nlhalyr. Pasture DM production, proportion of clover in herbage and N concentration in herbage were as for the north aspect paddock in Table 7 . 1 A. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, A T= Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P . u . = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope category in that paddock. Campsites Input F=24 AT=385 Excretal N 66.6% P.U. - 79.2% 12.2% .--------=----, Output AP=12 AV=1 1 5 L=1 05 N Balance + 178 Easy 45.S% .--------=c-:--:-----, Input Output F=45 AP=15 AT=79 Campsites 12.2% Input Output F=24 AT=236 AP=8 AV=71 � Excretal N 66.6% Easy Input F=45 4S.5% N Balance + 1 17 Output AP=9 AT=47 AV=8 Campsites Input F=24 AT=1 1 8 Excretal N 66.6% P.U. = 2S% Easy Input F=45 12.2% 4S.S% 225 Output AP=4 AV=36 L=32 t N Balance + 7 1 Output AP=5 AT=24 AV=4 Excretal N 28.6% P.U. = 82.8% ...--____ ...., Excretal N 28.6% .----�::_:_--..., Excretal 28.6% N Balance P.U. = SO% N Balance P.U. = 25% N Balance + 13 Steep Input F= 1 6 Excretal N 4.8% P.U. = 76.2% 42.3% - 62 Output AP=4 AT=34 AV=1 Overall paddock balance -1 4 Steep Input F= 1 6 Excretal N 1 % 42.3% - 1 9 Output AP=3 AT=1 8 AV=2 P. U. = 50% r l -:-NC-:B:- � -:- � -nc-e--' Overall paddock balance 3 Steep Input F= 1 6 Excretal N 1 % 42.3% Output AP=1 AT=9 AV=1 P.U. = 25% I r---:-N:-:B:-- : -:-I ; -nc-e--' Overall paddock balance 1 7 Fig. 7 . 5 N Balances for hill country paddocks with different pasture utilisations. Values are kg N/ha/yr. Pasture DM production, proportion of clover in herbage and N concentration in herbage were as for the north aspect paddock in Table 7 . 1 A. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, A V = Ammonia volatilisation, L = Leaching, P.U. = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope category in that paddock. 7.4.3 Impact of soil fertility, as affected by P fertiliser addition. 226 Blennerhassett (2002) measured pasture and clover production on different land slopes under low P as well as high soil P fertility regimes. This gave an opportunity to check the effect of high P fertiliser addition on N cycling in hill soils using the current model. For the low P conditions the same data as in Table 7. 1 were used, except that pasture utilisation was assumed to be 5 0% on all slope categories. The input data for the high P paddocks are listed in Table 7.3 . Table 7 . 3 Data used to evaluate the model on a paddock with a high level o f P fertility. G = measured data of Gillingham ( 1 978), B = measured data of (B lennerhassett (2002), T = Findings from this thesis. A) North aspect Input Campsites Easy slopes Steep slopes Land area (%)U 12 .2 45.5 42.3 Pasture DM Production (kglha/yr) B 6078 9022 2740 N concentration in herbage (%N) " 3.2 2 .6 1 .7 Clover in herbage (% by weight of total DM) l> 1 14 0 Pasture utilisation (%)U 50 50 50 Proportion of total excretal N deposited o n slope category (%)" 66.6 28.6 4.8 Proportion of urine N leached (%)1 30 0 0 Proportion of urine N volatilised (%) 33 33 33 B) South aspect Input Campsites Easy slopes Steep slopes Land area (%)U * 1 2.2 45.5 42.3 Pasture DM Production (kglha/yr) " 6078 5 1 89 3785 N concentration i n herbage (%N) B 3.2 2.7 2.5 Clover i n herbage (% b y weight of total DM) " 1 1 0 7 Pasture utilisation (%)" 50 50 50 Proportion of total excretal N deposited on slope category (%)" 66.6 28.6 4.8 Proportion of urine N leached (%) 30 0 0 Proportion of urine N volatilised (%)1 33 33 33 2 2 7 Blennerhassett' s (2002) data revealed that at the Waipawa site, addition of high rates of P fertiliser increased pasture growth and clover production (compare Table 7. 1 and 7 . 3 ) . In the model, this increased clover production resulted in higher inputs through N fixation. However, the N balances illustrated in Fig 7.6 and 7 . 7 clearly reveal that under both the low P and high P fertiliser regimes the overall predicted N balances remain close to zero, indicating all the annual N inputs are lost from the system. Although adding P fertiliser increased N fixation, when this N goes through the animal cycle it increases N losses too. Low P High P Campsites Excretal N 66.6>;, P.U. = 50% Easy Excretal N 28.6% P.U. = 50% Excretal N 1 °/. 12.2'10 N Balance + 1 1 7 Campsit .. r-�"""---' Excretal N 28.6'10 N Balance I _ 19 P.U. = 50'10 Steep Input F= 1 6 P.U. = 50% I '-'N�B�""' � -nc-e--" I Excretal N 4.8% P.U. = 50% 12.2% 42.3'10 N Balance + 124 N Balance - 18 Output AP=2 AT=13 AV=2 Overall paddock balance 3 Overall paddock balance 7 Fig. 7.6 N B alances for north aspect hill country paddocks under low P and high P conditions. Values are kg Nlhalyr. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, AV = Ammonia volatilisation, L = Leaching, P.D. = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope category in that paddock Low P Campsites 12.2% Output Input F=24 AT=1 64 AP=8 AV=54 L=49 + N Balance Excretal T 66.6% + 77 P.U.= 50% Easy 45.5% r----:::--,---,-----, Input Output F=20 AP= 6 AT=25 High P Campsites Input F = 1 8 AT=226 � Excretal N 66.6% P.U. = 50% Easy Input F=37 1 2.2% Output AP= 1 0 AV=73 45.5% L=67 N Balance + 94 Output AP= 7 AT=27 AV=9 Excretal N 28.6% P.U. = 50% r--:-:-::--:----, Excretal N 28.6% N Balance I N Balance - 6 _ 1 7 P.U. = 50% Steep Input F= 1 8 Excretal N 4.8% P.V. = 50% 42.3% Output AP=3 AT=21 AV=1 Overall paddock balance -1 Steep Input F = 27 Excretal N 4.8% P.V. = 50% 42.3% Output AP=5 AT=36 AV=2 Overall paddock balance 2 228 Fig. 7 . 7 N Balances for south aspect hill country paddocks. Values are kg Nlha/yr. F = N input by legume N fixation, non symbiotic fixation and atmospheric deposition, AT = Animal transfer, AP = Animal products, A V = Ammonia volatilisation, L = Leaching, P . u . = Pasture utilisation, Excretal N = percentage of excretal N deposited on each slope category in that paddock These results are supported by the results of Blennerhassett (2002). He observed the N limitation in annual pasture yield (difference in pasture yield between N unlimited (potential) and N limited sites) was relatively even between low P and high P sites. He observed the highest pasture yield limitation in the HPNE (high P north easy) site, 229 which also showed the highest negative annual N balance in the current model (Fig. 7.4). Interestingly, his clover growth data showed that the highest annual clover growth was also recorded on the HPNE site. Thus he commented that in hill pastures, N is deficient to such an extent that even sites which have the best conditions for clover growth, are still severely restricted by N supply. These results draw attention to a key i ssue with respect to N fertility in hill country pasture. In hill country, farmers apply P fertiliser to increase clover growth to overcome N deficiency by increasing the N fixation. However as mentioned above, when this increased N in herbage goes through the animal cycle it also increases the N losses . Hence, adding P fertiliser may increase annual pasture growth but residual soil N fertility remains the same. Therefore, the pasture remains highly N responsive. Therefore, as Blennerhassett (2002) noted, the potential for N fertiliser application on hill country pasture is high and the question arises as to the value of large P fertiliser applications to overcome N deficiency through clover growth. 7.5 Improvement of efficiency of hill country N cycle A framework for considering the efficiency of nutrient cycling is outlined below. It is based on two simple propositions. .. Sustainable management of hill country pastures requires that nutrients lost from the soil/plant/animal system in animal products, or through other loss processes, must be replaced. For some nutrients such as potassium, the nutrients may be supplied by the weathering of parent materials. For nutrients such as N however, that are not present in large quantities in soil parent materials, annual inputs are required from outside the system. .. The purpose of hil l country pastoral systems is to produce animal products. Therefore a measure of the efficiency of N use might be the proportion of N inputs to the system that are converted to animal products. 2 3 0 Applying this concept t o the modelled data i n Table 7 . 2 for paddocks with campsites, easy slopes and steep slopes occupying 1 2 .2%, 4 5 . 5 % and 42 . 3 % of the paddock area respectively, gave N effi ciencies of 3 3 % and 39% for north and south facing paddocks. This calculation assumed N inputs from fixation and atmospheric deposition 3 0 .4 and 1 9 . 5 kg Nlha/yr, for north and south facing paddocks respectively. However the data in Table 7 . 2 also suggested that there was a net negative N balance of 1 3 . 5 kg Nlha/yr for the north facing paddock, and 1 3 . 7 kg Nlha/yr for the south facing paddock. These notional deficits would be made up by "n;tining" the reserves of N in the soil organic matter, and they could therefore b e considered as another input. When these inputs from soil reserves are included in the total N inputs, the estimated efficiency of N use drops to 2 3 % for both the north and south facing paddocks. Clearly, these efficiencies of N use are low, and theoretically at least, there is considerable room for improvement. This i mprovement could be achieved in a number of ways. In the model it was assumed that only 1 0% of the N ingested by grazin g animals was retained in animal products . This estimate was similar to those of a number of other workers (Gillingham, 1 97 8 ; Lambert et al., 1 982a). Any processes that increase the proportion of ingested N that is retained in animal products will immediately increase the efficiency of the N cycle. The use of forages high in condensed tannins (Waghom et al., 1 998) shows promise in this regard. Efficiency of N use is also dependent on the number of times N excreted by the animal is retained in the soil for the next cycle of uptake by pasture, and ingestion by the animal. If there are no losses, other than in animal products, then eventually all the added N will be retained by the animal and the efficiency of N use will be 1 00%. Losses from the system occur in two ways. Processes such as volatilisation, denitrification and leaching represent an immediate, direct loss from the system. In contrast, animal transfer to camp areas does not necessarily represent a total loss from the system, as the N could still potentially be taken up by plants. If however, the annual input of N to camp areas is larger than the maximum amount of N that could be taken 23 1 up by pasture, given the existing environment conditions, then some of this transfer to camp areas does represent an effective loss of N from the system. In addition, it is apparent that soil conditions in camp areas, particularly nitrification activity, are such that urine N transferred to these sites is at high risk of loss by leaching. A key therefore to improving the efficiency o f N use in hill country, is to reduce transfer to camp areas. If this could be achieved the benefits would be two-fold. Firstly, a greater proportion of the paddock would receive an input of urine N to offset the chronic N deficiency of hill country pastures . Secondly, the soil conditions o n slopes are likely to be less conducive to rapid nitrification and subsequent loss of the applied N . Subdivision of paddocks and grazing management offer some potential for minimising N transfer to stock campsites. These ideas are explored further in Chapter 8 . 2 3 2 CHAPTER 8 SUMMARY AND IMPLICATIONS FOR FUTURE RESEARCH In New Zealand, the traditional way to build up N fertility in pasture soils has been to apply P fertiliser to provide adequate fertility for legume growth. These legumes then provided N through biological nitrogen fixation. However, studies conducted in hill country pastures (Luscombe, 1 980; Ball and Field, 1 982; Lambert and C lark, 1 98 6 ; Clark and Lambert, 1 9 89; Gillingham e t al. , 1 998; Blennerhassett, 2002) have revealed that these pastures are still highly N responsive. Thus, N fertiliser application to hill country pastures has been suggested as a cost effective way of boosting production. Some scientists have gone further, suggesting precise aerial fertiliser application on hill country using GIS and GPS techniques (Gillingham et al. , 1 999). However, to make best use of these new technologies, scientists require more detailed information on N fertility and responsiveness on contrasting topographic land units . This in turn, requires an understanding of the cycling of N in hill country paddocks, and in particul ar, the role of grazing animals and their excretal returns in that cycle. The above-ground N balances developed in Chapter 2, demonstrated that N is accumulating in animal campsites that are in fl at areas of the paddock, and that N is being depleted from sloped areas. When the whole paddock is considered, these above­ ground balances predict that N is conserved in the system. However, whether this predicted N accumulation actually occurs depends on the below ground components o f the N cycle, particularly i n urine patches. Previous studies in the literature have revealed that N transformations in urine patches vary with different environmental and soil conditions. Information on N transformations under urine patches in hill country pastures is scarce. The two field experiments and the incubation experiment described in this thesis aimed to broaden the information available on N transformations in urine patches, and to explore the implications of that information to the N cycle in hill country pastures. The maj or observations are summarised below. 233 Urine application increased the mineral N availability in soil and this increase lasted about 2 to 3 months. Immediately after urine application, soi l mineral N sometimes increased by more than the amount of N added. This was due to a priming effect. The extent of the priming effect was variable across different sites, but if a priming effect is common after urine application, this may lead to accelerated losses of N from the soil organic matter, as well as from the urine N. More work is required on priming effects after urine application to identify exactly what is causing them particularly as the priming effect observed in the field could not be repeated in the laboratory incubation study. The dominant mineral N form (NH/-N or N03--N) remammg m the unne patch depends on the rate of nitrification. Soil nitrification rates were highly variable across the hill country sites, but were much lower than the nitrification rates existing in more developed, flatland soils. Within the soils from hill country sites, nitrification rates were higher in soils from flat, camp areas than in soils from steep slopes. In soils from some steep sites (eg. Waipawa), nitrification was virtually negligible. It appears that at least part of thi s variation can be explained by the previous history of N inputs to the soil. Numerous authors have observed higher nitrification rates in soils that have previously received high inputs ofN fertiliser (Jarvis et al. , 1 98 9 ; Jarvis and Barraclough, 1 99 1 ; Watson and Mills, 1 998). In the current study, soils on camp areas would have received regular inputs of urine N that is readily converted to NH4 +. In contrast, soils on steep slopes receive few inputs of N from external sources - and even the NH/-N released within the soil by mineralisation is competed for strongly by the dominant heterotrophic orgamsms. These factors together appear to result in inherently low nitrification activity. A significant effect of organic matter quality (CIN, ratio of soil labile organic C to total C) on nitrification was also observed. This variation of nitrification in hil l pastures is an important area for future research work, as nitrification i s the decisive process governing N losses from denitrification and leaching. The variation in nitrification activity across the hill country sites led to differentia l potentials for leaching. The potential for N loss by leaching from campsites in hil l 234 country i s high. When the nitrification rate is high, the leaching losses might be as large as 3 0% of added urine N. In this study, l eaching losses were not measured directly. Rather, a simple model was used to assess the likelihood of leaching being the cause of the apparent loss of N03 --N from the soil in the Ballantrae field experiment, reported in Chapter 5 . Although a modelling approach, such as that used in Chapter 5, provides a cost effective way of assessing the likely significance of leaching losses, validation measurements will be required at selected sites in future, to check on the reliability of the model. This is important, as it is apparent that the extent of leaching losses determines whether current farming systems are building up the N fertility of hill country pastures. It appeared from the experiments described in this thesis that the N loss by ammonia volati lisation from urine patches in hill country is substantial, ranging from 2 1 % to 5 1 % of added urine N . These estimated losses were surpri singly large, as the experiments were conducted under winter conditions. However, when these estimated losses were included in the calculation of overall urine N recoveries, the resulting totals were close to 1 00%, suggesting that the estimated volatilisation l o sses were reasonable. Future studies are required to confirm whether the measurement technique used to assess the extent of NH3 volatilisation was accurately calibrated for hill country pastures and also whether the use of synthetic rather than real urine may have affected NH3 volatilisation. The experiments described in this thesis confirmed the common VIew of pasture responsiveness to added urine N . U p to a three fold increase i n pasture dry matter production could be expected in urine patches. The N concentration (%) in herbage can also increase by up to 1 .7 units in pastures receiving urine, compared to controls. The urine N recoveries by pasture ranged from only 1 % to 26% (mean 1 5 %), highlighting the potential for the remaining 74-99% of added urine N to be lost from the soil. Throughout this thesis it was assumed that urine N not accounted for by the sum of mineral and mineralisable N, together with the urine N lost from soil through plant uptake, leaching and volatilisation, was immobi lised to complex organic matter in the soil. This estimated immobilisation into organic matter was large, ranging from 8-57% (mean 3 5 %) of added urine N. The percentage immobilisation was larger from the 2 3 5 lower rate of urine N addition than the higher rate and larger o n steep sites than flat sites . When the above information was incorporated into a simple model of the N cycle in hill country pastures, the crucial role of urine N in determining the overal l N balance was highlighted. The uneven urine N distribution within a paddock causes N transfer from sloped areas to fl at campsites, resulting in a net negative N balance in sloped areas and a net N gain to campsites. As the N losses by ammonia volatilisation and leaching from campsites are substantial , the overall N outputs from hill paddocks c an exceed the N inputs by N fixation. Key factors in determining whether the overall N balance in hill country p addock is positive or negative include the level of pasture utilisation on different slopes and the actual division of excreta between campsites and the remainder of the paddock. This in turn is determined by paddock topography and aspect. The N balances developed by this model, using the measured pasture production by B l ennerhassett (2002) under low and high P fertility regimes, suggested that any increase in clover growth brought about by P fertiliser addition, had l i ttle effect on the residual N fertility of the soil. This is mainly b ecause most N inputs are lost by animal transfer, and through volatilisation and leaching from urine patches. Blennerhassett (2002) noted that, after c limate and its interaction with topography, N availability was far and away the largest detenninant of pasture production . In the absence of added N, pasture production was less than half of its potential . He commented that the mineralisable N test appeared to show some potential for estimating N availability, and thereby, pasture production. However, whilst the mineralisable N test c an measure the amount of N that is potentially available to b e mineralised, it cannot detect what percentage will actually be mineralised under the environmental conditions present in the field. The in situ N measurement by ion exchange resin membrane spikes highlighted in this thesis is a possible alternative method to measure N availability. The N adsorption to resin spikes depends on the soil moisture, soil temperature and soil available N 2 3 6 concentration. Thus, reSIn spikes may provide a more realistic indication of N availability, under the conditions existing in different hill country aspects and slopes. In a series of exploratory experiments, resin spikes appeared to be able to identify real differences between soils in their ability to supply N to plants - even when traditional extraction methods using 2 M KCI or 0 . 5 M K2S04 could not differentiate between the soils . It appears that resins may be capable of detecting fluxes of N through the plant avai lable pool, as well as measuring the size of the pool at the time of sampling. This may provide information on net mineralization of N in hill country pastures. The resin spikes could also potentially be used to measure downward movement of N . To do this i t would be necessary t o design a new configuration of spike t o measure resin-adsorbed N at different depths. This thesis contributes to an understanding of the N cycle in hill country pastures in 2 ways. F irstly, it provides information on the transformations of N within urine patches. Secondly, it combines that information with other published data from a variety of sources to construct a N cycle for a notional hill country paddock. It is apparent from this and earlier work that much of the N fixed by clovers on sloping land is transferred almost immediately to animal camp areas, where varying quantities may be lost through volatilisation and leaching. This transfer and subsequent loss of N places theoretical constraints on the quantity of pasture that can sustainably be produced on sloping land. These ideas were discussed briefly in S ection 7 . 5 and are considered in more detail here. Consider two extreme, hypothetical situations. In the first, N inputs through fixation are assumed to be 50 kg Nlha/yr, pasture utilisation is 1 00%, loss in animal product is 1 0% of N ingested, N concentration in herbage is 3% and excretal N is returned more or less evenly to the grazing area, with no N losses from urine patches. In such a hypothetical situation, the only N loss from the system is in animal product. The 90% of ingested N not retained in the animal is returned to the soil, and can then be used to grow more herbage. 237 At a herbage N concentration of 3%, the 5 0 kglha of fixed N would support an initial pasture production of 1 667 kg DMlha. In the next c ycle, the 90% (45 kg/ha) of remaining N would support a further 1 500 kg DMlha of pasture production. By continuing with this approach it can be demonstrated that, under these hypothetical conditions, the input of 50 kg/ha of fixed N could produce 1 6,667 kg DMlha, before all the N was lost in animal product. If environmental constraints were such that this amount of herbage could not be grown annually, then N would accumulate in the soil. The second hypothetical situation is identical to the first, except that none of the excreted N is returned to the grazed area. In this case, as none of the N is recycled, the input of 5 0 kg/ha of fixed N would support only 1 667 kg DMlha of pasture production. It is unlikely that environmental constraints would restrict pasture production to below this level, and so N would not accumulate in the soi l . Any pasture production in excess of 1 667 kg DMlha would involve "mining" the soil reserves, and would theoretically be unsustainable. Clearly, any real life situation would fall between these two simplistic scenanos. However this approach may provide a conceptual framework within which the insertion of more realistic data would enable the likely sustainable maximum pasture production in the absence of N fertiliser to be estimated and the potential for N fertilisers to be assessed. Blennerhassett (2002) assessed the relative impacts of environment and N supply on pasture production at the Waipawa site. He demonstrated clearly that N supply was insufficient to achieve potential pasture production, even at sites where that potential production was severely constrained by environmental factors, such as moisture availability and temperature. The modelling approach of Blennerhassett (2002) provides a way of assessing potential p asture production on different slopes and aspects, in the absence o f any restriction on growth from a shortage of N. If this information is combined with the N cycling approach developed in this and previous chapters, it may be possible to identify areas in hill country where potential growth is high, but the combination of low inputs in N fixation and high losses resulting from high pasture utilisation, low excretal return and 238 losses from urine patches, mean that N supply will always be a maj or limitation to pasture growth. It is these areas where N fertilisers may have their greatest application. To illustrate this approach, data on steep and easy slopes on north and south aspects from the current study were considered within the conceptual framework outlined above. The results are presented in Table 8 . 1 . Pasture utilisation on all slopes was assumed to be 5 0% and the proportion of excreted N that was transferred off site was estimated from Fig. 7 . 1 and 7 . 2 . The N inputs through N fixation and deposition on each slope and aspect were as calculated in the previous models in Chapters 2 and 7. Given these estimated N inputs and animal transfer rates, the predicted sustainable level of pasture production ranged from 1 3 72 kg D Mlha/yr on steep northerly slopes to 6 1 64 kg DMlha/yr on easy northerly slopes. Sustainable production on southerly slopes was intermediate between these two extremes (Table 8 . 1 ). To demonstrate how these numbers were generated consider the data for steep northerly slopes. At a herbage N concentration of 3 % the input of 1 6 kg of N would support initial production of 533 kg D M . At 5 0% pasture utilisation with 1 0% of ingested N retained in animal and 72% of excreted N transferred to off site, 9.4 kg N is returned to the soil and can then be used to grow a further 3 1 5 kg DM. Similarly in the next cycle another 5 . 9 kg N would support another 1 97 kg DM. This approach was continued until the entire N input was consumed. These sustainable (in terms of N supply) production levels are very much less than the potential yi elds as estimated by Blennerhassett (2002) in the absence of any N limitation. These potential yields ranged from 9,000 kg DMlha/yr on northerly steep sites to 1 8, 000 kg DMlha/yr on no�herly easy sites. The estimated N inputs required to maintain these potential yields on a sustainable basis range from 1 05 kg N/ha/yr on northerly steep slopes to 1 96 kg N/ha/yr on northerly easy slopes. Table 8 . 1 .Comparison between sustainable levels of pasture production with current N inputs and theoretical maximum pasture production in different slope categories of hill country Slope Current Proportion Sustainable level of pasture Theoretical maximum Estimated N input required to category estimated N of ingested N that i s production with current N pasture production if N is non maintain theoretical maximum inputs* transferred off site inputs l imiting* * production kg N/ha kg DM/ha/yr kg DM/ha/yr kg N/ha/yr N E 6 7 0 . 5 8 6 1 64 1 8000 1 96 N S 1 6 0 . 72 1 3 72 9000 1 05 SE 3 7 0 . 5 0 3 5 3 6 1 2000 1 26 S S 27 0 . 8 0 2220 1 1 000 1 3 4 * Values from Fig. 7 . 6 and 7 . 7 * * measured data o f Blennerhassett (2002) 240 The current N inputs are clearly very much smaller than those required to sustain maximum pasture production. More work is required to quantify inputs through N fixation, to determine the extent to which optimising soil fertility for legume growth can boost N inputs towards the levels needed to enable near-maximum pasture production . Calculations such as these, that are based on annual balances, do not take into account seasonal nutrient transformations. Thus, even if annual N fixation could be boosted to the levels indicated as being required in Table 8 . 1 , environmental constraints (such as soil temperature) would mean that mineralisation could not provide N at a sufficient rate to allow maximum pasture growth at some times of the year. Future research should focus on the framework for the N cycle presented in Chapters 2, 7 and 8 . As has been demonstrated in this thesis, much information can be derived from related studies in the literature, but there will be a need for focus sed experiments to provide information on specific issues. Of particular importance are quantitative estimates of N fixation, animal transfer and losses of excretal N . 24 1 REFERENCES Agrawal, A . S . , Singh, B . R. and Kanehiro, Y . 1 97 1 . Ionic effects o f salts on mineral nitrogen release in an allophonic soil. Soil Sci. S oc. Am. Proc. 3 5 : 454-457. Ardakani, M.S., Rehbock, J.T. and McLaren, A.D. 1 97 3 . Oxidation o f nitrite to nitrate in a soil column. Soil Sci. Soc. Am. Proc . 3 7 : 5 3-56. Ardakani, M . S . , Schulz, R .K. and McLaren, A . D . 1 974. A kinetic study of ammonium and nitrite oxidation in a soil field plots. S oil Sci. Soc. Am. Proc. 3 8 : 273-277. Baber, H.L. 1 97 8 . A study of some nitrate and phosphate problems in New Zealand Agriculture. PhD Thesis, University ofWaikato, Hamilton. Ball, P.R., Inglis, lA.H. and Mauger, 1 . H . 1 976. Tactical application of fertiliser nitrogen to offset seasonal feed shortage on heavily stocked sheep farm in southern Hawkes ' B ay. Proc. N.Z. Grassland Assoc. 3 7 : 1 66- 1 8 1 . Ball, R., Keeney, D.R., Theobald, P.W. and Nes, P . 1 979. Nitrogen b alances in urine ­ affected areas of a New Zealand pasture. Agron. 1. 7 1 : 3 09-3 1 4. Ball, P.R. and Field, T.R.O. 1 982. Responses to nitrogen as affected by pasture characteristics, season and grazing management. In : Nitrogen Fertilisers in New Zealand Agriculture. Ed. Lynch, P . B . pp. 45 -64. New Zealand Institute of Agricultural Sciences, Wellington, New Zealand. Ball, P .R . , Luscombe, P.c. and Grant, D.A. 1 98 2 . Nitrogen on hill country. In : Nitrogen Fertilisers in New Zealand Agriculture. Ed. Lynch, P . B . pp. 1 3 3 - 1 47 . New Zealand Institute of Agricultural Sciences, Wellington, New Zealand. Ball, P . R. and Keeney, D.R. 1 98 3 . Nitrogen losses from urine-affected areas of a New Zealand pasture, under contrasting seasonal conditions. pp. 3 4 1 -344. Proc. XIV Int. Grassland Congr. Lexington, Kentucky, U . S .A. 242 Ball, P . R. and Ryden, l . C . 1 984. Nitrogen relationships m intensively managed temperate grasslands. Plant Soil 76 : 23-3 3 . Ball, P . R. and Tillman, R.W. 1 994. Efficient use o f nutrients i n the intensive pastoral farming industry in New Zealand. In : The Efficient Use of Fertilisers in a Changing Environment: Reconciling productivity and sustainability. Eds. Currie, L.D. and Loganathan, P. pp. 96- 1 1 7 . Occasional Report No. 7. Fertilizer and Lime Research Centre, Massey University, Palmerston North. B ardgett, R . D . , Leemans, D.K., Cook, R. and Hobbs, P.J. 1 997. Seasonality of the soil biota grazed and ungrazed hill grasslands. Soil BioI. Biochem. 29: 1 285- 1 294. Bardgett, R . D . , Lovell, R., Hobbs, P. and larvis, S . c . 1 999. Seasonal changes in soil microbial communities along a fertility gradient of temperate grasslands. Soil BioI . Biochem. 3 1 : 1 02 1 - 1 03 0 . B ardgett, R.D., lones, A . c . , lones, D . L., Kemmitt, S . J . , Cook, R . and Hobbs, P.J . 200 1 . Soil microbial community patterns related to the history and intensity of grazing in sub­ montane ecosystems. Soil BioI. Biochem. 3 3 : 1 65 3 - 1 664. Barrow, N.l. and Lamboume, L.1. 1 962. Partition of excreted nitrogen, sulphur and phosphorus between the faeces and urine of sheep being fed pasture. Aust. 1 . Agric. Res. l 3 : 46 1 -47 1 . BDH Chemicals Ltd. 1 98 1 . Ion exchange resms. Broom Road, Poo1e BH 1 2 4NN England. Belser, L.W. 1 979. Population ecology of nitrifying bacteria. Ann . Rev. Microbiol. 3 3 : 3 09-3 3 3 . B inkley, D . and Matson, P . 1 983 . Ion exchange resin bag method for assessing forest soil nitrogen availability. Soil Sci. Soc. Am. 1. 47: 1 050- 1 05 2 . 243 Binkley, D . 1 984. Ion exchange resin bags : factors affecting estimates of nitrogen availability. Soil Sci. Soc. Am. J. 4 8 : 1 1 8 1 - 1 1 84. Blakemore, L . e . , Searle, P.L. and Daly, B .K. 1 9 8 7 . Methods for chemical analysis of soils. N . Z . Soil Bureau Sci. Rep. No. 80, Lower Hutt. Blennerhassett, J . D . 2002. Pasture growth constraints on dry steep east coast hill country. PhD Thesis, Massey University, New Zealand. Bramley, R.G.V. 1 9 89. Quantitative analysis of the variability in the activity of nitrifying organisms in a soil under pasture, PhD Thesis, Massey University, New Zealand. Bramley, R.G.V. and White, R. E. 1 989. The effects of pH, liming, moisture and temperature on the activity of nitrifiers in a soil under pasture. Aust. J . Soil Res . 2 7 : 7 1 1 -724. Bramley, R.G.V. and White, R.E. 1 990. The variability of nitrifying activity in field soils. P lant Soil 1 26 : 203-208. Brernner, 1.M. and Mulvaney, R . L. 1 97 8 . Urease activity in soils. In : Soil Enzymes. Ed. Bums, R.G. pp. 1 49- 1 96. Academic Press, London. Broadbent, F.E. and Tyler, K. B . 1 962 . Laboratory and greenhouse investigations of nitrogen immobilisation. Soil Sci. Soc. Am . Proc. 26: 459-469. Broadbent, F . E . 1 965. E ffect of fertilizer nitrogen on the release of soil nitrogen. Soil Sci. Soc. Am. Proc. 29: 692-696. Broadbent, F.E. and Nakashima, T. 1 97 1 . Effect of added salts on nitrogen mineralisation in three California soils. Soil Sci. Soc. Am. Proc. 3 5 : 45 7-460. Brock, J . L. 1 97 3 . Growth and nitrogen fixation of pure stands of three pasture legumes with high/low phosphate. N .Z. 1. Agric. Res. 1 6 : 4 8 3 -49 1 . 244 Brookes, P . c . , Landman, G., Pruden, G . and Jenkinson, D . S . 1 98 5 . Chloroform fumigation and the release of soil nitrogen: a rapid direct extraction method for measuring microbial biomass nitrogen in soil. S oil BioI. Biochem. 1 7 : 83 7-842 . Burns, LG. 1 97 5 . An equation to predict the leaching of surface applied nitrate. J. Agric . Sci . 8 5 : 443 -454 . Cabrera, M.L. and Beare, M.H. 1 99 3 . Alkaline persulfate oxidation for determining total nitrogen in microbial biomass extracts . Soil Sci. Soc. Am. J. 5 7 : 1 007- 1 0 1 2 . Cain, M . L . , Subler, S . , Evans, l P . and Fortin, M J . 1 999. S ampling spatial and temporal variation in soil nitrogen availability. Oecologia 1 1 8 : 397-404. Cameron, K.c. and Haynes, RJ. 1 9 86. Retention and movement of nitrogen in soils. In : Mineral Nitrogen in the Plant Soil System. pp . 1 66-220. Academic Press, Inc, Orlando , Florida, USA. Campbell, A.G. 1 95 1 . A study of principles and methods of pasture improvement on New Zealand hill land, with particular reference to the effects of climatic, biotic, and anthropogenic influences upon the sward. PhD Thesis. Massey University. Carlyle, J . c . and Ma1colm D . C . 1 98 6 . The use of ion exchange resin bags to assess N availability beneath pure spruce and larch + spruce stands growing on a deep peat soil . Plant Soil 93 : 1 23- 1 27. Carran, R.A., Ball, P.R., Theobald, P.W. and CoIl ins, M.E.G. 1 9 82. Soil nitrogen balances in urine affected areas under two moisture regimes in Southland. N.Z. l Exp. Agric. 1 0 : 3 77-3 8 1 . Carran, R.A., Theobald, P.W. and Evans, J . P . 1 99 5 . Emission of nitrous oxide from some grazed pasture soils in New Zealand. Aust. 1 . Soil Res . 3 3 : 34 1 -52. 245 Carran, R.A., Theobald, P.W. and Brock, l L . 2000. Novel paSSIve samplers for assessing NH3 volati lisation from difficult sites. In : Soil 2000: New Horizons for a New Century. Australian and New Zealand Second Joint Soils Conference Volume 2: Oral papers . Eds. Adams, J. A. and. Metherell, A.K. pp. 47-4 8 . 3-8 December 2000. Lincoln University. New Zealand Society of Soil Science. Chapman, D . F . , Clark, D.A., Land, c.A. and Dymock, N. 1 984. Leaf and tiller or stolon death of Lolium perenne, Agrostis spp. and Trifolium repens in set stocked and rotationally grazed hill pastures. N.Z. J. Agric. Res. 27: 3 03-3 1 2 . Clark, D .A . , Chapman, D.F., Land, C.A. and Dymock, N. 1 9 84. Defoliation of Lolium perenn e and Agrostis spp. tillers, and Trifolium repens stolons in set-stocked and rotationall y grazed hill pastures. N.Z. J. Agric . Res. 27: 289-30 1 . Clark, D . A . and Lambert, M.G. 1 98 9 . The use of fertiliser nitrogen for i ncreased animal production from New Zealand hill pastures . pp. 1 269- 1 27 0 . XVI Int. Grassland Congr. Nice, France. Crush, J . R. and Evans, lP.M. 1 98 8 . Clay fixed ammonium levels in four Manawatu pasture soils. N.Z. l Agric. Res. 3 1 : 7 1 -7 5 . Dale, W .R. 1 96 1 . Some effects of sheep urine o n pasture. N.Z. Grassland Assoc. Proc. 1 3 : 1 1 8 - 1 24. Deluca, T . H . and Keeney, D .R. 1 993 . S oluble organics and extractable nitrogen III paired prairie and cultivated soils of central Iowa. Soil Sci. 1 5 5 : 2 1 9-22 7 . Distefano, J . a n d Gholz, H. 1 986. A proposed u s e o f ion exchange resin t o measure nitrogen mineralisation and nitrification in intact soil cores. Commun. Soil Sci . Plant Anal. 1 7 : 989-998. Doak, B.W. 1 95 2 . Some chemical changes in the nitrogenous constituents of urine when voided on pasture. J. Agric. Sci. 42 : 1 62- 1 7 1 . 246 Fraser, P . M . , Cameron, K.C. and Sherlock, R.R. 1 994. Lysimeter study of the fate of nitrogen in animal urine returns to irrigated pasture. European J. S oil Sci. 45 : 439-447. Freney, J.R. and Black, A.S. 1 9 8 8 . Importance of ammonia volatil isation as a loss process . In : Advances in Nitrogen Cycling in Agricultural Ecosystems. Ed. Wilson, J.R. CAB International. Wallingford. Ghani, A . , Sarathchandra, u. , Perrott, K.W . , Singleton, P. , Wardle, D . , Dexter, B.M., and Ledgard, S .F. 1 999. Are soil microbial and biochemical indicators sensitive to pastoral soil management. In : Best Soil Management Practises for Production. Eds. Currie, L . D . , Hedley, MJ., Home, D J . and Loganathan, P. pp. 5 3- 62. Occasional Report No. 1 2 . Fertiliser and Lime Research Centre, Massey University, Palmerston North . Ghani, A . , Dexter, M., Sarathchandra, u. , Perrott, K.W. and Singleton, P. 2000. Assessment of extractable hot water carbon as an indicator of soil quality on soils under long term pastoral, cropping, market gardening and native vegetation. In : Soil 2000: New Horizons for a New Century. Austral ian and New Zealand S econd Joint S oils Conference Volume 2: Oral papers. Eds. Adams, lA. and. Metherell, A.K. pp. 1 1 9- 1 2 0. 3-8 December 2000, Lincoln University. New Zealand Society of Soil Science. Gillingham, A.G. and During, C. 1 97 3 . Pasture production and transfer of fertility within a long-established hill pasture. N.Z. l Exp. Agric. 1 : 227-23 2 . Gillingham, A.G. 1 974. Influence o f physical factors o n pasture growth on hill country. Proc. N . Z . Grassland Assoc. 3 5 : 77- 8 5 . Gill ingham, A . G . 1 978. Phosphorus cycling i n grazed, steep hill country. PhD Thesis, Massy University, Palmerston North, New Zealand. Gillingham, A., Syers, lK. and Gregg, P.E . H . 1 9 80. Phosphate uptake and return in grazed steep hill pastures 2. Above-ground components of the phosphorus cycle. N.Z. J . Agric. Res. 2 3 : 3 23-330. 247 Gillingham, A.G. 1 9 82. Topographic and measurement effects on dung distribution by grazing sheep. Proc. N.Z. Grassland Assoc. 43 : 1 6 1 - 1 70. Gillingham, A.G., Gray, M . H. and Smith, D.R. 1 99 8 . Pasture responses to phosphorus and nitrogen fertilisers on dry hill country. Proc. N.Z. Grassland Assoc. 60: 1 3 5 - 1 40 . Gillingham, A . G . , Maber, J. , Morton, 1. and Tuohy, M. 1 99 9 . Precise aerial fertiliser application on hill country. Proc. N.Z. Grassland Assoc. 6 1 : 22 1 -226. Gilmour, 1.T. 1 984. The effects of soil properties on nitrification and nitrification inhibition. Soil Sci. Soc. Am . J. 4 8 : 1 262-66. Grant, D.A. and Brock, J.L. 1 974. A survey of pasture composition in relation to soils and topography on a hill country farm in the southern Ruahine range, New Zealand. N.Z. J. Exp. Agric. 2: 243-2 5 0 . Grant, D . A . and Lambert, M . G . 1 979. Nitrogen fixation in pasture. 5 . Unimproved North Island hill country, "Ballantrae". N.Z. 1. Exp . Agric. 7: 1 9-22. Harland, C.E. 1 994. Ion Exchange. Theory and Practice. Royal S ociety of Chemistry, Turpin Distribution Services Limited. UK. Haynes, R.1. 1 98 6a. Nitrification. In : Mineral Nitrogen in the Plant Soil System. pp. 1 27- 1 5 7 . Academic Press, Inc, Orlando, Florida, USA. Haynes, R.1. 1 986b. The Decomposition Process : Mineralisation, immobilisation, humus formation, and degradation. In : Mineral Nitrogen in the Plant Soil System. pp. 52- 1 09 . Academic Press, Inc, Orlando, Florida, USA. Haynes, R.1. 1 986c. Uptake and assimilation of mineral nitrogen by plants. In : Mineral Nitrogen in the Plant Soil System. pp. 303-3 5 8 . Academic Press, Inc, Orlando, F lorida, USA. 248 Haynes, R.J. and Sherlock, R . R. 1 98 6 . Gaseous losses of nitrogen. In : M ineral Nitrogen in the Plant Soil System. pp . 242-3 02 . Academic Press, Inc, Orlando, Florida, USA. Haynes, R.J. and Williams, P . H . 1 992 . Changes in soil solution composition and pH in urine-affected areas of pasture. 1 . Soil Sci. 43 : 323- 3 3 4 . Haynes, R.J. and Williams, P . H . 1 99 3 . Nutrient cycling and soil fertility i n the grazed pasture ecosystem. Adv. Agron. 49 : 1 1 9- 1 99 . Haynes, R.J . and Williams, P . H . 1 999. Influence o f stock camping behaviour o n the soil microbiological and biochemical properties of grazed pastoral soils. BioI. Fertil . Soils 28: 2 5 3 - 2 5 8 . Haynes, R . J . 2 0 0 0 . Labile organic matter as a n indicator of organic matter quality i n arable and pastoral soils i n New Zealand. Soil Boil. Biochem. 3 2 : 2 1 1 -2 1 9. Heanes, D.L. 1 984. Determination of total organic C in soils by an improved chromic acid digestion and spectrophotometric procedure. Commun. Soil Sci. Plant Anal. 1 5 : 1 1 9 1 - 1 2 1 3 . Hewitt, A.E. 1 99 8 . New Zealand Soil C lassification. Land Care Research Science Series No. I . Manaaki Whenua Press, Lincoln. Hoglund, 1 . H . , Crush, 1.R., Brock, J . L., Ball, R. and C arran, R .A. 1 979. Nitrogen fixation in pasture. 1 2. General discussion. N.Z. 1. Exp . Agric . 7 : 45-5 l . Holland, P.T. and During, C . 1 977. Movement of nitrate-N and transformations of urea­ N under field conditions. N.Z. J . Agric. Res. 2 0 : 479- 4 8 8 . Huang, W . Z . and Schoenau, J .J . 1 996. Microsite assessment o f forest soil nitrogen, phosphorus and potassium supply rates in field using ion exchange membranes. Commun. Soil Sci. Plant Anal. 27: 2909-29 1 5 . 249 Hubner, c., Redi, G. and Wurst, F. 1 99 1 . In Situ methodology for studing N­ mineralisation in soils using anion exchange resin s . Soil BioI. Biochem. 23 : 70 1 -702. Islam, A., White, R.E. and Chen, D.L. 2000. In : Soil 2000: Horizons for a New Century. Australian and New Zealand Second loint Soils Conference Volume 2: Oral papers. Eds. Adams, l.A. and Metherell, A.K. pp. 1 5 3 - 1 54. 3 -8 December 2000, Lincoln University. New Zealand Society of Soil S cience. Jansson, S . L. 1 95 8 . Tracer studies on nitrogen transformations in soil with special attention to mineralisation-immobilisation relationships . Ann. R. Agric. ColI. Sweden 24 : 1 0 1 -3 6 1 . Jarvis, S . C . , Macduff, l.R., Wi lliams, J.R. and Hatch, D J . 1 9 89. Bal ances of forms of mineral N in grazed grassland soils impact of N losses. XVI Int. Grassland Congr. Nice, France. Jarvis, S .c . and Barrac1ough, D. 1 99 1 . Variation in mineral nitrogen under grazed grassland swards. Plant Soil 1 3 8 : 1 77- 1 8 8 . Jarvis, S . c . 1 999. Accounting for nutrients i n grassland: challenges and needs . In : Accounting for Nutrients: a Challenge for Grassland Farmers in the 2 1 st Century. pp. 3 - 1 2 . Proceedings of the British Grassland Society Conference, Great Malvern, UK, 22-23 November. British Gras sland Society (BGS), Reading, UK. Jenkinson, D. S . , Fox, R.H. and Rayner, 1 . H . 1 98 5 . Interactions between fertiliser nitrogen and soil nitrogen - the so-called "priming effect". 1. Soil Sci . 3 6 : 425 -444. Joblin, A.D.H. 1 9 8 3 . Production potentials of hill country in the Northern South Island. Proc. N.Z. Grassland Assoc. 44: 1 08- 1 1 9 . J ones, J . M . and Richards, B .N. 1 977. Effect of reforestation on turnover of 1 5N -labelled nitrate and ammonium in relation to changes in soil microflora. Soil BioI. Biochem. 9 : 3 8 3 -3 92. 2 5 0 Jungk, A. and C laassen, N. 1 997. I o n diffusion in the soil root system. Adv. Agron. 6 1 : 54- 1 07. Keeney, D.R. and MacGregor, A.N. 1 978. Short term cycling of 1 5N_urea in a ryegrass­ white clover pasture. N.z. 1. Agric. Res. 2 1 : 443 -448. Kester, R.A., De Boer, W. and Lanbroek, HJ. 1 99 7 . Production of NO and N20 by pure cultures of nitrifying and denitrifying bacteria during changes in aeration. Appl. Environ . Microbiol . 63 : 3 872-3 877. Lambert, M.G. 1 97 3 . An investigation of some differences between aspects in hill country. MSc Thesis, Massey University, New Zealand. Lambert, M.G. Renton, S .W. and Grant, D.A. 1 9 82a. Nitrogen balance studies in some North Island hill pastures. In : Nitrogen Balances in New Zealand Ecosystems. Ed. Gandar, P . W. pp. 3 5 -39. D . S . I . R . , Palmerston North, New Zealand. Lambert, M . G . , Luscombe, P . C . and Clark, D.A. 1 982b. Soil fertility and hill country production. Proc. N.Z. Grassland Assoc. 43 : 1 5 3 - 1 60 . Lambert, M . G . and Clark, D.A. 1 986. Effects of late-auturnn nitrogen application o n hill country pastures and sheep production. Proc. N.Z. Grassland Assoc. 47: 2 1 1 -2 1 5 . Lambert, M . G . , Barker, DJ., Mackay, A.D. and Springett, J. 1 996. Biophysical indicators of sustainability of North Island hill pasture systems . Proc. N.Z. Grassland Assoc. 5 7 : 3 1 -36. Ledgard, S.F., Steel , K.W. and Saunders, W.M.H. 1 982a. Effects o f cow urine and its major constituents on pasture properties. N.Z. 1. Agric. Res. 2 5 : 6 1 -6 8 . Ledgard, S .F. , Sheath, G.W. and Gillingham, A.G. 1 982b. Influence o f some soil and pasture components on the growth of hill country pastures. 1 . Winter and spring production. N.Z. J . Exp. Agric. 1 0 : 2 3 9 -244 . 2 5 1 Ledgard, S . F . , Brier, G.J. and Littler, R.A. 1 987. Legume production and nitrogen fixation in hill pasture communities. N.Z. J. Agric. Res. 3 0 : 4 1 3-42 1 . Ledgard, S . F . and Steele, K.W. 1 99 2 . Biological nitrogen fixation In mixed legume/grass pastures. P l ant Soil 1 4 1 : 1 3 7- 1 5 3 . Ledgard, S . F . 200 l . Nitrogen cycling i n low input legume-based agriculture, with emphasis on legume/grass pastures. Plant S oil 228: 43-5 9 . Lee, C . H . 1 97 1 . Improvement of hill country pasture. Dissertation, Post Graduate Diploma in Agricultural Sciences. Massey University, Palmerston North, N ew Zealand. Leon, M., Laine, P. , Ourry, A. and Boucaud, 1. 1 99 5 . Increased uptake of native soil nitrogen by roots of Lolium multiflorum Lam. after nitrogen fertilisation is explained by a stimulation of the uptake process itself. Plant Soil 1 73 : 1 97-203. Levy, E.B. 1 97 0 . Grass l ands of New Zealand. Third edition; Government printer, Wellington. Littell, R.e., Henry, P.R. and Ammerman, e . B . 1 99 8 . Statistical analysis of repeated measures data using SAS procedures. J. Anim. Sci. 76 : 1 2 1 6- 1 23 1 . Lovell, R.D. and Jarvis, S . e . 1 996. Effects of urine on soil microbial biomass, methanogenesis, nitrification and denitri fication in grassland soils. Plant Soil 1 86 : 265- 273 . Luo, 1 . , Tillman, R.W. and Ball, P.R. 2000. Nitrogen loss through denitrification in a soil under pasture in New Zealand. Soil BioI. Biochem. 3 2 : 497-509. Luscombe, P.e. 1 9 80. Nitrogen fertilizer responses on hill country pastures. Proc. N.Z. Grassland Assoc. 4 1 : 1 5 5 - 1 62 . Lyster, S . , Morgan, M . A . and O ' Toole, P . 1 980. Ammonia volatilisation from soils fertilized with urea and ammonium nitrate. J. Life Sci. 1 : 1 67- 1 76 . 2 5 2 Macfarlane, M.J. and Sheath, G . W . 1 984. Clover-what types for dry hill country. Proc. N.Z. Grassland Assoc. 4 5 : 1 40- 1 50 . Machado, A S .D.P. 1 994. Relationship between soil fertil ity and the components of and seasonality of forage supply of a hill pasture. M aster of Agricultural Science in S oil Science Thesis. Massey University. New Zealand. Mackay, A D . 1 990. Comparison of phosphate fertilisers III a grazed hill country pasture. Proc. N.Z. Grassland Assoc. 5 1 : 97- 1 00. Mackay, A.D., Wedderbum, M.E. and Lambert, M . G. 1 99 3 . Sustainable management of hill land. Proc. N.Z. Grassland Assoc. 5 5 : 1 70- 1 76. Malhi, S . S . and McGill, W.B. 1 982. Nitrification in three Alberta soils: effect of temperature, moisture and substrate concentration. Soil BioI . Biochem. 1 4 : 393-399. Mauger, J.H. 1 979. Hill country - our life blood. N .Z. Agric . Sci. 1 3 : 1 -9 . McKenzie, A.H. and Wallace, H . S . 1 95 4 . The kj eldhal determination o f nitrogen. A critical study of digestion conditions-temperature, catalyst, and oxidising agent. Aust. J . Chem. 7 : 5 5-70 . McLaren, A . D . 1 97 1 . Kinetics of nitrification i n soil : Growth of the nitrifiers. S o i l S c i . S o c . A m . Proc. 3 5 : 9 1 -9 5 . McLaren, R . G . and Cameron, K.C. 1 996. Soil orgamc matter In : Soil S cience, Sustainable Production and Environmental Protection, Oxford University Press. McLaughlin, M J . , Lancaster, P .A., S ale, P .W.G., Uren, N . C . and Peverill, K.I . 1 99 3 . Use of cation/anion exchange membranes for multi-element testing o f acidic soils . Plant Soil 1 5 6 : 223-226. Moat, AG. and Foster, J.W. 1 999. Microbial Physiology. Third edition. John Wiley and Sons, Inc. 2 5 3 Moir, 1 .L. 2 0 0 0 . Investigations into the influence o f fertiliser history a n d climate regime on the soil fertility, soil quality and pasture production of Wairarapa hill soils. PhD Thesis, Massey University, New Zealand. Molina, J.A.E., Hadas, A. and Clapp, C. E. 1 990. Computer simulation of nitrogen turnover in soil and priming effect. Soil B ioI. Biochem. 22: 3 49-3 5 3 . Monaghan, R M . , Cameron, K . c . and McLay, C . D .A. 1 98 9 . Leaching losses o f nitrogen from sheep urine patches. N .Z. J . Agric. Res. 3 2 : 23 7-244. Monaghan, R.M. and Barraclough, D. 1 992. Some chemical and physical factors affecting the rate and dynamics of nitrification in urine-affected soil. Plant Soil 1 43 : 1 1 - 1 8 . Monaghan, RM. and Barrac1ough, D. 1 99 3 . Nitrous oxide and dinitrogen emissions from urine-affected soil under controlled conditions. Plant S o i l 1 5 1 : 1 27 - 1 3 8 . Morton, J . D . , Korte, C.J., Smith, D . R , Watt, B . D . and Smith, R G . 1 99 3 . Nitrogen use and farm performance on Wairarapa sheep and beef farms. Proc. N.Z. Grassland Assoc. 5 5 : 5 3 - 5 7 . Muller, C. 1 99 5 . Nitrous oxide emission from intensive grassland i n Canterbury, New Zealand. PhD Thesis. Lincoln University. New Zealand. Murphy, M . D . and Riley, 1 .P . 1 962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta. 2 7 : 3 1 -3 6 . Norton, J. 1 999. Nitrification. In : Handbook of Soil Science. E d . M a1com. E. Summer. pp. 1 48- 1 8 1 . CRC Press, USA. NZLRI. 1 979. New Zealand Land Resource Inventory, worksheets and extended legend. Produced for the National Water and Soil Conservation Organisation by the Water and Soil Division, Ministry of Works and Development. Government Printer, Wellington, New Zealand. 2 5 4 Paul, E .A. and Clark, F . E . 1 98 8 . Transformation of nitrogen between the organic and inorganic phase and to nitrate. In : S oil Microbial B iochemistry. Academic Press, Inc . S an Diego, California. P ilbeam, D.l and Kirkby, E.A. 1 99 2 . S ome aspects of the utilization of nitrate and ammonium by plants. In : Nitrogen Metabolism of Plants. Eds. Mengel, K. and Pilbeam, D J . C larendon Press. Oxford. Qian, P.Y., Schoenau, J J . and Huang, W.Z. 1 992. Use of ion exchange membranes in routine soil testing. Commun. Soil Plant Anal . 23 : 1 79 1 - 1 804. Qian, P.Y. and Schoenau, 1.1. 1 995. Assessing nitrogen mineralisation from soil organic matter using anion exchange membranes. Fert. Res. 40: 1 43 - 1 48 . Quin, B . F . 1 977. The fate of sheep urine-nitrogen on surface-irrigated pasture III Canterbury. New Zealand Soil News 2 5 : 4 . Quin, B . F . 1 979. The effects of surface irrigation o n drainage and groundwater quality. In : Groundwater: a symposium on current research in New Zealand. Ed. Noonan, M J . p p . 1 3 9- 1 45 . Technical Publication No . 2 . Lincoln College Department of Agricultural Microbiology. Quin, B . F . 1 982. The influence of grazing animals on nitrogen balances. In : Nitrogen B alances in New Zealand Ecosystems. Ed. Gandar, P . W . pp . 95- 1 02 . D . S . I. R, Palmerston North, New Zealand. Radc1iffe, lE. 1 982. Effects of aspect and topography on pasture production in hill country. N.Z. 1. Agric. Res. 2 5 : 48 5-496 . Reisenauer, H . M . , Clement, C . R . and lones, L.H.P. 1 98 2 . Comparative efficacy o f ammonium and nitrate for grasses. Proc . Int. Plant Nutr. : 5 3 9-544. Rice, C.W. and Tiedje, J.M. 1 9 8 9 . Regulation of nitrate assimilation by ammonium in soils and in isolated soil microorganisms. Soil BioI. B iochem. 2 1 : 597-602. 2 5 5 Rowarth, J . S . and Gillingham, A . G . 1 990. Effects of withholding fertiliser o n pasture production and phosphate cycling in hill country. Proc. N.Z. Grassland Assoc. 5 1 : 1 7- 20. Rumble, PJ. and Esler, A.E. 1 96 8 . Pasture pattern on grazed slopes. N . Z . l Agric. Res. 1 1 : 5 7 5 - 5 5 8 . Ryden, J . C . , Ball, P.R. and Garwood, E.A. 1 98 4 . Nitrate leaching from grassland. Nature 3 1 1 : 50-5 3 . Ryden, J . C . 1 986. Gaseous losses o f nitrogen from grassland. In : Nitrogen Fluxes in Intensive Grassland Systems. Eds. Van Der Meer, H . G . , Ryden, l C . and. Ennik, G.C. pp. 5 9-73 . Martinus Nijhoff Publishers. Saggar, S . , Hedley, MJ. and White, R.E. 1 990. A simplified resin membrane technique for extracting phosphorus from soil. Fert. Res. 24: 1 73- 1 80. S akadevan, K . 1 99 1 . F actors influencing the transformation and fate of sulfur and nitrogen in grazed hill country pastures. PhD thesis, Massey University. Palmerston North, New Zealand. Sakadevan, K . , Mackay, A. and Hedley, MJ. 1 993 . Influence of sheep excreta on pasture uptake and leaching losses of sulfur, nitrogen, and potassium from grazed pastures. Aust. J. Soil Res. 3 1 : 1 5 1 - 1 62 . Sakadevan, K. and Hedley, M . J . 1 993 . Mineralisation and fate of soil sulfur and nitrogen in hill pastures. N .Z. l Agric. Res. 3 6 : 27 1 -28 1 . Sakadevan K, Hedley M . l and Mackay, A . D . 1 994. An In Situ Mini Lysimeter with a removable ion exchange resin trap for measuring nutrient losses by leaching from grazed pastures. Aust. J. Soil Res. 3 2 : 1 3 8 9- 1 400. Sarathchandra, S .u. 1 97 8 . Nitrification activities of some New Zealand soils and the effect of some c l ay types on nitrification. N .Z. J. Agric . Res. 2 1 : 6 1 5 -62 1 . 2 5 6 Schnabel, R.R, Messier, S . R . and Purnell, R F . 1 99 3 . An evaluation o f anion exchange resin used to measure nitrate movement through soil . Comrnun. Soil Sci. Plant Anal. 24: 863-879. Schoj erring, 1.K., Sommer, S.G. and Ferm, M . 1 992. A simple passive samplers for measuring NH3 volatilisation in the field. Water Air and Soil Pollut. 62 : 1 3 -24. Scotter, D . R , White, RE. and Dyson 1 . S . 1 99 3 . The Bums leaching equation. 1. Soil Sci. 44: 2 5 -3 3 . Searle, P.L. 1 97 5 . Automated colorimetric determination o f ammonium ions i n soil extracts with "Technicon Autoanalyzer II equipment' . N.Z. 1. Agric . Res. 1 8 : 1 83 - 1 87. Sears, P . D . 1 950. S oil fertility and pasture growth. 1 . British Grassl and Soc. 5: 267-280. Sears, P.D. 1 95 3 . Pasture growth and soil fertility. The influence of red and white clovers, superphosphate, lime and sheep grazing, on pasture yields and botanical composition. N.Z. 1. Sci. Technology. 35A supplement 1 : 1 -2 9 . S ears, P.D. 1 956. The effect of the grazing animal o n pasture. Proc. V I I Int. Grassland Congr. : 92- 1 02 . S elvaraj ah, N . 1 996. Determination o f sustainabl e nitrogen loading rates for land treatment systems without adequate soil and ground water information : Dairy farm effluent application onto grazed pasture in the Waikato region. In : Proceedings of the Workshop on Recent Developments in Understanding Chemical Movement in Soils. S ignificance in relation to water quality and efficiency of fertiliser use. Eds. Currie, L.D. and Loganathan, P. 1 4- 1 5 February 1 99 6 . Massey University. Palmerston North, New Zealand. S herlock, R.R and Goh, K.M. 1 98 3 . Initial emission of nitrous oxide from sheep urine applied to pasture soil. Soil BioI. Biochem. 1 5 : 6 1 5-6 1 7. 2 5 7 S herlock, R.R. and Goh, K.M. 1 984. Dynamics o f ammonia volatilisation from simulated urine patches and aqueous urea applied to p asture. I. Field experiments. Fert. Res. 5 : 1 8 1 - 1 95 . Sherwood, M . and Fanning, A. 1 9 8 9 . Leaching of nitrate from simulated urine patches. In : Management S ystems to Reduce Impact of Nitrates. Ed. Germon, l . C . pp. 32-44. Elsevier Science Publishers, Barking, Essex. S ibbesen, E. 1 977. A simple ion-exchange resin procedure for extracting plant available elements from soil. Plant S oil 46: 665-669. Sibbesen, E. 1 97 8 . An investigation of the am on exchange resm method for soil phosphate extraction. Plant Soil 50: 305 -32 1 . S ilva, l.A. and Brernner, l.M. 1 966. Determination and isotope-ratio analysis of different forms of nitrogen in soils. Fixed ammonium. S oil Sci. S oc. Am. Proc. 30: 5 87- 5 94. Skogley, E . O . , Georgitis, l .E. and Schaff, B . E . 1 990. The phytoavailability soil test­ PST. Commun. Soil Sci. Plant Anal . 2 1 : 1 229- 1 243 . Skogley, E.O. and Dobermann, A. 1 996. Synthetic IOn exchange resm: Soil and environmental studies. 1. Environ. Qual. 2 5 : 1 3 -24. Sparling, G.P . and West, A.W. 1 98 8 . Modifications to the fumigation-extraction technique to pennit simultaneous extraction and estimation of soil microbial C and N . Commun. Soil Sci. Plant Anal. 1 9 : 327-344. Sparling, G. and S chipper, L. 1 999. Soil quality changes under intensive and extensive land use in New Zealand. In : Best Soil M anagement Practises for Production. Eds. Currie, L.D., Hedley, M . l . , Home, D J . and Loganathan, P. pp. 5 3 -62. Occasional Report No. 1 2 . Fertiliser and Lime Research Centre, Massey University, Palmerston North. 2 5 8 Steel, K.W. , Wilson, A.T. and Saunders, W.M.H. 1 980. Nitrification activity i n New Zealand soils. N.Z. l . Agric . Res. 2 3 : 249-256. Steel, K.W. and ludd, M.l. 1 984. Leaching of nitrate and other nutrients from a grazed pasture. N . Z . 1. Agric. Res. 2 7 : 5- 1 1 . Subler, S . , Blair, l.M. and Edwards, C.A. 1 995 . Using anion exchange membranes to measure soil nitrate availability and net nitrification. Soil. BioI. B iochem. 27: 9 1 1 -9 1 7 . Suckling, F . E .T. 1 959. Pasture management trials on unploughable hill country at Te Awa H. Results for 1 95 1 -5 7 . N.Z. l . Agric. Res. 2: 48 8-543 . Suckling, F . E . T. 1 966. Hill pasture improvement. In : A Guide to Better Use of the Hills. pp. 5 8 . Printed by Newton King Group of Companies in association with Department of S cientific and Industrial Research. Theobald, P .W . 1 98 3 . Some aspects of the nitrogen economy of a rye grass-white clover sward. Report. Diploma in Science. Technical Institute. Palmerston North. Theobald, P .W . and Carran, R.A. 2000. Urine powers pasture production. In : Soil Research. A knowledge industry for land-based exporters . Eds . Currie, L.D. and Loganathan, P. Occasional Report No. 1 3 . Fertiliser and Lime Research Centre, Massey University, Palmerston North. Thomas, R . l. , Logan, K. A . B . , Ironside, A.D. and Bolton, G.R. 1 9 8 8 . Transformations and fate of sheep urine N applied to an upland u.K. pasture at different times during the growing season. Plant Soil 1 07 : 1 73 - 1 8 1 . Tillman, R. 1 99 3 . Principles of fertiliser use on dairy farms . Dairy Farming Annual 45 : 46-49. Tinker, P . B . and Nye, P.H. 2000. Solute movement m the rhizosphere. Oxford University Press, Inc. New York. 259 Vallis, I . , Harper, L.A, Catchpoole, V. R. and Weier, K. L. 1 98 2 . Volatilisation of ammonia from urine patches in a sUbtropical pasture. Aust. J. Agric . Res. 3 3 : 97- 1 07. Vance, E.D., Brookes, P.e. and Jenkinson, D . S . 1 987. An extraction method for measuring soil microbial biomass e. Soil BioI. Biochem. 1 9 : 703-707. Waghorn, G.e., Douglas, G.B., Niezen, J.H., McNabb, W . C . and Foote, A.G. 1 998. Forages with condensed tannins-their management and nutritive value for ruminants. Proc. N.Z. Grassland Assoc. 6 0 : 89-9 8 . Watson, C J . and Mills, e.L. 1 998. Gross nitrogen transformations i n grass land soils as affected by previous management intensity. Soil BioI. B iochem. 3 0 : 743-75 3 . Walker, N . and Wickramasinghe, K.N. 1 979. Nitrification and autotrophic nitrifying bacteria in acid tea soils. Soil BioI. Biochem. 1 1 : 23 1 -236. White, J .G.H. 1 990. Hill and high country pasture: Their ecology and management, Chapter 1 0 . Ed. Langer, R.H.M: 299-3 3 6 . Whitehead, D . e . , Bristow, A.W. and Pain, B . F . 1 98 9 . The influence o f some cattle and pig slurries on the uptake of nitrogen by ryegrass in relation to the fractionation of the slurry N . P lant SoiI . 1 l 7: 1 1 1 - 1 20. Whitehead, D . e . and Bristow, AW. 1 990. Transformations of nitrogen following the application of 1 5N-Iabelled cattle urine to an established grass sward. J. App l . Ecol . 27: 667-6 7 8 . Wild, A 1 98 1 . Mass flow and diffusion. In : The Chemistry of Soil Processes. Eds . Greenland, D J . and Hayes, M . H . B . pp. 3 7-80. John Wiley and Sons Ltd. Williams, P . H . and Haynes, RJ. 1 994 . Comparison of initial wetting pattern, nutrient concentrations in soil solution and the fate of 1 5N labelled urine in sheep and cattle urine patch areas of pasture soil . Plant Soil . 1 62 : 49- 5 9 . 260 Williams, P . H. and Haynes, R.1. 2000. Transfonnations and plant uptake of urine N and S in long and short tenn pastures. Nutr. eycI. Agro Ecosys. 5 6 : 1 09- 1 1 6. Wrage, N . , Velthof, G.L., Beusichem, M.L. and Oenema, O. 200 1 Role of nitrifier denitrification in the production of nitrous oxide. Soil B ioI . B iochem. 3 3 : 1 723- 1 732. Youngdahl, L.J., Pacheco, R., Street, J.1., and Vlek, P . L . G . 1 9 82. The kinetics of ammonium and nitrate uptake by young rice plants. Plant Soil 69: 225-232. 26 1 APPENDIX 1 1 . 1 Statistical significance (P Value) of mean comparisons for the total soil (0- 1 5 cm) mineral N levels shown in Fig. 3 . 6. Table 1 . 1 . 1 Steep site Treatment Days after urine application comparison 1 6 27 1 00 142 SO-S200 0.002 0.0003 0. 1 0.9 0.9 SO-S400 <0.000 1 <0.000 1 0.00 1 0.9 0.9 S200-S400 0.000 1 0.00 1 0.03 0.9 0.9 Table 1 . 1 .2 Flat site Treatment Days after urine application comparison 1 6 27 100 142 FO-F200 0.002 0.Q1 0.08 0.7 0.9 FO-F400 <0.0001 <0.0001 0 .00 1 0.9 0.8 F200-F400 <0.000 1 <0.000 1 0.05 0.8 0.8 1 .2 Statistical significance (P value) of mean comparisons for soil (0- 1 5 cm) NH/ -N levels shown in Fig. 3 . 7 . Table 1 . 2 . 1 Steep site Treatment Days after urine application comparison 1 6 27 1 00 142 SO-S200 0.0003 0.0003 0.2 0.9 0.9 SO-S400 <0.000 1 <0.000 1 0.00 1 0.9 0.9 S200-S400 0.0001 0.000 1 0.02 0.9 0.9 Table 1 .2 .2 Flat site Treatment Days after urine application 1 6 27 1 00 142 FO-F200 0.001 0.01 0.09 0.8 0.9 FO-F400 <0.000 1 <0.000 1 0.003 0.9 0.8 F200-F400 <0.0001 <0.000 1 0 . 1 0.8 0.8 262 1 . 3 Statistical significance (P value) of mean comparisons for soil (0- 1 5 cm) N03--N levels shown in Fig. 3 . 8 . Table 1 . 3 . 1 Steep site Treatment Days after urine application comparison 1 6 27 100 1 42 SO-S200 0.6 0.5 0.004 0.9 0 .8 SO-S400 0.6 0.5 0.2 0.9 0.9 S200-S400 0.9 0.9 0.08 0.9 0 .8 Table 1 . 3 .2 Flat site Treatment Days after urine application companson 1 6 27 100 1 42 FO-F200 0.9 0 .8 0.01 0.8 0.9 FO-F400 0.6 0.5 <0.0001 0.7 0.2 F200-F400 0.6 0.6 <0.000 1 0.8 0.6 263 APPENDIX 2 2 . 1 Statistical significance (P value) of mean comparisons for the total mineral N levels shown in Fig. 5 . 3 . Table 2. 1 . 1 Steep site Treatment Days after urine application Comparison 3 1 2 27 45 66 88 SO - S280 < 0 .000 1 < 0.000 1 < 0.0001 0.06 0.4 0.8 SO - S560 < 0.0001 < 0.000 1 <0.0001 <0.000 1 < 0.000 1 0.9 S280- S560 0.02 0.03 <0.00 1 < 0.000 1 < 0.000 1 0.8 Table 2 . 1 .2 Flat site Treatment Days after urine application comparison 3 1 2 27 45 66 88 FO - F280 0.002 < 0.0001 0.007 0.0006 0.3 0.7 FO - F560 0.0006 < 0.0001 <0.0001 <0.0001 0.0006 0.5 F280- F560 0.6 0.26 0 .000 1 0.008 0 .01 0.3 2.2 Statistical significance (P value) of mean comparisons for the 2 M KCI -extractable NH/-N (0- 1 0 cm) levels shown in Fig. S A . Table 2 .2 . 1 Steep site Treatment Days after urine application comparIson 3 1 2 27 45 66 88 SO - S280 < 0.0001 < 0.000 1 0. 0002 0.4 0.07 0.6 SO - S560 < 0.0001 < 0.0001 <0.000 1 <0.000 1 0.3 0.9 S280- S560 0 .02 0.003 <0.000 1 < 0.000 1 < 0.007 0.8 Table 2.2.2 Flat site Treatment Days after urine application comparison 3 12 27 45 66 88 FO - F280 <0.0001 < 0.000 1 0 .01 0 . 1 0.3 0.4 FO - F560 < 0.000 1 < 0.000 1 <0.000 1 0.0005 0.6 0.6 F280- F560 0.9 0.4 0.0006 0.02 0.6 0.2 2 64 2 . 3 Statistical significance (P value) of mean comparisons for the 2 M KCI -extractable N03--N (0- 1 0 cm) levels shown in Fig. 5 . 6 . Table 2 .3 . 1 Steep site Treatment Days after urine application Comparison 3 1 2 27 45 66 88 SO - S280 0.2 0.007 0.00 1 0.02 1 1 SO - S560 .08 0.001 <0.000 1 <0.000 1 0.0006 1 S280- S560 0.5 0.3 0.01 0.001 0.0006 1 Table 2 . 3 .2 Flat site Treatment Days after urine application companson 3 1 2 27 45 66 88 FO - F280 0.4 0.004 0. 1 0.0005 1 1 FO - F560 0.02 < 0.000 1 <0.0001 0.000 1 0.002 1 F280- F560 0. 1 0.09 0.0002 0.002 0.002 1 2.4 Statistical significance (P value) of mean comparisons for the resin-adsorbed NH/ -N (0- 10 cm) levels shown in Fig. 5 .4. Table 2.4. 1 Steep site Treatment Days after urine application Comparison 3 6 9 12 15 1 8 21 24 27 30 33 36 42 48 55 62 69 76 SO-S280 < 000 1 <.007 <.000 1 <.0001 <.000 1 <.000 1 <.0001 <.0001 <.0001 0.001 .05 .02 .001 .4 .07 .08 . 3 . 1 SO-S560 <.0001 .0008 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.000 1 <.0001 <.0001 <.0001 <.0001 .01 .09 .6 . 1 .9 S280-S560 < .3 .4 . 1 .04 .06 .04 .0005 .0005 .005 .0001 .0002 .03 . 1 .07 .8 .2 .5 . 1 Table 2 .4.2 Flat site Treatment Days after urine application comparison 3 6 9 12 1 5 1 8 2 1 24 27 30 33 36 42 48 55 62 69 76 SO-S280 < <.000 1 <.003 <.000 1 <.0001 <.000 1 <.000 1 <.000 1 <.0001 I 0.001 <0.001 .0007 .000 1 . . 07 .09 .. 3 .004 .2 SO-8560 <.0001 .0008 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 .0006 <.000 1 <.0001 <.0001 .0002 .008 .03 .0 1 . 2 .3 8280-8560 . 1 .3 . 1 . 1 .2 .02 .01 .002 .0006 <.007 .005 .08 .8 .4 .6 . 1 . 1 .7 83 9 1 97 .8 .5 .8 .5 .07 . 1 .6 .2 . 1 83 91 97 . . 3 .6 .7 .3 .6 .9 .9 .9 .7 2 . 5 Statistical significances (P value) of mean comparisons for the resin-adsorbed N03--N (0- 1 0 cm) levels shown in Fig. 5 .6 . Table 2 . 5 . 1 Steep site Treatment Days after urine application comparison 3 6 9 1 2 1 5 1 8 2 1 24 27 30 33 36 42 48 55 62 69 76 SO-S280 . 1 4 .05 <.000 1 <.0001 <.000 1 <.0001 .000) .0008 .0001 <0.001 <.000 1 <.0001 .0005 <.000 1 .07 .02 <.000 1 .006 SO-S560 . 1 7 .) .0005 <.0001 <.000 1 <.0001 <.0001 <.000 1 <.0001 <.0001 <.000 1 <.0001 <.0001 <.000 1 <.0001 <.000 1 <.000 1 <.000 1 S280-S560 .8 .4 .007 .6 .5 .5 .6 .4 .05 . 1 .009 .00 1 <.0001 .0002 .0002 <.000 1 .0 1 <.0001 Table 2 . 5 .2 Flat site Treatment Days after urine application companson 3 6 9 1 2 1 5 1 8 2 1 24 27 30 33 36 42 48 55 62 69 76 SO-S280 .2 .5 .6 . 1 .02 .003 .06 .004 1 .000 1 <.0001 <.0001 .00 1 .05 .004 .8 .7 .000 1 SO-S560 .2 .9 .0) .0002 <.0001 .002 .006 <.0001 .2 <.000 1 <.0001 <.0001 <.000 1 <.0001 <.0001 .2 <.000 1 <.000 1 S280-S560 < .8 .6 .08 .02 .04 .8 .) . 1 .2 .04 .2 .2 .00) .000) .02 .) .000 1 .05 83 9 1 97 . 2 .2 . 5 .009 .009 .00) .001 .001 .0001 83 9 1 97 . 5 1 .04 .01 .02 .) .00) .02 .2 2 . 6 Water balance for Ballantrae flat site from 09/07/00 to 1 01 1 0100 R= Rainfall (mm); ET= Evapotranspiration (mm); W= Water depth (mm) D = Drainage (mm) Date 00/7/09 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 23 24 25 26 27 28 29 30 31 00/08/01 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 23 24 25 26 27 28 29 30 31 Day o 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 R (mm) 0.4 0.2 o o o o o o o o 0.4 3.2 1 3.2 3.4 4.4 o 0.8 2.6 1 2 . 2 9.4 o 0.2 0.4 1 .2 o o o o o o o 1 .8 0.8 20.6 3.8 o o o o o 0.6 2 1 . 6 1 .2 1 5.2 o o 0.2 5.6 0.6 1 3.4 o 1 .2 8 ET(mm) 0.7 0.8 0.9 1 . 1 0.1 1 . 0 1 . 1 0.9 0.7 1 .0 0.6 0.1 0.2 0.3 0.5 1 . 1 0.9 0.6 0.6 0.2 1 .2 1 .0 0.7 1 .2 1 . 1 1 .2 0.9 1 .5 1 .4 1 .5 1 . 1 0.8 0.7 0.4 0.8 1 .2 1 .5 1 .8 1 .3 1 .9 0.3 0.1 0.4 0.8 0.5 1 .2 1 .2 1 .5 0.5 1 .7 1 . 1 1 .6 0.4 0.6 R-ET -0.3 -0.6 -0.9 - 1 . 1 -0.1 - 1 .0 - 1 . 1 -0.9 -0.7 -1 . 0 -0.2 3.2 1 3. 1 3 . 1 3 . 9 - 1 . 1 -0. 1 2.0 1 1 .6 9.2 - 1 .2 -0.8 -0.3 0.0 - 1 . 1 -1 .2 -0.9 - 1 . 5 -1 .4 - 1 . 5 - 1 . 1 1 .0 0 . 1 20.2 3.0 -1 .2 - 1 . 5 -1 . 8 - 1 . 3 - 1 .9 0.3 21 .5 0 . 8 0 . 2 4 . 7 -1 .2 -1 .2 -1 .3 5.1 -1 . 1 1 2.3 - 1 . 6 0.8 7.5 W(mm) 48.5 48.1 47.5 46.7 45.6 45.4 44.4 43.3 42.4 4 1 .5 4 1 .3 44.4 48.0 48.0 48.0 46.9 46.8 48.0 48.0 48.0 46.9 46.0 45.8 45.8 44.7 43.5 42.7 4 1 .2 39.8 38.2 37.1 38.2 38.3 48.0 48.0 46.8 45.3 43.6 42.2 40.3 40.6 48.0 48.0 48.0 48.0 46.9 45.6 44.4 48.0 46.9 48.0 46.4 47.2 48.0 D (mm) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.5 3.1 3.9 0.0 0.0 0.8 1 1 .6 9.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1 0.4 3.0 0.0 0.0 0.0 0.0 0.0 0.0 14.0 0.8 0.2 4.7 0.0 0.0 0.0 1 .5 0.0 1 1 .2 0.0 0.0 6.7 2 6 7 268 00/09/01 49 0 1 .6 -1 .6 46.4 0.0 2 50 3 1 .9 1 . 1 47.5 0.0 3 51 0.2 1 .8 -1 .6 45.9 0.0 4 52 0.4 2.0 -1 .6 44.3 0.0 5 53 1 .6 1 .5 0 . 1 44.3 0.0 6 54 1 .6 1 .6 0 . 0 44.3 0.0 7 55 3.6 0.9 2.7 47.0 0.0 8 56 1 8 0.9 1 7. 1 48.0 1 6.2 9 57 4.8 0.7 4 . 1 48.0 4.1 1 0 58 1 0.8 1 .6 9.2 48.0 9.2 1 1 59 1 .8 2.3 -0.5 47. 5 0.0 1 2 60 0 2.4 -2.4 45. 1 0.0 1 3 6 1 3.2 0.4 2.8 47.9 0.0 14 62 1 .2 0.5 0.7 48.0 0.6 1 5 63 0.2 1 .6 -1 .4 46.6 0.0 1 6 64 0 2.5 -2.5 44.2 0.0 1 7 65 0 3.2 -3.2 4 1 .0 0.0 1 8 66 0 2.4 -2.4 38.6 0.0 19 67 1 0.4 0.6 39.2 0.0 20 68 0.6 1.9 -1 .3 37.9 0.0 21 69 0.2 1 . 6 - 1 .4 36.6 0.0 22 70 0 2.1 -2 . 1 34.4 0.0 23 71 0 3.5 -3.5 30.9 0.0 24 72 0.6 2.8 -2.2 28.7 0.0 25 73 7.6 0.1 7 . 5 36.3 0.0 26 74 0.6 1 . 5 -0.9 35.4 0.0 27 75 0 2.8 -2.8 32.6 0.0 28 76 0 3.5 -3.5 29.2 0.0 29 77 0.8 0.9 -0 . 1 29.1 0.0 30 78 1 6 .8 0.2 1 6.6 45.8 0.0 00/10/01 79 1 1 . 8 0.7 1 1 . 1 48.0 8.8 2 80 27.2 0.3 26.9 48.0 26.9 3 8 1 5.6 3.4 2.2 48.0 2.2 4 82 9.6 1 .4 8.2 48.0 8.2 5 83 0 2 . 1 -2 . 1 45.9 0.0 6 84 0.2 2.7 -2.5 43.5 0.0 7 85 9.2 0.1 9.1 48.0 4.6 8 86 0 0.9 -0.9 47.1 0.0 9 87 1 5 1 .0 1 4.0 48.0 1 3. 1 1 0 88 0 4.1 -4.1 43.9 0.0 2 . 7 Water balance for Ballantrae steep site from 09/07/00 to 1 01 1 0/00 R = Rainfall (mm); ET = Evapotranspiration (mm); W = Water depth (mm) D = Drainage (mm) Date 0017/09 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 23 24 25 26 27 28 29 30 31 00/08/01 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 23 24 25 26 27 28 29 30 Day o 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 R (mm) 0.4 0.2 o o o o o o o o 0.4 3.2 1 3.2 3.4 4.4 o 0 . 8 2.6 1 2.2 9.4 o 0.2 0.4 1 .2 o o o o o o o 1 .8 0 . 8 20.6 3.8 o o o o o 0.6 2 1 .6 1 .2 1 5.2 o o 0.2 5.6 0 . 6 1 3.4 o 1 .2 ET(mm) 0.7 0.8 0.9 1 . 1 0 . 1 1 .0 1 . 1 0.9 0.7 1 .0 0.6 0 . 1 0.2 0.3 0.5 1 . 1 0.9 0.6 0.6 0.2 1 .2 1 .0 0.7 1 .2 1 . 1 1 .2 0.9 1 .5 1 .4 1 .5 1 . 1 0.8 0.7 0.4 0.8 1 .2 1 .5 1 .8 1 .3 1 .9 0.3 0.1 0.4 0.8 0.5 1 .2 1 .2 1 .5 0.5 1 .7 1 . 1 1 .6 0.4 R-ET -0.3 -0.6 -0.9 -1 . 1 -0. 1 -1 .0 -1 . 1 -0.9 -0.7 - 1 . 0 -0.2 3.2 1 3. 1 3 . 1 3.9 -1 . 1 -0. 1 2.0 1 1 .6 9.2 -1 .2 -0.8 -0.3 0.0 -1 . 1 -1 .2 -0.9 -1 .5 -1 .4 - 1 .5 - 1 . 1 1 .0 0 . 1 20.2 3.0 -1 .2 -1 .5 - 1 . 8 - 1 . 3 - 1 .9 0.3 2 1 . 5 0.8 0.2 4.7 - 1 . 2 -1 .2 - 1 . 3 5 . 1 -1 . 1 1 2.3 -1 .6 0.8 W(mm) 48.5 48.1 47.5 46.7 45.6 45.4 44.4 43.3 42.4 4 1 .5 4 1 .3 44.4 48.0 48.0 48.0 46.9 46.8 48.0 48.0 48.0 46.9 46.0 45.8 45.8 44.7 43.5 42.7 4 1 .2 39.8 38.2 37.1 38.2 38.3 48.0 48.0 46.8 45.3 43.6 42.2 40.3 40.6 48.0 48.0 48.0 48.0 46.9 45.6 44.4 48.0 46.9 48.0 46.4 47.2 D (mm) 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 0 . 0 9 . 5 3 . 1 3 . 9 0 . 0 0.0 0.8 1 1 .6 9.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1 0.4 3.0 0.0 0.0 0.0 0.0 0.0 0.0 1 4. 0 0.8 0.2 4.7 0.0 0.0 0.0 1 .5 0.0 1 1 .2 0.0 0.0 269 270 3 1 48 8 0.6 7.5 48.0 6.7 00/09/01 49 0 1 .6 -1 .6 46.4 0.0 2 50 3 1 .9 1 . 1 47.5 0.0 3 5 1 0 . 2 1 .8 -1 . 6 45.9 0.0 4 52 0.4 2.0 -1 .6 44.3 0 . 0 5 53 1 .6 1 .5 0 . 1 44.3 0.0 6 54 1 .6 1 .6 0.0 44.3 0.0 7 55 3.6 0.9 2.7 47.0 0 . 0 8 56 1 8 0.9 1 7 . 1 48.0 1 6.2 9 57 4.8 0.7 4 . 1 48.0 4 . 1 1 0 5 8 1 0. 8 1 .6 9.2 48.0 9.2 1 1 59 1 .8 2.3 -0.5 47.5 0.0 12 60 0 2.4 -2.4 45.1 0.0 1 3 6 1 3.2 0.4 2.8 47.9 0.0 1 4 62 1 .2 0.5 0.7 48.0 0.6 1 5 63 0.2 1 .6 -1 .4 46.6 0.0 1 6 64 0 2 . 5 -2.5 44.2 0.0 17 65 0 3.2 -3.2 4 1 . 0 0.0 18 66 0 2.4 -2.4 38.6 0.0 1 9 67 1 0.4 0.6 39.2 0.0 20 68 0.6 1 .9 -1 .3 37.9 0.0 21 69 0.2 1 .6 -1 .4 36.6 0.0 22 70 0 2.1 -2 . 1 34.4 0.0 23 71 0 3.5 -3 . 5 30.9 0.0 24 72 0.6 2.8 -2.2 28.7 0.0 25 73 7.6 0.1 7.5 36.3 0.0 26 74 0.6 1 . 5 -0.9 35.4 0.0 27 75 0 2.8 -2.8 32.6 0.0 28 76 0 3.5 -3.5 29.2 0.0 29 77 0.8 0.9 -0 . 1 29.1 0.0 30 78 1 6.8 0.2 1 6.6 45.8 0.0 00/10/01 79 1 1 .8 0.7 1 1 . 1 48.0 8 . 8 2 80 27.2 0.3 26.9 48.0 26.9 3 8 1 5 . 6 3.4 2.2 48.0 2.2 4 82 9.6 1 .4 8.2 48.0 8 . 2 5 83 0 2.1 -2. 1 45.9 0.0 6 84 0.2 2.7 -2.5 43.5 0.0 7 85 9.2 0.1 9 . 1 48.0 4.6 8 86 0 0.9 -0.9 47.1 0.0 9 87 1 5 1 .0 1 4.0 48.0 1 3. 1 1 0 88 0 4 . 1 -4 . 1 43.9 0.0 2 7 1 APPENDIX 3 3 . 1 Statistical significance (P value) of mean companson of soil mineral N levels between urine treated and control soils (Fig. 6 . 5 ) Days after urine application Soil 3 9 1 5 2 1 28 45 KAR <0.0001 <0.0001 <0.000 1 <0.000 1 <0.000 1 <0.000 1 KAl <0.0001 <0.0001 <0.000 1 <0.000 1 <0.000 1 <0.000 1 MD <0.000 1 <0.0001 <0.000 1 <0.0001 <0.000 1 <0.000 1 BF <0.000 1 <0.000 1 <0.0001 <0.0001 <0.000 1 <0.000 1 BS <0.000 1 <0.000 1 <0.000 1 <0.0001 <0.000 1 <0.000 1 WF <0.000 1 <0.000 1 <0.000 1 <0.0001 <0.000 1 <0.000 1 WS <0.000 1 <0.000 1 <0.000 1 <0.000 1 <0.000 1 <0.000 1 3 . 2 Statistical significance (P value) of mean comparison of O.SM K2S 04-extractable NH/-N l evels between urine treated and control soils (Fig. 6 . 7) Days after urine application Soil 3 9 1 5 2 1 28 45 KAR <0.000 1 <0.000 1 <0.0001 0.4 0.7 0.6 KAl <0.000 1 <0.000 1 0.9 0.9 0.9 0.9 MD <0.000 1 <0.000 1 <0.000 1 0 .01 0.4 0 .6 BF <0.000 1 <0.000 1 <0.000 1 <0.0001 <0.0001 0.06 BS <0.000 1 <0.0001 <0.0001 <0.0001 <0.000 1 <0.000 1 WF <0.000 1 <0.000 1 <0.000 1 <0.0001 <0.000 1 <0.0001 WS <0.000 1 <0.000 1 <0.000 1 <0.0001 <0.000 1 <0.0001 3 . 3 Statistical significances (P value) of mean comparison of resin-adsorbed NH/-N levels between urine treated and control soils (Fig. 6 . 8) Days after urine application Soil 3 9 1 5 2 1 28 45 KAR 0 .000 1 0.03 0.2 0.9 0.9 0.9 KAl <0.0001 0 .0004 1 0.8 0.9 0.9 MD <0.000 1 0.0006 0.02 0.06 0.5 0.5 BF <0.000 1 <0.000 1 0 .01 <0.0001 0.0027 0.4 BS <0.000 1 <0.0001 0.0003 <0.000 1 <0.000 1 <0.000 1 WF <0.0001 0.0 1 0.02 <0.0001 <0.000 1 0.00 1 1 WS <0.000 1 0. 1 0.09 <0.0001 <0.000 1 <0.0001 2 72 3 . 4 Statistical significances (P value) of mean comparison of 0 . 5 M K2S04-extractable N03--N levels between urine treated and control soils (Fig. 6.9) Days after urine application Soil 3 9 1 5 2 1 28 45 KAR 0. 1 <0.0001 <0.0001 <0.000 1 <0.000 1 <0.0001 KAI 0.08 <0.0001 <0.0001 <0.000 1 <0.000 1 <0.000 1 MD 0.004 <0.000 1 <0.000 1 <0.000 1 <0.000 1 <0.000 1 BF 0.8 0.0 1 0.0002 <0.000 1 <0.000 1 <0.000 1 BS 0.4 0.004 <0.000 1 <0.0001 <0.000 1 <0.000 1 WF 0.7 0.8 0.3 0.04 <0.000 1 <0.000 1 WS 0.2 0. 1 0.0 1 0.0008 <0.000 1 <0.0001 3.5 Statistical significances (P value) of mean comparison of resin-adsorbed N03--N levels between urine treated and controls ( Fig. 6 . 1 0) Days after urine application Soil 3 9 1 5 2 1 2 8 45 KAR 0.2 <0.000 1 <0.000 1 <0.0001 <0.000 1 <0.000 1 KAI 0.3 <0.000 1 <0.000 1 <0.000 1 <0.000 1 <0.000 1 MD 0.2 0.0007 0.0008 0.0009 0. 1 <0.0001 BF 0.5 0.00 1 4 <0.000 1 <0.000 1 <0.000 1 <0.000 1 BS 0.0001 0.2 0.02 0.0002 <0.000 1 <0.000 1 WF 0.06 0.5 0.03 0.007 <0.000 1 <0.0001 WS <0.0001 0 .9 1 0.3 0.004 <0.000 1 3 . 6 Statistical significances (P value) of mean comparison of mineralisable N levels between urine treated and control soils (Fig. 6. 1 1 ) Days after urine application Soil 3 9 1 5 2 1 28 45 KAR 0.2 0.06 0.08 0.0 1 0.02 0 .8 KAI 0.2 0.2 0.004 0.6 0 . 1 0 .5 MD 0.009 O.o I 0.0 1 0.03 0.007 0. 1 BF 0.07 0.0006 0.0005 0.6 0.5 0.4 BS 0.9 0.0003 0.00 1 0.6 0.2 0.05 WF 0.002 0.0034 0.3 1 0.2 0.4 0. 1 WS 0.09 0.009 0.02 0.5 0.9 0.006 273 3 . 7 Statistical significances (P value) of mean comparison of dissolved organic carbon (DOe) levels between urine treated and control soils (Fig. 6 . 1 2) Days after urine application Soil 3 9 1 5 2 1 28 45 KAR 0.8 0.06 0.5 0.2 0.2 0.8 KAI 0.9 0.2 0.8 0 .07 0.06 0.001 MD 0.004 0.00 1 0.0002 0.0002 0.01 0.02 BF 0.02 0.5 0.05 0.4 0.0007 0.02 BS 0.7 0 .3 0.3 0.8 0. 1 0.6 WF 0.7 0.07 0.9 0.3 0.08 0 .009 WS 0.6 0 .8 0 .9 0. 1 0.4 0.07 3 . 8 Statistical significances (P value) of mean comparison of soil microbial biomass N ( S MB-N) levels between urine treated and control soils (Fig. 6. 1 3) Days after urine application Soil 3 9 1 5 2 1 28 45 KAR 0.002 0.2 0 .01 0.006 0.4 0 . 1 KAI 0.2 0.2 0.004 0.6 0. 1 0 .5 MD 0 .8 <0.0001 0.008 0.004 0.2 0 . 1 BF 0.7 0.09 0.003 0.01 0 .01 0.04 BS 0.2 0.0003 0.003 1 <0.000 1 0.8 0.2 WF 0.02 <0.0001 <0.000 1 <0:000 1 0.08 <0.0001 WS 0. 1 0.03 0 .01 0.9 0.4 0 .2 274 APPENDIX 4 Computer program (Visual Basic) for diffusion model described m Section 4 . 6 of C hapter 4. Darker texts are descriptions o f model parameters. Sub twoDO 2-D diffusion, units cm, Ilg, days Dim x As Integer: Dim y As Integer Dim M(20, 20), dM(20, 20), fy(20, 20), fx(20, 20), C(20, 20) BD = 1 .39: W = 0.26: theta = W * BD BD= Bulk Density, W= Gravimetric water content, theta= volumetric water content G = 1 92 G= Soil nitrate-N concentration (Ilg/g soil) Ci = (G * BD) / theta: Cmax = (Ci / 80) * 1 00 Ci= Initial soil nitrate-N concentration (llg/cm3 soil solution) Cmax= Operational maximum nitrate-N concentration (Ilg/ cm3 soil solution) dt = 0 . 0 1 : tmax = 7 dt= Time step (days) tmax= Resin spike burial period (days) dy = 0 .25 : dx = 0.25: ytot = 20: xtot = 1 0: L = 5 dy=Length of cell in x direction, dy= Length of cell in y direction xtot= Number of cells in x direction ytot=Number of cells in y direction L= Length of resin membrane Dl = 1 .3 DJ= Diffusion coefficient of nitrate in free solution (cml/day) f= 0.66 / ( 1 + 77.2375 * Exp(- 1 3 .9964 * theta)) f= Impedance factor Ds = Dl * f: Ds= Diffusion coeffient of nitrate in soil (cml/day) CumfOl = 0: Cumf02 = 0: Cumfmax = 2 1 79 / 4 CumfOl= Nitrate N adsorbed to resin from 10,9 cell Cumf02= Nitrate N adsorbed to resin from 10,10 cell Cumfmax=Maximum Nitrate-N adsorption capacity of resin (llg/cm1 resin) For y = 1 To ytot: For x = 1 To xtot C(y, x) = Ci: M(y, x) = Ci * theta * dy * dx * L M(y,x)= Ilg nitrate-N in each cell C(y,x)=llg nitrate-N in cm"3 soil solution Next x: Next y For y = 1 To ytot fx(y, 0) == 0 : fx(y, xtot) == 0 Next y For x == 1 To xtot fy(O, x) == 0 : fy(ytot, x) == 0 Next x Do Until t >== tmax Worksheets("Sheet4").Cells(2, 3).Value == t Worksheets("Sheet4").Cells(3, 3).Value = (CumfOl + Cumf(2) * 2 For y = 1 To ytot: For x == 1 To xtot - 1 fx(y, x) == theta * Ds * dy * L * (C(y, x) - C(y, x + 1 » / dx fx(y,x)==Nitrate-N flux in X direction Next x: Next y For y == 1 To ytot - 1 : For x == 1 To xtot fy(y, x) == theta * Ds * dx * L * (C(y, x) - C(y + 1 , x)) I dy