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. I Are low-producing plants sequestering carbon at a greater rate than high-producing plants? A test within the genus Chionochloa A thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Ecology Massey University, Palmerston North, New Zealand Matthew Phillip Sijbe Dickson 2016 II III Abstract Plant life and primary production play an important role in the global carbon (C) cycle through the fixing of atmospheric C into the terrestrial biosphere. However, the sequestration of C into the soil not only depends on the rate of plant productivity, but also on the rate of litter decomposition. The triangular relationship between climate, litter quality, and litter decomposition suggests that whilst low-producing plants fix C at a slower rate than high- producing plants, they may release C at an even slower rate, due to the production of a recalcitrant litter. Here, the relationships between environment, productivity, litter quality and decomposition are investigated to determine their relative influences on C sequestration for taxa in the genus Chionochloa. Annual productivity was measured in situ for 23 taxa located across New Zealand, whilst litter and soil were collected for analyses and two ex-situ decomposition experiments; litter incubation on a common alpine soil, and litter incubation on each taxon's home-site soil. Plant growth rate was found to be positively correlated with both litter nitrogen and litter fibre content. Litter decomposition on the common soil was instead negatively correlated with lignin content, which showed a strong correlation with phylogeny, as opposed to environment or growth rate. When incubated on home-site soils, litter quality had no influence on decomposition, which was instead positively correlated with the rate of soil C decomposition, and negatively correlated with both soil organic matter and soil water content. On the common soil there were weak correlations between productivity and decomposition; however the proportional increase in productivity was greater than the corresponding increase in decomposition, resulting in high-producing plants sequestering C at a greater rate than low-producing plants. However, there was no correlation between productivity and decomposition on the home-site soil, with soil water content being a better predictor of C sequestration rate than productivity. Despite the range of variation in morphology, ecophysiology, productivity and habitat displayed within the Chionochloa genus, taxa all produced litter of a very similar quality. Breakdown of that litter is then most strongly influenced by the environment in which decomposition occurs, as opposed to the quality of the litter. Any subsequent differences in rates of C sequestration are therefore most influenced by the environment decomposition occurs in, with wet and cool environments likely to result in increased rates of C sequestration, independent of the rate of productivity. IV V Acknowledgements Firstly, thank you to my supervisor Dr Jill Rapson. Your knowledge, guidance, and critique have been of great assistance. I have greatly enjoyed our ecological and non- ecological discussions and debates. My appreciation goes also to those who gave insight and comment into topics integral to this research: Dr Kevin Tate, Dr Matthew Krna, and PhD candidate Helen Walker. Thank you for sharing your knowledge. Thank you to Matthew Krna, Hamish Baird, Zuni Steer, Mark Dickson, Jennifer Dickson, Thurza Andrew, James Voss, and Josh Olsen for assistance with field work, and similarly to Stacey Gunn for laboratory assistance. Thanks also to Ian Furkert, Helen Walker, Paul Barrett, and Cleland Wallace of Massey University for providing expertise in the laboratory, and technical experience. All your help was much appreciated. Thank goes to the National Institute for Water and Atmospheric Research (NIWA) for providing annual climate data, Rachel Summers of Massey University for GIS assistance, and PhD candidate Gregory Nelson of Otago University for provision of and assistance with Chionochloa genetic distance matrices. My appreciation goes also to Iris and Kate Scott for kindly providing land access to Chionochloa vireta in the Rees Valley, Kevin Henderson for kindly providing accommodation and land access to Poplars Range, and to the Department of Conservation (DoC), with particular thanks to the Stewart Island DoC boat Hananui and Skipper Stephen Meads, for kindly providing water transport. This research was made possible through funding by the Miss E.L. Hellaby Indigenous Grasslands Research Trust, Project Tongariro, the Heseltine Trust, and the Ecology Group of the Institute for Agriculture and Environment, Massey University. Thank you for supporting environmental research. Lastly, much love and thanks to my family for supporting me in my studies, and to God, who has been generous and faithful in all things. VI VII Contents Abstract III Acknowledgements V Contents VII Chapter 1: Introduction 1 Carbon Cycling and Carbon Sequestration 3 CO2 and Global Warming 3 Litter Quality, Productivity, and Decomposition 4 Theory and Hypotheses 5 Chionochloa as a Suitable Study System 7 Research Sites 11 Objectives 11 References 13 Chapter 2: Variation in Litter quality within a congeneric group: Is litter quality more related to environment or phylogeny? 17 Introduction 19 Litter Quality 19 Factors Influencing Litter Quality 19 Measures of Litter Quality 20 Plant Reponses to Stress 22 Genetic Distance and Plant Functional Group 22 Hypotheses and Aims 23 Methods 23 Experimental Design 23 Environmental Data 24 Soil Collection and Preparation 24 Litter Collection and Preparation 25 Litter Chemical Analyses 25 Analysis 26 Results 27 Litter Chemistry 27 General linear models 31 Genetic relatedness 32 Discussion 34 Leaf Nitrogen Content 34 Leaf Structural Components 35 Phenolics, C:N, and Soluble Compounds 36 Environmental Control of Litter Quality 37 Genotypic Control of Litter Quality 38 Conclusions and Implications 39 VIII References 40 Chapter 3: Investigating relationships between environment, plant growth rate, and litter quality: Can litter quality be determined from plant growth rate? 49 Introduction 51 Plant Productivity 51 Influence of Growth Rate on Litter Quality 53 Environmental and Resource Stresses 53 Aims and Hypothesis 54 Methods 54 Species Sampled and Locations 54 Experimental Design 54 Plant Growth Measurements 55 Annual Productivity 55 Litter Production 58 Productivity Measures 59 Analysis 60 Results 61 Measures of Productivity 61 General Linear Models 66 Productivity and Litter Quality 66 Productivity and Genotype 68 Discussion 68 Productivity Measures 68 Productivity and Environmental Stress 69 Influence of Genotype on Productivity 71 Links between Productivity and Litter Quality 71 Conclusions and Implications 73 References 74 Chapter 4: Is litter quality the determining factor in litter decomposition within the genus Chionochloa? A test under controlled conditions 79 Introduction 81 Decomposition and C sequestration 81 Factors Determining Decomposition 82 Litter Quality Parameters and Decomposition 82 Hypotheses and Aims 84 Methods 85 Location, Species Sampled, Litter Collection and Preparation 85 Experimental Design 85 Soil Collection 85 Soil Preparation and Analysis 86 Incubation Chambers 86 Titration 87 Common Soil Incubation 88 Site Soil Incubation 88 Analysis 89 b) IX Results 89 Decomposition Substrate 89 Temporal Trends in Decomposition 91 Rates of Litter Decomposition 91 Cumulative litter carbon loss 93 General Linear Models of Litter Decomposition 96 General Linear Models for Soil C Decomposition 97 Discussion 99 Common Soil Litter Decomposition 100 Litter Quality as a Predictor of Litter Decomposition 101 Home-Site Soil Litter Decomposition 102 Soil Characteristics as Predictors of Litter Decomposition 102 Conclusions, Implications, and Limitations 104 References 106 Appendix 111 Chapter 5: Synthesis and Discussion: Are low-producing plants sequestering C at a greater rate than high-producing plants? 113 Introduction 115 Synthesis of Findings in Chionochloa 115 P:D Ratios and C Sequestration 116 Relationships between Productivity and Decomposition 117 Relationships between Productivity and C sequestration 119 Rate of Productivity, Litter Quality, and C Sequestration 123 Soil Characteristics and C Sequestration 124 C Sequestration in Chionochloa Grassland 126 Climate Change and C Sequestration 127 Limitations and Future Research 128 Conclusions and Implications 128 References 130 X 1 Chapter 1 Introduction Chionochloa tussock grassland, Mt Burns, Fiordland – M. Dickson 2 3 Introduction Carbon Cycling and Carbon Sequestration The long term storage of atmospheric carbon dioxide (CO2) into the soil carbon (C) pool is an important part of the global C cycle. This transfer of C from the atmosphere to the biosphere is made possible through photosynthesis, resulting in net production by primary producers (Pregitzer et al., 2007). In general, a large proportion of this C is released back into the atmosphere through biotic respiration by C consumers, but a small proportion of it is stored long-term in the soil C pool (Wigley and Schimel, 2005). This storage of C in the soil is called C sequestration, and occurs when the ratio of the total C fixed to total C released, within a system over a set time period, is greater than one (Krna and Rapson, 2013). Thus, C sequestration within an ecosystem is equal to the balance between productivity and decomposition (Bradley and Pregitzer, 2007). Determining this balance between productivity and decomposition will provide an insight into the factors driving C sequestration, and may help to identify biotic and abiotic characters that increase C sequestration. The importance of C sequestration has been accentuated by an imbalance in the C cycle as a result of increased anthropogenic CO2 emissions. Cycling of C between oceanic, terrestrial, and atmospheric pools has historically, prior to anthropogenic interference, had long time periods of near net balance (Wigley and Schimel, 2005). Anthropogenic influences, especially those associated with the eighteenth century’s industrial revolution, have disrupted this balance through a number of ways, with the most significant being the destruction of existing vegetation, land use change, fossil fuel burning, and cement production (Houghton, 1991; Watson et al., 1996; Vitousek et al., 1997; Wigley and Schimel, 2005). This trend continues to occur, with anthropogenic CO2 emissions increasing dramatically over the last century, resulting in an increase in the atmospheric C pool (Friedlingstein et al., 2010). CO2 and Global Warming The global atmospheric CO2 concentration has increased dramatically since the industrial revolution from approximately 285 parts per million to now exceeded 400 parts per million (Pachauri et al., 2014; Gall and Nazaroff, 2015). One of the lead-on effects from an increase in atmospheric CO2 has been a change in climate, due to the 4 green-house effect of CO2 and other green house gases. The Intergovernmental Panel on Climate Change (IPCC) has reported an unequivocal rise in average global temperature over the past 50 years, associated with an increase in anthropogenic green-house emissions, with further increases predicted (Pachauri et al., 2014). Of these anthropogenic emissions, CO2 is by far the most prevalent. In 2010 anthropogenic CO2 emissions made up 76% of the total equivalently weighted anthropogenic green-house gas emissions (with Methane at 16%, Nitrous oxide at 6.2%, and Fluorinated gases at 2% making up the remainder) (Pachauri et al., 2014). Hence, reducing this imbalance in the carbon cycle has become an issue of significance worldwide, both for ecosystem preservation and anthropoic reasons. The long term sequestering of atmospheric C into the soil pool in the form of potentially stable humus to reduce atmospheric CO2 is a possible solution for the mitigation of climate change. Long term storage of C in the soil is considered a better solution than the temporary sequestration of C into biomass through afforestation and reforestation (Batjes, 1998; Krna and Rapson, 2013). Hence, understanding relationships between productivity and decomposition is vital for accurate C budgeting and decision-making in the mitigation of global climate change. Litter Quality, Productivity, and Decomposition The decomposition of organic matter ultimately determines how much C is emplaced in the soil and how much C is released (Berg and McClaugherty, 2008). This process is controlled by a number of factors including temperature, moisture, litter quality, and the microbial community (Aerts, 1997; Zhang et al., 2008; Prescott, 2010). Whilst climate is thought to be the primary determinant of decomposition rate, there are suggestions litter quality may play an equally important role, particularly in the formation of stable humus in the latter stages of decomposition (Couteaux et al., 1995). The physical and chemical characteristics of plant litter play an important role in its rate of decomposition (Aerts, 1997). These characteristics can be generally summarised as litter quality. High quality litters contain metabolites and constituents that are readily metabolised and broken down by microorganisms, resulting in rapid rates of decomposition and C release (Cadisch and Giller, 1997). Conversely, poor quality litters contain metabolites and constituents that are recalcitrant to decomposition. Thus, the quality of litter is an important determinant of the amount of C stored and released within the soil. 5 A plants litter quality is thought to be related to the growth strategy implemented by that plant and growth rate at which the material is produced (Coley et al., 1985). The influence of environmental and resource stress on plant growth often results in trade- offs between productivity and survival (Grime, 2006). Adaptations for survival and persistence not only influence the rate of growth, but also the underlying plant physiology, morphology, and resource use strategy, which in turn influences the chemistry of the plant material produced (Poorter and Villar, 1997). Environments that contain favourable conditions and ample resources for growth tend to be dominated by plants with characters that allow them to exploit these conditions (Darwin, 1991; Grime, 2006). These characters, such as rapid rates of growth and a large photosynthetic area, tend to result in the production of a high quality litter that is readily decomposed. In contrast, environments that contain conditions and levels of resources that are limiting for growth, tend to favour plants with characters than allow them to endure and persist in that environment (Grime, 2006). Characters associated with this strategy, such as a slow rate of growth and production of tough leaf material, tend to result in the production of a poor quality litter, recalcitrant to decomposition (Coley, 1988). Theory and Hypotheses The relationships between environment, productivity, litter quality, and decomposition are likely to influence C sequestration. The productivity to decomposition ratio (P:D), as used by Kirschbaum (2000) and Krna (2015) indicates that C sequestration increases when the ratio of productivity to decomposition increases. Where C is added to the surface of the soil at a great rate, as occurring under high producing plants, C sequestration may be zero if C is lost from the soil at an equivalent rate, as occurs in readily decomposing litter (Figure 1; red dashed lines). Underlying the assumption that higher-producing plants are sequestering more carbon than lower-producing plants (Figure 1;b), is the assumption that productivity has little influence on the rate of litter decomposition, assuming the rate litter of decomposition remains relatively constant, independent of the rate at which the litter is produced (Figure 1; a). 6 Figure 1: Hypothetical relationships between productivity and decomposition (green lines), and the corresponding relationships between productivity and C sequestration (P:D ratio; blue lines). The Red dashed line indicates a 1:1 relationship between productivity and decomposition, resulting in no C sequestration where the P:D ≤ 1. a & b) rate of decomposition remains constant independent from rate of productivity. c & d) rate of decomposition increases linearly with rate of productivity. e & f) rate of decomposition is relatively lower in low-producing plants compared to high producing plants. 0 0 De co m po si tio n Productivity a) 0 0 P : D ra tio Productivity b) 0 0 De co m po si tio n Productivity c) 0 0 P : D r at io Productivity d) 0 0 De co m po si tio n Productivity e) 0 0 P : D r at io Productivity f) 7 However, according to the relationships introduced above, there is evidence that the rate of litter decomposition is not independent of the rate of productivity at which that litter was produced. Where the rate of decomposition increases at a constant rate relative to productivity (Figure 1; c), low-producing plants can be expected to sequester approximately equal amounts of C per gram of productivity (Figure 1; d). Due to the tendency of stress-tolerating and low-producing plants to produce litter recalcitrant to decomposition, it could be possible that the relationship between productivity and decomposition is not linear (Figure 1; e), where the rate of decomposition is much lower for low-producing plants. If this decrease in decomposition associated with low- producing plants, is proportionally greater than the associated reduction in productivity, C sequestration per gram of productivity produced may be greater in low-producing plants (Figure 1; f). Chionochloa as a Suitable Study System An appropriate system to test the relationships between productivity, plant litter, and decomposition would require the following: i) ease and accuracy in measurement of annual productivity; ii) variation in productivity, morphology, climate, habit, and subsequent variation in litter quality; and iii) minimal phylogenetic dissimilarities between taxa that may alter litter quality and productivity. Testing for relationships in C sequestration within a congeneric group will allow for the detection of factors influencing C sequestration, independent of major genotypic variation. The genus Chionochloa provides such a suitable system for investigating C sequestration. An Australasian genus of 25 species, of which 23 species are endemic to New Zealand, these tussock grasses predominantly occur in native grassland and scrubland, where they are often the dominant species (Connor, 1991; Connor and Lloyd, 2004). Chionochloa tussocks are long-lived perennials grasses, made up of modular tiller units, which allow for ease in productivity measurements (Mark, 1965b; Williams, 1977), and in addition produce little somatic C sequestration, as woody species do, which complicates the measurement of productivity (Krna and Rapson, 2013). The distribution of Chionochloa ranges from localised to widespread, with species occurring throughout the North, South, Stewart, and offshore Islands, with variation in habitat and niche occurring (Connor, 1991). This, in combination with an altitudinal range from sea-level to high-alpine grasslands, provides a vast range in environmental conditions and resulting productivity 8 Figure 2: Some of the variation in morphology and habit present in smaller taxa in the Chionochloa genus. Taxa are displayed in order of increasing size (height and width) from a – k. a) C. australis, Mt Robert, b) C. crassiuscula ssp. torta, Mt Eldrig, c) C. vireta, Richardson Mts., d) C. juncea, Denniston Plateau, e) C. teretifolia, Mt Burns, f) C. spiralis, Mt Luxmore. Figure 2 (g – k) continues below. (Mark, 1965a; Mark, 1965b; Mark, 1969). Chionochloa also has a variety of growth forms and habits, ranging from less than 15 cm in height to over 130 cm in height, and shows variation in leaf morphology and tiller size (Connor, 1991) (Figure 2). Relationships between litter quality and decomposition have been tested before (Melillo et al., 1982; McClaugherty et al., 1985; Trofymow et al., 2002). However, these studies often test for relationships between plant functional groups and unrelated taxa, which can result in large differences in plant physiology and morphology due to differences in phylogeny. In support of this, Cornelissen et al. (2004) and Cornwell et al. (2008) found the greatest differences in litter quality and rates of litter 9 Figure 2 continued: Larger Chionochloa tussock taxa. g) C. pallens, ssp. pilosa, Poplars Range, h) C. defracta, Red Hills, i) C. rigida ssp. amara, Mt Anglem, j) C. rubra spp. cuprea, Pukerau, k) C. flavescens ssp. lupeola, Mt Rochfort. decomposition between taxa were due to differences in plant functional group and plant species traits, as opposed to environment. However, there is evidence that environment and climate do influence the traits of green leaves and resulting litter quality (Aerts, 1997; Wright et al., 2005), though these relationships may be hidden if phylogenetic differences occur between the taxa studied. From this it is proposed that to accurately detect the relationships between environment, productivity, and decomposition, any phylogenetic differences between taxa need to be minimal, and-or taken into account. g) h) i) j) K) 10 Figure 3: Chionochloa taxa sampled and sampling locations. The Chionochloa genus should provide sufficient variation in environment, productivity, morphology, and consequent litter quality, without any critical genetic differences between taxa to distort this relationship. Relationships between leaf litter traits and decomposition have been found to occur between congeneric species (Geng et al., 1993), supporting the notion that a congeneric group may be a good system to test for the relationships above. This research tests for relationships between a plant’s 11 environment, leaf productivity, leaf litter quality, and leaf litter decomposition, principally focusing on C sequestration of aboveground plant tissues. Research Sites For the 23 species of Chionochloa endemic to New Zealand, as many taxa as possible were located within their natural geographic range. A total of 23 taxa were located, comprising 14 species (61% of genus) and 13 subspecies and or varieties (72% of genus subspecies and varieties), resulting in a wide representation of the different growth forms present in the Chionochloa genus. Taxa occurring in indigenous grassland or the alpine zone were selected for predominantly, although some taxa sampled do occur in montane herbfield-scrubland, with one taxon on montane forest margins, and another under open montane forest canopies. A wide geographic sampling range was also achieved, with species occurring throughout the North Island, South Island, and Stewart Island (Figure 3). Due to the wide geographic range and altitude range, sites differed in their vegetation composition and environmental characters, but at these sites Chionochloa tended to be the dominant vegetation Objectives The aim of this research is to investigate the relationships between productivity, litter quality, and decomposition, and their subsequent roles in C sequestration. It is hypothesised that low-producing plants may be sequestering equal or greater amounts of C per gram of productivity when compared to high producing plants, due to the production of litter recalcitrant to decomposition. Chapter two investigates the range in litter quality and chemistry occurring in the Chionochloa genus, with the aim of determining how climate and environment may be influencing these. It is hypothesised that taxa under greater environmental and climatic stress will produce a poorer quality litter. Chapter three investigates the range of productivity occurring in the Chionochloa genus, with the aim of testing for relationships between productivity and environment, and productivity and litter quality. It is hypothesised that litter quality will be higher in plants with greater productivity, due to a reduction in the concentration of leaf secondary plant metabolites and constituents, and an increase in concentration of leaf nutrients. 12 Chapter four investigates the influence of litter quality on the rate of litter decomposition. The influence of soil characteristics on litter decomposition are also considered, and compared with the influence of litter quality. It is hypothesised that litter decomposition will be lower in taxa with a poorer quality litter, due to the recalcitrant nature of secondary plant metabolites and constituents. Chapter five investigates the relationships between productivity and decomposition, providing a synthesis of the previous chapters. P:D ratios are created to test for relationships between productivity and C sequestration, and ultimately test the hypothesis; are low-producing plants, per gram of productivity, sequestering C at an equal or greater rate than high-producing plants? 13 References Aerts, R. 1997. Climate, leaf litter chemistry and leaf litter decomposition in terrestrial ecosystems: a triangular relationship. Oikos, 439-449. Batjes, N. 1998. Mitigation of atmospheric CO2 concentrations by increased carbon sequestration in the soil. Biology and Fertility of Soils, 27, 230-235. Berg, B. & Mcclaugherty, C. 2008. Plant litter. Decomposition, humus formation, carbon sequestration. 2nd Ed Springer. Bradley, K. L. & Pregitzer, K. S. 2007. Ecosystem assembly and terrestrial carbon balance under elevated CO 2. Trends in Ecology & Evolution, 22, 538-547. Cadisch, G. & Giller, K. E. 1997. Driven by nature: plant litter quality and decomposition, CAB international. Coley, P. 1988. Effects of plant growth rate and leaf lifetime on the amount and type of anti-herbivore defense. Oecologia, 74, 531-536. Coley, P. D., Bryant, J. P. & Chapin, F. S. 1985. Resource availability and plant antiherbivore defense. Science, 230, 895-899. Connor, H. 1991. Chionochloa Zotov (Gramineae) in New Zealand. New Zealand journal of botany, 29, 219-282. Connor, H. & Lloyd, K. 2004. Species novae graminum Novae-Zelandiae II. Chionochloa nivifera (Danthonieae: Danthonioideae). New Zealand Journal of Botany, 42, 531-536. Cornelissen, J. H. C., Quested, H. M., Gwynn-Jones, D., Van Logtestijn, R. S. P., De Beus, M. a. H., Kondratchuk, A., Callaghan, T. V. & Aerts, R. 2004. Leaf digestibility and litter decomposability are related in a wide range of subarctic plant species and types. Functional Ecology, 18, 779-786. Cornwell, W. K., Cornelissen, J. H., Amatangelo, K., Dorrepaal, E., Eviner, V. T., Godoy, O., Hobbie, S. E., Hoorens, B., Kurokawa, H. & Pérez-Harguindeguy, N. 2008. Plant species traits are the predominant control on litter decomposition rates within biomes worldwide. Ecology letters, 11, 1065-1071. Couteaux, M.-M., Bottner, P. & Berg, B. 1995. Litter decomposition, climate and liter quality. Trends in Ecology & Evolution, 10, 63-66. Darwin, C. 1991. On the origin of species by means of natural selection. 1859. Murray, London, 502. 14 Friedlingstein, P., Houghton, R., Marland, G., Hackler, J., Boden, T. A., Conway, T., Canadell, J., Raupach, M., Ciais, P. & Le Quere, C. 2010. Update on CO2 emissions. Nature Geoscience, 3, 811-812. Gall, E. T. & Nazaroff, W. W. 2015. New directions: Potential climate and productivity benefits from CO2 capture in commercial buildings. Atmospheric Environment. Geng, X., Pastor, J. & Dewey, B. 1993. Decay and nitrogen dynamics of litter from disjunct, congeneric tree species in old-growth stands in northeastern China and Wisconsin. Canadian Journal of Botany, 71, 693-699. Grime, J. P. 2006. Plant strategies, vegetation processes, and ecosystem properties, John Wiley & Sons. Houghton, R. A. 1991. Tropical deforestation and atmospheric carbon dioxide. Tropical Forests and Climate. Springer. Kirschbaum, M. U. 2000. Will changes in soil organic carbon act as a positive or negative feedback on global warming? Biogeochemistry, 48, 21-51. Krna, M. A. 2015. Productivity, decomposition and carbon sequestration of Chionochloa species across altitudinal gradients in montane tussock grasslands of New Zealand. PhD, Massey Univeristy, New Zealand. Krna, M. A. & Rapson, G. L. 2013. Clarifying ‘carbon sequestration’. Carbon Management, 4, 309-322. Mark, A. 1965a. Ecotypic differentiation in Otago populations of narrowleaved snow tussock, Chionochloa rigida. New Zealand journal of botany, 3, 277-299. Mark, A. 1965b. The environment and growth rate of narrow-leaved snow tussock, Chionochloa rigida, in Otago. New Zealand journal of botany, 3, 73-103. Mark, A. 1969. Ecology of snow tussocks in the mountain grasslands of New Zealand. Vegetatio, 18, 289-306. Mcclaugherty, C. A., Pastor, J., Aber, J. D. & Melillo, J. M. 1985. Forest litter decomposition in relation to soil nitrogen dynamics and litter quality. Ecology, 266-275. Melillo, J. M., Aber, J. D. & Muratore, J. F. 1982. Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology, 63, 621-626. Pachauri, R. K., Allen, M., Barros, V., Broome, J., Cramer, W., Christ, R., Church, J., Clarke, L., Dahe, Q. & Dasgupta, P. 2014. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. 15 Poorter, H. & Villar, R. 1997. The fate of acquired carbon in plants: chemical composition and construction costs. Pregitzer, K. S., Zak, D. R., Loya, W. M., Karberg, N. J., King, J. S., Burton, A. J., Cardon, Z. & Whitbeck, J. 2007. The contribution of root–rhizosphere interactions to biogeochemical cycles in a changing world. The rhizosphere: An ecological perspective, 155-178. Prescott, C. E. 2010. Litter decomposition: what controls it and how can we alter it to sequester more carbon in forest soils? Biogeochemistry, 101, 133-149. Trofymow, J., Moore, T., Titus, B., Prescott, C., Morrison, I., Siltanen, M., Smith, S., Fyles, J., Wein, R. & Camiré, C. 2002. Rates of litter decomposition over 6 years in Canadian forests: influence of litter quality and climate. Canadian Journal of Forest Research, 32, 789-804. Vitousek, P. M., Mooney, H. A., Lubchenco, J. & Melillo, J. M. 1997. Human domination of Earth's ecosystems. Science, 277, 494-499. Watson, R. T., Zinyowera, M. C., Moss, R. H. & Dokken, D. J. 1996. Climate change 1995. Impacts, adaptions and mitigation of climate change: scientific-technical analyses. Wigley, T. M. & Schimel, D. S. 2005. The carbon cycle, Cambridge University Press. Williams, P. 1977. Growth, biomass, and net productivity of tall-tussock (Chionochloa) grasslands, Canterbury, New Zealand. New Zealand journal of botany, 15, 399- 442. Wright, I. J., Reich, P. B., Cornelissen, J. H., Falster, D. S., Groom, P. K., Hikosaka, K., Lee, W., Lusk, C. H., Niinemets, Ü. & Oleksyn, J. 2005. Modulation of leaf economic traits and trait relationships by climate. Global Ecology and Biogeography, 14, 411-421. Zhang, D., Hui, D., Luo, Y. & Zhou, G. 2008. Rates of litter decomposition in terrestrial ecosystems: global patterns and controlling factors. Journal of Plant Ecology, 1, 85-93. 16 17 Chapter 2 Congeneric variation in litter quality: Is litter quality more related to environment or to phylogeny? Chionochloa rubra ssp. rubra var. rubra, Mt Tongariro – M. Dickson 18 19 Introduction Litter Quality Litter quality is one of the key factors influencing the rate of litter decomposition (Couteaux et al., 1995; Aerts, 1997; Zhang et al., 2008), and subsequently the sequestration of carbon (C) in the soil (De Deyn et al., 2008). Determining correlations between litter quality and decomposition is thus important for accurately calculating and estimating C sequestration. Research into the influence of increased atmospheric carbon dioxide on litter quality (Coûteaux et al., 1999; Norby et al., 2001) has highlighted the importance of the relationship between the environment and leaf litter quality. Leaf litter can be described by the quality and quantity of resources it contains, and how easily those resources are utilised by other organisms, including animals, soil animals, soil microorganisms, and plants (Cadisch and Giller, 1997). In addition, leaf litter can generally be described by is chemistry. Poor quality litters generally comprise low quality resources, and/or a low abundance of resources, as well as potentially containing constituents or characters that prevent consumers from accessing to those resources within the litter. Alternatively, high quality litter is comprised of either high quality resources, and/or an abundance of resources, which in turn are easily accessible to consumers. As a result, differences in initial litter chemistry between species directly influences the process of microbial litter decomposition in the soil (Meier and Bowman, 2008). Here, the relationships between plant litter quality and plant environment are explored within a congeneric group to determine the relative influences of the environment and phylogeny on litter quality. Factors Influencing Litter Quality The resource quality of plant material is broadly defined by a species' plant functional group, and by the environmental conditions under which that plant material is grown (Grime, 1988; Cornelissen et al., 2004). The ecological strategies used by plants to allocate C and N in the growth of live material, as well as the production of plant protective compounds and other secondary metabolites, have obvious implications for the litter once that material has been shed by the plant. Many of the physiological and protective characters that occur in the live green leaf of a plant persist in discarded litter due to incomplete re-absorption of nutrients during senescence. As a result, there is a 20 strong correlation between live leaf chemistry and the chemistry of leaf material discarded as litter (Aerts, 1996; Killingbeck, 1996). Plants can be separated into functional groups based on the resource acquisition strategy they implement for survival, growth, and development. As a result, variation in litter quality seen between species can be attributed to variation in growth strategy used produce leaf material (Cornwell et al., 2008a) and references therein). Plants growing in a resource-rich environment with low-stress tend to have higher productivity, and produce a litter that is more readably decomposed compared to plants growing in a resource-poor and high-stress environment (Coley et al., 1985; Grime, 2006; Ordoñez et al., 2009) (See Chapter 3). Therefore, it can be hypothesised that the environment plays a large part in a plant’s growth strategy, which in turn determines the makeup of live leaf material and consequent litter quality. Measures of Litter Quality Litters with a high nutrient content tend to decompose rapidly as they provide microorganisms with resources that allow for rapid rates of metabolism and growth. In a review of the literature Enriquez et al. (1993) found litter nitrogen (N) and phosphorus (P) to both have strong positive linear relationships with rates of litter decomposition, which are attributed to the high nutrient requirements of microorganisms. Where nutrients are readily available, microbial activity and population growth is greatly increased resulting in increased decomposition of substrate, whilst where limiting, decomposition can be greatly reduced (Gulis and Suberkropp, 2003). The C:N ratio has been commonly used to describe this effect, with C rich and N poor litters found to decompose slowly (Cadisch and Giller, 1997). Secondary plant metabolites also greatly influence the rate of litter decomposition through ‘afterlife’ effects on decomposers (Cornelissen et al., 2004). Secondary plant metabolites are most simply defined as plant constituents not essential or necessary for basic plant functioning (Seigler, 2012). However, these metabolites may enhance plant growth and development through structural support, protection from stress, herbivory, and disease, as well as by allowing plant signalling (Karlovsky, 2008). As a result, it can be expected that plants under greater stress are likely to contain greater concentrations of secondary plant metabolites. Plant material can be separated broadly into two components, soluble compounds able to be leached from the litter, and insoluble components constituting 21 plant fibre. The water soluble fraction of litter is generally made up of simple sugars, lower fatty acids, proteins and peptides (Cadisch and Giller, 1997). Most of these components are easily taken up and metabolized by microorganisms, providing high quality litter which is rapidly and entirely decomposed within the first phase of decomposition. Plant fibre on the other hand, including cellulose, hemicellulose, and lignin, are not as rapidly decomposed. Structural carbohydrates cellulose and hemicellulose provide a good energy source for decomposers, though tend to decompose more slowly than the soluble fraction, decomposing at a steady rate and remaining at a constant concentration in the litter throughout the decomposition (Berg and McClaugherty, 2008). The secondary metabolite lignin is also a structural component found in plant cell walls and has the main functions of mechanical support, aiding water transport in xylem vessels, and chemical defence against predators and microorganisms (Moura et al., 2010). Lignin content is well known to be an indicator of poor litter quality and reduced rates of decomposition (Meentemeyer, 1978; Melillo et al., 1982; Taylor et al., 1989) due to decomposition being restricted to limited types of fungi and bacteria, such as ‘white rot fungi’ (Guerriero et al., 2016). Whilst high N levels are known to increase initial rates of decomposition, in later decomposition where litter lignin concentration is greater, N may inhibit litter decomposition. Berg and Matzner (1997) found N to reduce lignin decomposition, which is attributed to N inhibition of synthesis of ligninolytic enzymes (Carreiro et al., 2000; Sinsabaugh et al., 2002), as well as the formation of recalcitrant humic compounds when N reacts within degraded lignin products (Dijkstra et al., 2004; Berg and McClaugherty, 2008). These humic compounds may act as barriers, preventing microorganisms from accessing more readily decomposable structures such as cellulose and hemicellulose (Nommik and Vahtras, 1982; Liu et al., 1985). As a result, the Lignin: N ratio of litter is used as a key indicator of litter quality for long term decomposition. Therefore, these ratios were derived in this study. Phenols and tannins also can be used as measures of litter quality. A high tannin content can indicate a poor litter quality, as tannins can form recalcitrant complexes with many other compounds during decomposition (Horner et al., 1988). Similarly, phenolic content can also indicate a poor litter quality, as phenols tend to condense into less decomposable forms over time (Berg and McClaugherty, 2008). Phenols can inhibit the growth or function of decomposing organism by binding to enzymes, or 22 chemical binding to N leaving it unusable to decomposers (Martin and Haider, 1980; Waterman and Mole, 1994). Plant Reponses to Stress Plant responses and adaptations to environmental stress include both phenotypic and genotypic changes, including changes in biochemical systems, which allow survival and growth in environments that do not have favourable conditions (Yordanov et al., 2000). At the plant level, environmental and climatic stress often result in a reduction in growth and photosynthesis rates (Grime, 2006). Environment and climate stress also influencing the chemical composition of plant leaves due to the production of secondary metabolites, which functions often aid in the tolerance of this stress (Gershenzon, 1984; Akula and Ravishankar, 2011). The quality and quantity of plant lignin is known to be regulated by developmental and environmental cues, with environmental stress thought to increase lignin production (Campbell and Sederoff, 1996). Stresses known to increase lignin content include decreases in temperature, mineral deficiency, drought, increased solar radiation, as well as disease and herbivory (Moura et al., 2010). An increase in fibre is also known to occur in cold-stressed plants, with increased cold-stress thought to result in the production of tougher leaves, and other mechanisms including thickening of the cell wall (Huner et al., 1981; Stefanowska et al., 1999). Climate has also been shown to influence leaf litter N concentrations. In a meta- analysis Lui et al (2006) found significant positive relationships between leaf litter N concentrations in temporal and boreal forest, and temperature and precipitation. Similarly, along a altitudinal gradient, Craine and Lee (2003) found high altitude grasses to have lower N concentrations when compared to low altitude grasses in New Zealand, though tissue density remained constant. Production of phenols appears linked to UV light protection (Kefeli et al., 2003) as well as protection against herbivory and disease (Hättenschwiler and Vitousek, 2000). Genetic Distance and Functional Group There is mixed opinion in the literature as to whether litter quality is predominantly controlled by genotype and functional group, or by environment. In a meta-analysis of live leaf traits, climate was found to have a modest relationship with leaf traits (Wright et al., 2004). However, the analysis comprised a wide range of 23 function groups, with over 2500 species sampled at over 170 sites worldwide, which may indicate that functional groups are primarily responsible for variation in litter quality, as is indicated by other studies (Cornwell et al., 2008a; Cornwell et al., 2008b; De Deyn et al., 2008; Freschet et al., 2012). In addition, a large proportion of the variation in leaf traits was found to occur between co-existing species, suggesting a range of function plant types are likely to occur at a site independent of climate. A functional group approach is thus unlikely to detect changes in litter quality resulting from differences in environmental conditions (Bradley and Pregitzer, 2007). In this experiment, the use of a congeneric group containing closely related species allows for testing of the relationship between the environment and litter quality, without major phylogenetic differences in litter quality occurring. Hypotheses and Aims In this study, the litter quality of taxa in the genus Chionochloa is assessed using some of the key litter chemistry variables described above to determine if litter quality in the genus varies with environmental conditions. Litter quality is correlated against the climatic and environmental conditions with which the litter was produced in, as well as the genetic distance between taxa. It is hypothesised that a range of litter qualities will occur within the genus, due to a range of environmental and climatic conditions experienced between taxa. Poor quality litters are hypothesised to correlate with increased environmental stress, whilst high quality litters are hypothesised to correlate with environmental conditions that favour rapid rate of growth. Methods Experimental Design Sites were located for a total of 23 taxa, comprising 14 species (61% of genus) and 13 subspecies and varieties (72% of subspecies and varieties within the genus), resulting in a large representation of the different growth habits present in the Chionochloa genus. At each taxon location a 4x4m sampling plot was set up using restricted randomisation (Figure 1a). Plots were restricted to (i) a common location within the known species’ range, (ii) the approximate median altitude within the species’ altitudinal range at that location, (iii) a homogenous environment with 24 predominantly uniform indigenous vegetation, (iv) a site with the flattest available topography or on a ridge line. Sampling occurred over two visits made to these plots, approximately one year apart, with sampling Time 1 in February 2013 and sampling Time 2 in February 2014. Environmental Data Environmental data parameters were recorded during the first visit in February 2013. Parameters recorded included aspect, measured as degrees of deviation from true north, slope (measured along steepest side of plot using a clinometer), altitude, latitude and longitude measured in degrees, and maximum standing height of Chionochloa taxa in the plot. An estimate of annual climate data for the year February 2013 to February 2014 was provided by the National Institute of Water and Atmospheric Research (NIWA) using their modelled Virtual Climate Stations (VCS’s). Over 12,000 VCS’s are generated over a 5km grid across New Zealand, with modelling based on spatial interpolation of actual data observations made at over 6000 climate stations located around the country. A thin-plate smoothing spline model is used for the spatial interpolations (Tait et al., 2006; Tait and Woods, 2007). VCS’s were selected for each site in order of preference for (i) the least distance from each field site, and (ii) least difference in altitude from the field site. Where two VCS’s were equally distant from a site, the VCS with the most similar altitude to the site was chosen. The VCS’s climatic parameters used were mean maximum summer temperature, mean minimum winter temperature, mean summer potential evapotranspiration, average annual wind speed, mean summer solar radiation, mean summer soil moisture deficit, and total annual rainfall. Soil Collection and Preparation Soil samples were collected from the top 10cm of the soil profile (n=10) using a soil corer (diameter 2.8cm), haphazardly from within and directly adjacent to the plot. Any litter, vegetation, and stones >1cm diameter were removed, and samples bulked. As soon as possible (within 4 days), soil samples were stored intact in a dark deep freeze (-20˚C) until analysis. After thawing, a subsample of each soil was sieved to 4mm, and water holding capacity calculated from saturation on a ceramic pressure plate by means of a water bubble tower with 50cm of suction at 5kpa (Klute, 1986). Soil 25 samples were spread out and air dried at 25˚C for 72hrs, then sieved to 2mm to remove any further vegetative material and stones. Larger soil particles were finely ground with a mortar and pestle prior to chemical analyses. Organic soil C and total N content for all soils were analysed by flash combustion using Leco TruMac Analyzer (Leco, 2003), analysed at Landcare Research, Palmerston North, New Zealand. Litter Collection and Preparation Recently dead leaves, including the lamina and sheath, that were still attached to the plant were collected for each of the 23 taxa at its home plot in February 2014. This litter was collected from a minimum of 10 plants within each plot. Where not enough litter was available within the plot, litter was collected from plants directly adjacent to the plot. Between 50-150 recently dead leaves, depending on the leaf size, were bulked for each taxon and stored in the dark at -20˚C as soon as possible thereafter (within 4 days). In the lab, leaf litter was thawed before being oven dried at 30˚C for 72hrs to standardise moisture content, and then impact-ground to pieces less than 1mm in size (Figure 1b). Litter Chemical Analyses Chemical analyses were performed on the impact-ground litter for each of the 23 taxa. Litter was dried again, this time at 60C° for 48hrs before plant tissue carbon (C) and nitrogen (N) were analysed by flash combustion using a Leco TruMac Analyzer (Leco, 2003) at Landcare Research, Palmerston North, New Zealand. Plant C and N are expressed as a percentage of the oven dry weight of litter. Cellulose, hemicellulose, and lignin occurring in plant litter were analysed by sequential treatment of the dried impact-ground litter with a neutral detergent to separate the neutral detergent fibre (NDF) and neutral detergent solution (NDS). Then the sample is washed with an acid detergent to isolate the acid detergent fibre (ADF) from the NDF, and subsequently with 72% H2SO4, before ashing to isolate the lignin content from the ADF. Next NDF, ADF, and Lignin were analysed by a Tecator FibertecTM system (Mertens, 2002) at the Nutrition Laboratory, IAE, Massey University, Palmerston North, New Zealand. NDS is equivalent to the reciprocal of NDF, whilst ADF is equivalent to cellulose plus lignin. Total phenolic content in the impact ground plant litter was analyzed by Folin-Ciocalteau reagent according to Isabelle et al. (2010), at the Nutrition Laboratory, IAE, Massey University, Palmerston 26 Figure 1: a) 4x4m sampling quadrate for Chionochloa australis, Poplars Range. b) Impact ground litter from attached recently dead litter collected from Chionochloa rubra, Mt Tongariro. North, New Zealand. Total phenolic content is expressed as milligrams of gallic acid equivalents per gram of oven dry plant litter. Analysis General linear modelling was used to predict the influence of environmental and climatic variables on litter quality. To allow for parsimony and prevent over-fitting of models, explanatory variables were reduced down to only the strongest predictors (Anderson and Burnham, 2002). Firstly, one variable out of highly correlated pairs (R2 >0.8) were removed using Pearson’s correlation, as the pair closely reflect each other. Next the remaining variables were plotted in a PCA ordination using the statistical program Canoco (Ter Braak and Smilauer, 2012) (Figure 2). To further reduce the number of explanatory variables, only four variables in the PCA ordination, those with the greatest explanatory strength, were selected as well as soil organic C to reflect soil fertility. Explanatory variables included in the model were mean minimum winter temperature, altitude, total annual rainfall, mean summer potential evapotranspiration, and soil organic C. To correlate environment against litter quality, the strongest explanatory variables listed above were used to create general linear models in Systat (Wilkinson, 1992), with a forward stepwise logistic regression (Tabachnick et al., 2001) used to determine the strongest predictors. For each measure of litter quality, a total of seven candidate models were created, including single variable models for each of the five environmental variables, one combined model including linear interactions between all environmental variables, and one further combined model allowing non-linear interactions between altitude, temperature, and rainfall variables. b) a) 27 To indentify the strongest and most likely models for each litter quality parameter, models within the candidate set for each litter quality measure were compared and ranked using Akaike’s information criteria (AIC). The variation AICc was used to account for a small sample size of less than 40 taxa (Symonds and Moussalli, 2011). Models were ranked according to their AICc score and evaluated based on their score relative to the lowest AICc score (AICc ∆i) for that litter quality parameter. Models with an AICc ∆i of <2 can be considered essentially as good as the best candidate model (Richards, 2005), whilst models with an AICc ∆i of >10 were rejected as implausible models and are not reported (Anderson and Burnham, 2002). Akaike weights (AICc Wi) were also calculated, indicating the probability that a model is the best out of the candidate set (Symonds and Moussalli, 2011). Models were also assessed for goodness of fit using R2 (Symonds and Moussalli, 2011), with models with an R2 of <0.1 not reported due to extremely poor fit. To test for relationships between taxa and their litter quality variables, PCA ordinations of litter quality variables for each taxon were created using the statistical program Canoco (Ter Braak and Smilauer, 2012). A Mantel test was performed in the statistical program R (Team, 2014) using the Vegan package to test for correlations between genetic similarity and litter quality. First, a genetic similarity matrix was calculated by extracting phylogenetic branch distances for paired taxa using the Ape package in R. A similarity matrix for each litter quality measure was calculated using the absolute value of the difference between paired taxa. Mantel tests were performed in R using Pearson’s correlation and 999 permutations per test. Results Litter Chemistry Quantities of litter structural carbohydrates, including neutral detergent fibre (NDF), hemicellulose, and cellulose, were variable between taxa (Table 1). The percent NDF ranged from 77.1% (C. vireta) to 85.7% (C. crassiuscula spp. crassiuscula), with a generic mean of 82.1% (SE = 0.44). Hemicellulose showed similar variability ranging from 35.3% (C. australis) to 43.6% (C. pallens spp. pilosa), with a generic mean of 40.2% (SE = 0.43). There was less variation between taxa in cellulose content, ranging 28 Table 1: Summary of chemistry measures for Chionochloa leaf litter. NDF, hemicellulose, cellulose, and lignin are expressed as a percentage of oven dry weight. Phenolic content is expressed as mgg -1 of gallic acid equivalent. SI = Stewart Island. Table 1 is continued below. Taxa NDF (%) Hemi- cellulose (%) Cellulose (%) Lignin (%) Phenolic content (mg/g) C. australis 81.8 35.3 34.6 11.9 1.97 C. conspicua ssp. cunninghamii 84.4 36.8 37.7 9.8 1.59 C. crassiuscula ssp. crassiuscula 85.7 41.5 35.7 8.4 2.44 C. crassiuscula ssp. directa 83.1 40.1 34.7 8.3 1.85 C. crassiuscula ssp. torta 82.2 39.4 34.4 8.5 2.32 C. defracta 77.9 36.2 34.3 7.3 2.25 C. flavescens ssp. lupeola 82.8 40.1 36.9 5.8 1.97 C. juncea 82.8 41.9 32.6 8.3 2.53 C. lanea 84.1 41.2 36.5 6.5 1.94 C. macra 78.9 41.9 33.1 3.9 2.70 C. pallens ssp. cadens 82.6 41.1 36.1 5.4 2.38 C. pallens ssp. pallens 81.9 43.0 32.8 6.1 1.57 C. pallens ssp. pilosa 81.4 43.7 32.8 4.9 1.86 C. rigida ssp. amara 82.1 40.1 36.3 5.6 1.58 C. rigida ssp. amara (SI) 84.6 41.3 37.0 6.4 1.88 C. rigida ssp. rigida 81.2 40.6 36.1 4.5 3.15 C. rubra ssp. cuprea 84.6 42.4 35.6 6.6 2.36 C. rubra ssp. occulta 82.5 40.2 36.3 6.0 1.90 C. rubra ssp. rubra var. inermis 81.5 40.8 34.9 5.8 2.15 C. rubra ssp. rubra var. rubra 79.6 39.2 35.5 4.9 2.13 C. spiralis 83.5 38.6 36.8 8.1 1.37 C. teretifolia 81.1 38.9 35.5 6.7 2.29 C. vireta 77.1 39.5 33.0 4.6 2.58 Mean 82.1 40.2 35.2 6.72 2.12 SE 0.44 0.43 0.32 0.40 0.09 from 32.5% (C. defracta) to 37.8% (C. conspicua spp. cunninghamii), with a generic mean of 35.2% (SE = 0.32). Lignin content was low compared to structural carbohydrates, but showed a greater range, from 3.9% (C. macra) to 11.9% (C. australis), with a generic mean of 6.7% (SE = 0.40). Total phenolic content was also variable between taxa, ranging from 1.3 mg g-1 (C. spiralis) to 3.1 mg g-1 (C. rigida spp. rigida), with a generic mean of 2.2 (SE = 0.09). C content of litter was similar between taxa (Table 1 continued), ranging from 45.2% (C. defracta) to 49.7% (C. crassiuscula spp. directa) with a generic mean of 48.0% (SE = 0.21). Litter N content was particularly low across the genus, ranging 29 Table 1 continued: Summary of chemistry measures of Chionochloa leaf litter. Phenolic content is expressed as mgg-1 of gallic acid equivalent. Concentrations of carbon (C) and nitrogen (N) are expressed as a percentage of oven dry weight. Also included are the corresponding carbon to nitrogen ratio, and lignin (L) to nitrogen ratio. SI = Stewart Island Taxa Phenolic content (mg/g) C (%) N (%) C: N L: N C. australis 1.97 48.9 0.30 165 40.2 C. conspicua sp. cunninghamii 1.59 48.3 0.31 157 31.8 C. crassiuscula sp. crassiuscula 2.44 49.0 0.24 207 35.7 C. crassiuscula sp. directa 1.85 49.7 0.30 166 27.8 C. crassiuscula sp. torta 2.32 49.2 0.30 166 28.6 C. defracta 2.25 45.2 0.27 169 27.4 C. flavescens sp. lupeola 1.97 48.4 0.24 200 23.8 C. juncea 2.53 49.0 0.31 160 27.0 C. lanea 1.94 48.3 0.30 158 21.2 C. macra 2.70 46.7 0.31 153 12.8 C. pallens sp. cadens 2.38 47.7 0.24 200 22.7 C. pallens sp. pallens 1.57 48.2 0.27 180 22.7 C. pallens sp. pilosa 1.86 47.9 0.26 188 19.3 C. rigida sp. amara 1.58 47.6 0.34 140 16.6 C. rigida sp. amara (SI) 1.88 48.2 0.25 192 25.3 C. rigida sp. rigida 3.15 47.4 0.28 170 16.3 C. rubra sp. cuprea 2.36 49.0 0.40 123 16.6 C. rubra sp. occulta 1.90 47.9 0.18 268 33.9 C. rubra sp. rubra var. inermis 2.15 47.8 0.25 190 23.0 C. rubra sp. rubra var. rubra 2.13 46.8 0.34 139 14.6 C. spiralis 1.37 47.7 0.31 152 25.7 C. teretifolia 2.29 49.0 0.43 114 15.6 C. vireta 2.58 47.3 0.31 154 15.0 Mean 2.12 48.0 0.29 170 23.6 SE 0.09 0.21 0.01 6.70 1.53 from 0.18% (C. rubra spp. occulta) to 0.43% (C. teretifolia), with a generic mean of 0.29% (SE = 0.01). The litter C:N ratio was markedly different between taxa, more so than C and N individually, and ranged from 114 (C. teretifolia) to 268 (C. rubra spp. occulta), with a generic mean of 170 (SE = 6.7). A linear regression between C and C:N displayed no relationship (R2 = 0.0005), whilst N and C:N displayed a strongly negative linear relationship (R2 = 0.89). Similarly, taxa were markedly different in Lignin: N ratio, 30 Figure 2: PCA ordination of environmental parameters that may cause plant stress and influence litter quality. Physical parameters are aspect (DevNth), slope (Slope), latitude (Latitude) and altitude (Altm). Soil parameters are Soil organic C (SoilC), Soil total N (SoilN), Soil C to N ratio (SoilCN), Soil water holding capacity (WHCgg), and mean summer soil moisture deficit (SoilMS). Climate parameters are mean maximum summer temperature (MaxTS), mean minimum temperature winter temperature (MinTW), total annual precipitation (RainT), mean daily wind speed (WindA), mean summer solar radiation (SolRS), and mean potential summer evapotranspiration (EvapS). ranging from 12.8 (C. macra) to 40.2 (C. australis), with a mean of 23.6 (SE = 1.5) for the genus. A linear regression between N and Lignin: N showed a moderate negative relationship (R2 = 0.23), whilst Lignin and Lignin: N showed a strongly positive linear relationship (R2 = 0.70). 31 Table 2: Comparison of general linear models predicting the influence of environmental stress on litter quality. K= number of parameters in the model including the constant. AICc = Akaike’s Information Criterion. AICc ∆i = difference in AICc score between model i and the best model out of the candidate model set. AICc Wi = Akaike model weight. Only credible models are presented, with models with an AICc ∆i < 10 and R2 > 0.1 accepted as credible models. Strongest models within a candidate set (AICc ∆i <2) are shown in bold. * Bonferroni Significance to 0.10. ** Bonferroni Significance to 0.05. Litter Parameter Model K P r R2 AICc AICc ∆i AICc Wi NDF MinTW 2 0.0001** 0.721 0.520 21.78 0.94 0.384 EvapS + MinTW 3 0.0001** 0.767 0.589 20.84 0.00 0.613 Cellulose MinTW 2 0.053 0.408 0.166 18.59 0.00 0.446 MinTW*Alt 3 0.043 0.425 0.181 20.86 2.27 0.143 C Soil C 2 0.014** 0.506 0.256 -3.70 1.85 0.179 Alt 2 0.134 -0.322 0.104 0.58 6.13 0.021 EvapS 2 0.037* -0.436 0.190 -1.76 3.79 0.068 MinTW 2 0.009** 0.532 0.283 -4.56 0.99 0.275 Soil C + MinTW 3 0.007** 0.623 0.388 -5.55 0.00 0.451 N Alt 2 0.027* -0.460 0.212 -136.42 0.00 0.683 C:N Alt 2 0.081 0.372 0.138 159.78 0.00 0.510 General Linear Models A PCA ordination of the climatic and environmental variables thought to influence litter quality (Figure 2), showed an extremely strong gradient along axis 1, explaining 96.9% of the variation. Strongest parameters along this axis were total annual rainfall and mean summer moisture deficit, opposed by mean maximum summer temperature. Axis 2 explained a further 3% of the variation, with mean minimum temperature in the winter opposed by altitude. Variables selected for inclusion in general linear models were mean minimum winter temperature (MinTW), altitude (Altm), total annual rainfall (RainT), mean summer potential evapotranspiration (EvapS), and soil organic C (SoilC). Of the general linear models tested (Table 2), MinTW, and MinTW + EvapS, were the strongest predictors of litter NDF, with both being equally suitable models (AICc ∆i <2). Out of these two models, the combined model of MinTW + EvapS showed a stronger Akaike weight than MinTW (0.61 and 0.38 respectively), indicating a 61% probability, given the data used, that this is the strongest model. Individually, NDF 32 Figure 3: Correlations ions between measures of litter quality and environmental/climatic stress. shows a strong positive relationship with MinTW (Figure 3), and a weak negative correlation with EvapS (R2 = 0.21). Similarly, cellulose was best predicted by MintTW, displaying a positive relationship, though this was not strong or significant. Modelling of litter C resulted in three equally good predictors, with moderate to strong positive relationships with SoilC (Figure 3), MinTW, and SoilC + MinTW. Litter N and C:N were best predicted by altitude, with a moderate to weak negative relationship occurring between Litter N and altitude (Figure 3), and a moderate to weak positive relationship between Litter C:N and altitude. Other litter parameters, hemicellulose, lignin, lignin:N, and total phenolics could not be predicted by environment or climate, displaying not credible models.. Genetic Relatedness The PCA ordination of litter chemistry measures (Figure 4) displayed a strong leaf nitrogen gradient along axis 1, explaining 73.9% of the variance, with total N opposed against the C:N ratio. Axis 2 explained a further 22.6% of variance, and was dominated by a strong gradient between leaf structural components, with the strongest opposing variables being lignin and hemicellulose. Groups of related subspecies showed R² = 0.519172 76 80 84 88 -4 -2 0 2 4 N D F (% ) Mean Mininmum Winter Temperature (˚C) a) R² = 0.283 44 45.5 47 48.5 50 51.5 -4 -2 0 2 4 Li tt er C ( % ) Mean Mininmum Winter Temperature (˚C) b) R² = 0.260 45 46 47 48 49 50 0.0 20.0 40.0 60.0 Li tt er C ( % ) Soil C (%) c) R² = 0.212 0 0.1 0.2 0.3 0.4 0.5 0 400 800 1200 1600 Li tt er N ( % ) Altitude (m) d) 33 Figure 4: PCA ordination of litter chemistry variables measured in Chionochloa. Labels are as follows: Total nitrogen (N), total phenolics (PHEN), hemicellulose (HEM), carbon (C), cellulose (CELL), Lignin (LIG), acid detergent fibre (ADF), neutral detergent fibre (NDF), Lignin to nitrogen ratio (LN), and carbon to nitrogen ratio (CN). CN data have been square-root transformed to dampen down the influence of CN on the ordination. Figure 5: PCA ordination of Chionochloa taxa based on their litter chemistry (as per Figure 3). The C:N ratio data have been transformed by square-root to dampen down its influence on the ordination. Labels are as follows: C. australis (AUS), C. conspicua ssp. cunninghamii (CON), C. crassiuscula ssp. crassiuscula (CRAcra), C. crassiuscula ssp. directa (CRAdir), C. crassiuscula ssp. torta (CRAtor), defracta (DEF), C. flavescens ssp. lupeola (FLA), C. juncea (JUN), C. lanea (LAN), C. macra (MAC), C. pallens ssp. cadens (PALcad), C. pallens ssp. pallens (PALpal), C. pallens ssp. pilosa (PALpil), C. rigida ssp. amara (RIGama), C. rigida ssp. amara Stewart Island (RIGamaSI), C. rigida ssp. rigida (RIGrig), C. rubra ssp. cuprea (RUBcup), C. rubra ssp. occulta (RUBocc), C. rubra ssp. rubra var. inermis (RUBine), C. rubra ssp. rubra var. rubra (RUBrub), C. spiralis (SPI), C. teretifolia (TER), and C. vireta (VIR). Species subgroups are underlined in colour: crassiuscula (Yellow), pallens (Green), rigida (Blue), and rubra (Red). 34 Table 3: Mantel test results for correlations between genetic similarity and litter quality. Number of permutations used = 999. Significant (< 0.10) correlations highlighted in bold. Litter quality variable Mantel r p-value NDF 0.024 0.369 ADF 0.215 0.057 Cellulose 0.090 0.150 Hemicellulose 0.434 0.001 Phenolics 0.002 0.461 C 0.062 0.310 N -0.166 0.927 C:N -0.171 0.955 Lignin 0.163 0.098 Lignin:N 0.022 0.351 strong grouping along axis 2, though were less related along axis 1 (Figure 5). A Mantel test between genetic similarity and litter quality measures showed significant correlations between genetic distance and hemicellulose, ADF, and lignin, with no detectable relationship occurring for the other litter quality measures (Table 3). Discussion Whilst taxa in the genus Chionochloa were distinctly similar in some measures of litter quality, in other measures they showed a greater variability. Taxa tended to be similar in measures of fibre, including NDF, ADF, hemicellulose, cellulose, and C, displaying 1.1, 1.3, 1.2, 1.2, and 1.1 fold differences respectively. The greatest differences between taxa in litter chemistry were seen in measures of total phenolics, nitrogen, and lignin, with 2.4, 2.4, and 3.1 fold differences respectively. Similarly, the C:N and Lignin:N ratio also showed a large range between taxa, with 2.4 and 3.2 fold differences respectively. Leaf Nitrogen Content Craine and Lee (2003) suggest that the N content of New Zealand’s indigenous grasses may be among the lowest recorded for C3 grasses anywhere. In support of this, 35 the Chionochloa litter in this study was found to contain very low levels of N, which may indicate a poor litter quality. These measures were also similar to those reported in other studies of Chionochloa. In C. rubra and C. rigida, Young et al. (1994) found recently fallen litter to have approximately 0.25% N per dry weight, whilst Lee and Fenner (1989) found slightly higher levels (0.35% - 0.71%) in the attached dead material of a number of Chionochloa taxa shared with this study. Connor et al. (1970) found mean live leaf N concentrations in C. macra, C. rigida, C. rubra, and C. flavescens to range from 0.51% to 0.89% of dry weight, with concentrations rarely exceeding 1%. When these results are compared to those of pastoral grasses in New Zealand, Chionochloa can be seen to contain much lower N levels. N in the recently dead leaves of a New Zealand perennial ryegrass have been reported to range from 1.5% - 3.2% on unfertile and fertile soils respectively, with live leaf N ranging from 2.9% - 4.2% on unfertile and fertile soils respectively (Hunt, 1983). Concentrations of live leaf N in Chionochloa are not only low compared to other grasses, but are also low when compared to other indigenous montane vegetation. Live leaf N concentrations in New Zealand montane trees were found to range from 1.04% to 1.73% of oven dry weight, Shrubs from 0.92% to 1.25%, and forbs from 1.39% to 2.55% (Körner et al., 1986), suggesting that Chionochloa tussocks may have some of the most N poor litter in New Zealand’s montane environment. According to positive correlations between litter N and litter decomposition (Köchy and Wilson, 1997; Berg and McClaugherty, 2008), these results initially suggest generically Chionochloa litter may be slow to decomposition, not just relative to other grasses, but also relative to other montane vegetation. However, there is a range in litter N content occurring between taxa in the genus Chionochloa, which may also result in differential rates of decomposition within the genus. Leaf Structural Components Leaf structural components, i.e. hemicellulose, cellulose, and lignin, were found in generally similar concentrations to other studies of Chionochloa (Connor et al., 1970; Bailey and Connor, 1972). However, the hemicellulose content measured here was notably higher, with a mean of 40% of dry weight, compared to mean measures of 24% to 30% of dry weight reported in Connor et al. (1970) and Bailey and Connor (1972). This difference is most likely due the above studies sampling from the leaf laminal portion only, whereas this study includes both the leaf lamina and the leaf sheath. 36 Connor and Bailey (1972) found sheaths of Chionochloa to contain increased concentrations of hemicellulose and decreased concentrations of cellulose relative to laminae. The above studies also used live green leaves in contrast to recently dead leaves sampled in this study. Hemicellulose and cellulose may therefore be expected to have a greater relative percentage in dead leaves due to the re-absorption of other plant compounds during senescence. Lignin content was similar to other findings for Chionochloa, though a greater range in lignin content was found here. Connor et al. (1970) found lignin in the tall Chionochloa tussocks, C. macra, C. rigida, C. rubra, and C. flavescens, to range on average from 6.4% of dry weight to up to 8.6%. The greater range in lignin content occurring in this study is most likely due to the wider range of taxa sampled, particularly the due to the inclusion of smaller taxa such as C. australis and C. crassiuscula which tended to have a higher lignin content than tall tussocks. The lignin content of Chionochloa litter appears on average to be intermediate between pastoral grasses (2% to 5% lignin) and cereal grasses (12 -14% lignin) (Connor et al., 1970; Allison et al., 2009), though Chionochloa has a much larger range compared to pastoral and cereal grasses. Thomas and Asakawa (1993) found leaf Lignin content and lignin:N ratio to be significant predictors of leaf decomposition in tropical grasses and legumes, where lignin was found to be negatively correlated with decomposition. The wide range in leaf lignin, and N content, found here in the Chionochloa genus may result in varying rates of litter decomposition between taxa. Phenolics, C:N, and Soluble Compounds Measures of NDF and NDS appear to be relatively constant between Chionochloa taxa, which may translate to minimal differences in decomposition. Chionochloa appears to have much greater NDF, and hence reduced NDS, compared to other pastoral grasses in New Zealand (Bailey and Ulyatt, 1970), resulting in a poorer quality litter and possibly result reduced rates of litter decomposition. Variation in phenolic content between taxa could result in variable rates of decomposition, with higher phenolic content associated with reduced rates of decomposition (Martin and Haider, 1980; Waterman and Mole, 1994). Chionochloa C:N levels were high compared to other grasses and forbs. Cadisch and Giller (1997) report ratios of 20 or less as comprising rapidly decomposable litter, with green leaves having ratios between 25 - 75 decomposing relatively quickly, whilst 37 ratios above 100 result in greatly reduced rates of decomposition. Chionochloa taxa appear to be approaching C:N ratios of bark and soft wood (200 to 500) (Cadisch and Giller, 1997), with ratios ranging from 114 to 268. The range in C:N between taxa also suggests varying litter quality within the genus. Environmental Control of Litter Quality Not all measures of litter quality displayed a response to apparent environmental or resource stress. Plant structural compounds were expected to be greater in taxa experiencing lower temperatures. Thickening of the cell wall and greater concentrations of structural components such as fibre, cellulose, hemicellulose and lignin, have been reported as a plant response to cold stress (Huner et al., 1981; Stefanowska et al., 1999). However, in this study mean winter temperature (MinTW) displayed a positive correlation with structural components C, NDF, and cellulose. The greater concentrations of C, NDF, and cellulose occurring in plants at warmer temperatures is likely due to an increase in plant height associated with increased productivity, where more productive and taller plants require greater structural components to support a tall up-right stature. This is supported by a moderate positive correlation between cellulose and plant height (R2 = 0.38). Leaf C is generally considered to be relatively constant in plants (Cadisch and Giller, 1997); however the results here suggest that combined with warmer winter temperatures, leaf litter C is also greater on soils with greater soil organic C. Leaf litter C:N ratios in trees have been reported to influence the underlying soil C and soil N contents through litter turnover rates (Vesterdal et al., 2008), suggesting that plants can influence their underlying soil conditions through inputs. It is possible that greater organic C in the soil here correlates with greater leaf C:N ratios due to the negative correlation between C:N and decomposition, though this is more likely due to low leaf N, as opposed to greater leaf C. Live leaf N content, expressed per unit leaf area, has been shown to increase with increasing in altitude (Körner et al., 1986; Friend et al., 1989; Hultine and Marshall, 2000), however, this relationship is not so clear for leaf N content when expressed by unit dry weight, as is relevant for measuring N available to decomposers. The results found in this study show leaf litter N per dry weight to be negatively correlated with altitude. In support, Körner et al. (1986) found live leaf N per dry weight to be negatively correlated with altitude for some New Zealand montane taxa. 38 Live leaf N per dry weight for both indigenous and introduced grasses in montane New Zealand have also been found to be negatively correlated with altitude (Craine and Lee, 2003). This negative correlation between leaf litter N and altitude may be explained by soil nitrogen limitation occurring at higher altitudes, as a result of lower rates of N mineralization associated with colder temperatures (Klingensmith and Cleve, 1993). In addition, taxa at low altitude sites in New Zealand are more likely receive N fertilization from livestock or agricultural application. This reduction in leaf N content may also be attributed to the strong relationship between leaf photosynthetic rate per unit mass and leaf N concentration per unit mass (Hikosaka, 2004). If so, this would suggest that the lower leaf N occurring in plants at higher altitudes is linked to a reduction in growth rate associated with increased stress at higher altitudes. Genotypic Control of Litter Quality The grouping of related subspecies seen in the PCA ordination of Chionochloa taxa litter quality measures (Figure 5) indicates there may be some genetic control of litter quality. The subspecies groups C. pallens and C. crassiuscula showed the strongest grouping, whilst subspecies groups C. rigida and C. rubra were less closely related, particularly along axis 1. Grouping for all subspecies groups is strongest along axis 2, which displays a strong structural components gradient, suggesting some structural components may have a strong genetic component. In contrast, subspecies show poorer grouping along axis 1, which displays a strong litter N gradient, suggesting litter N may be influence more by environment than genetic relatedness. In a study of relationships between leaf nutrients, environment, and phylogeny, He et al. (2008) found leaf N concentrations in the taxa of Chinese grasslands to be explained by both phylogeny and environment. However, this study investigated these trends using a large range of unrelated taxa, with over 41 different families sampled. Thus, the strong influence of phylogeny on leaf N reported is most likely due to major differences in plant growth strategy and plant functional group, as highlighted earlier. Where taxa are closely related, as occurring in Chionochloa, leaf N is proposed to be predominantly determined environment, due to the relationships between environment, photosynthetic rate, and leaf N, as discussed above. This is supported by the Mantel test performed here, which reported strong relationships between genetic similarity and the structural components lignin, ADF and 39 hemicellulose, but no relationships for leaf N and other litter quality variables. Campbell and Sederoff (1996) report lignin synthesis to be influenced by both genetic and environmental cues; however the findings in this study suggest that genotype is a greater determinate of lignin content in Chionochloa. The same probably applies to hemicellulose and ADF, which also display strong relationships with phylogeny, and no apparent relationship with environment or resource stress. Conclusions and Implications Congeneric Chionochloa taxa were shown to be variable in litter quality, which is attributable to both differences in environment and genotype. Chionochloa litter quality appears to be generically poor due to low N content, high structural components, and presence of significant lignin. However, there is variability between taxa in lignin, total phenolics, N, and C:N concentrations which is likely to result in variable rates of litter decomposition. Greater N content is likely to result in greater rates of decomposition, whilst greater lignin, total phenolics, and C:N ratios are likely to result in lower rates of decomposition. The relationship between litter N content and altitude suggests that poorer quality litters may be related to environments with increased stress. However, some measures of litter quality also appear more related to genotype, which may have an influence on apparent relationship between litter quality and an environment. These results show there is potential for differential rates of decomposition between congeneric taxa, and imply the environment and genotype both have strong controls over litter quality, but operate differentially across the genus. This suggests plant growth rate, a combination of genotype and environment, may have merit as a predictor of litter quality, as is tested for in the following chapter. 40 References Aerts, R. 1996. Nutrient resorption from senescing leaves of perennials: are there general patterns? Journal of Ecology, 597-608. Aerts, R. 1997. Climate, leaf litter chemistry and leaf litter decomposition in terrestrial ecosystems: a triangular relationship. Oikos, 439-449. Akula, R. & Ravishankar, G. A. 2011. Influence of abiotic stress signals on secondary metabolites in plants. Plant signaling & behavior, 6, 1720-1731. Allison, G. G., Thain, S. C., Morris, P., Morris, C., Hawkins, S., Hauck, B., Barraclough, T., Yates, N., Shield, I. & Bridgwater, A. V. 2009. Quantification of hydroxycinnamic acids and lignin in perennial forage and energy grasses by Fourier-transform infrared spectroscopy and partial least squares regression. Bioresource technology, 100, 1252-1261. Anderson, D. R. & Burnham, K. P. 2002. Avoiding pitfalls when using information- theoretic methods. The Journal of Wildlife Management, 912-918. Bailey, R. & Connor, H. 1972. Structural polysaccharides in leaf blades and sheaths in the Arundinoid grass Chionochloa. New Zealand journal of botany, 10, 533-544. Bailey, R. & Ulyatt, M. 1970. Pasture quality and ruminant nutrition: II. Carbohydrate and lignin composition of detergent-extracted residues from pasture grasses and legumes. New Zealand Journal of Agricultural Research, 13, 591-604. Berg, B. & Matzner, E. 1997. Effect of N deposition on decomposition of plant litter and soil organic matter in forest systems. Environmental Reviews, 5, 1-25. Berg, B. & Mcclaugherty, C. 2008. Plant litter. Decomposition, humus formation, carbon sequestration. 2nd Ed Springer. Bradley, K. L. & Pregitzer, K. S. 2007. Ecosystem assembly and terrestrial carbon balance under elevated CO 2. Trends in Ecology & Evolution, 22, 538-547. Cadisch, G. & Giller, K. E. 1997. Driven by nature: plant litter quality and decomposition, CAB international. Campbell, M. M. & Sederoff, R. R. 1996. Variation in Lignin Content and Composition (Mechanisms of Control and Implications for the Genetic Improvement of Plants). Plant physiology, 110, 3. 41 Carreiro, M., Sinsabaugh, R., Repert, D. & Parkhurst, D. 2000. Microbial enzyme shifts explain litter decay responses to simulated nitrogen deposition. Ecology, 81, 2359-2365. Coley, P. D., Bryant, J. P. & Chapin, F. S. 1985. Resource availability and plant antiherbivore defense. Science, 230, 895-899. Connor, H. & Bailey, R. 1972. Leaf strength in four species of Chionochloa (Arundineae). New Zealand journal of botany, 10, 515-532. Connor, H. E., Bailey, R. & O'connor, K. F. 1970. Chemical composition of New Zealand tall-tussocks (Chionochloa). New Zealand journal of agricultural research, 13, 534-554. Cornelissen, J. H. C., Quested, H. M., Gwynn-Jones, D., Van Logtestijn, R. S. P., De Beus, M. a. H., Kondratchuk, A., Callaghan, T. V. & Aerts, R. 2004. Leaf digestibility and litter decomposability are related in a wide range of subarctic plant species and types. Functional Ecology, 18, 779-786. Cornwell, W. K., Cornelissen, J. H., Amatangelo, K., Dorrepaal, E., Eviner, V. T., Godoy, O., Hobbie, S. E., Hoorens, B., Kurokawa, H. & Pérez-Harguindeguy, N. 2008a. Plant species traits are the predominant control on litter decomposition rates within biomes worldwide. Ecology letters, 11, 1065-1071. Cornwell, W. K., Godoy, Ó. & Westoby, M. 2008b. The leaf economic spectrum drives litter decomposition within regional floras worldwide. Couteaux, M.-M., Bottner, P. & Berg, B. 1995. Litter decomposition, climate and liter quality. Trends in Ecology & Evolution, 10, 63-66. Coûteaux, M.-M., Kurz, C., Bottner, P. & Raschi, A. 1999. Influence of increased atmospheric CO2 concentration on quality of plant material and litter decomposition. Tree Physiology, 19, 301-311. Craine, J. & Lee, W. 2003. Covariation in leaf and root traits for native and non-native grasses along an altitudinal gradient in New Zealand. Oecologia, 134, 471-478. De Deyn, G. B., Cornelissen, J. H. & Bardgett, R. D. 2008. Plant functional traits and soil carbon sequestration in contrasting biomes. Ecology letters, 11, 516-531. Dijkstra, F. A., Hobbie, S. E., Knops, J. M. & Reich, P. B. 2004. Nitrogen deposition and plant species interact to influence soil carbon stabilization. Ecology Letters, 7, 1192-1198. 42 Enriquez, S., Duarte, C. M. & Sand-Jensen, K. 1993. Patterns in decomposition rates among photosynthetic organisms: the importance of detritus C: N: P content. Oecologia, 94, 457-471. Freschet, G. T., Aerts, R. & Cornelissen, J. H. 2012. A plant economics spectrum of litter decomposability. Functional Ecology, 26, 56-65. Friend, A., Woodward, F. & Switsur, V. 1989. Field measurements of photosynthesis, stomatal conductance, leaf nitrogen and δ 13 C along altitudinal gradients in Scotland. Functional Ecology, 117-122. Gershenzon, J. 1984. Changes in the levels of plant secondary metabolites under water and nutrient stress. Phytochemical adaptations to stress. Springer. Grime, J. P. 1988. The CSR model of primary plant strategies—origins, implications and tests. Plant evolutionary biology. Springer. Grime, J. P. 2006. Plant strategies, vegetation processes, and ecosystem properties, John Wiley & Sons. Guerriero, G., Hausman, J. F., Strauss, J., Ertan, H. & Siddiqui, K. S. 2016. Lignocellulosic biomass: Biosynthesis, degradation, and industrial utilization. Engineering in Life Sciences, 16, 1-16. Gulis, V. & Suberkropp, K. 2003. Effect of inorganic nutrients on relative contributions of fungi and bacteria to carbon flow from submerged decomposing leaf litter. Microbial Ecology, 45, 11-19. Hättenschwiler, S. & Vitousek, P. M. 2000. The role of polyphenols in terrestrial ecosystem nutrient cycling. Trends in ecology & evolution, 15, 238-243. He, J.-S., Wang, L., Flynn, D. F., Wang, X., Ma, W. & Fang, J. 2008. Leaf nitrogen: phosphorus stoichiometry across Chinese grassland biomes. Oecologia, 155, 301-310. Hikosaka, K. 2004. Interspecific difference in the photosynthesis–nitrogen relationship: patterns, physiological causes, and ecological importance. Journal of plant research, 117, 481-494. Horner, J. D., Gosz, J. R. & Cates, R. G. 1988. The role of carbon-based plant secondary metabolites in decomposition in terrestrial ecosystems. American Naturalist, 869-883. Hultine, K. & Marshall, J. 2000. Altitude trends in conifer leaf morphology and stable carbon isotope composition. Oecologia, 123, 32-40. 43 Huner, N., Palta, J., Li, P. & Carter, J. 1981. Anatomical changes in leaves of Puma rye in response to growth at cold-hardening temperatures. Botanical Gazette, 55-62. Hunt, W. 1983. Nitrogen cycling through senescent leaves and litter in swards of Ruanui and Nui ryegrass with high and low nitrogen inputs. New Zealand journal of agricultural research, 26, 461-471. Isabelle, M., Lee, B. L., Lim, M. T., Koh, W.-P., Huang, D. & Ong, C. N. 2010. Antioxidant activity and profiles of common vegetables in Singapore. Food Chemistry, 120, 993-1003. Karlovsky, P. 2008. Secondary metabolites in soil ecology, Springer. Kefeli, V. I., Kalevitch, M. V. & Borsari, B. 2003. Phenolic cycle in plants and environment. J. Cell Mol. Biol, 2, 13-18. Killingbeck, K. T. 1996. Nutrients in senesced leaves: keys to the search for potential resorption and resorption proficiency. Ecology, 77, 1716-1727. Klingensmith, K. & Cleve, K. V. 1993. Patterns of nitrogen mineralization and nitrification in floodplain successional soils along the Tanana River, interior Alaska. Canadian Journal of Forest Research, 23, 964-969. Klute, A. 1986. Water retention: laboratory methods. Methods of soil analysis. Part 1. Physical and mineralogical methods. 2nd Edition ed. Köchy, M. & Wilson, S. D. 1997. Litter decomposition and nitrogen dynamics in aspen forest and mixed-grass prairie. Ecology, 78, 732-739. Körner, C., Bannister, P. & Mark, A. 1986. Altitudinal variation in stomatal conductance, nitrogen content and leaf anatomy in different plant life forms in New Zealand. Oecologia, 69, 577-588. Leco 2003. Carbon and Nitrogen in plant tissue. LECO Corporation. St. Joesph, MO: Organic Application Note 203-821-171. Lee, W. & Fenner, M. 1989. Mineral nutrient allocation in seeds and shoots of twelve Chionochloa species in relation to soil fertility. The Journal of Ecology, 704- 716. Liu, S.-Y., Freyer, A. J., Minard, R. D. & Bollag, J.-M. 1985. Enzyme-catalyzed complex-formation of amino acid esters and phenolic humus constituents. Soil Science Society of America Journal, 49, 337-342. 44 Martin, J. & Haider, K. 1980. Microbial degradation and stabilization of 14C-labeled lignins, phenols, and phenolic polymers in relation to soil humus formation. Lignin biodegradation: microbiology, chemistry, and potential applications. Meentemeyer, V. 1978. Macroclimate and lignin control of litter decomposition rates. Ecology, 59, 465-472. Meier, C. L. & Bowman, W. D. 2008. Links between plant litter chemistry, species diversity, and below-ground ecosystem function. Proceedings of the National Academy of Sciences, 105, 19780-19785. Melillo, J. M., Aber, J. D. & Muratore, J. F. 1982. Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology, 63, 621-626. Mertens, D. R. 2002. Gravimetric determination of amylase-treated neutral detergent fiber in feeds with refluxing in beakers or crucibles: collaborative study. Journal of AOAC international, 85, 1217-1240. Moura, J. C. M. S., Bonine, C. a. V., De Oliveira Fernandes Viana, J., Dornelas, M. C. & Mazzafera, P. 2010. Abiotic and biotic stresses and changes in the lignin content and composition in plants. Journal of integrative plant biology, 52, 360- 376. Nommik, H. & Vahtras, K. 1982. Retention and fixation of ammonium and ammonia in soils. Nitrogen in agricultural soils, 123-171. Norby, R. J., Cotrufo, M. F., Ineson, P., O’neill, E. G. & Canadell, J. G. 2001. Elevated CO2, litter chemistry, and decomposition: a synthesis. Oecologia, 127, 153-165. Ordoñez, J. C., Van Bodegom, P. M., Witte, J. P. M., Wright, I. J., Reich, P. B. & Aerts, R. 2009. A global study of relationships between leaf traits, climate and soil measures of nutrient fertility. Global Ecology and Biogeography, 18, 137-149. Richards, S. A. 2005. Testing ecological theory using the information-theoretic approach: examples and cautionary results. Ecology, 86, 2805-2814. Sinsabaugh, R., Carreiro, M. & Repert, D. 2002. Allocation of extracellular enzymatic activity in relation to litter composition, N deposition, and mass loss. Biogeochemistry, 60, 1-24. Stefanowska, M., Kuras, M., Kubacka Zebalska, M. & Kacperska, A. 1999. Low temperature affects pattern of leaf growth and structure of cell walls in winter oilseed rape (Brassica napus L., var. oleifera L.). Ann. Bot, 84, 313-319. 45 Symonds, M. R. & Moussalli, A. 2011. A brief guide to model selection, multimodel inference and model averaging in behavioural ecology using Akaike’s information criterion. Behavioral Ecology and Sociobiology, 65, 13-21. Tabachnick, B. G., Fidell, L. S. & Osterlind, S. J. 2001. Using multivariate statistics. Tait, A., Henderson, R., Turner, R. & Zheng, X. 2006. Thin plate smoothing spline interpolation of daily rainfall for New Zealand using a climatological rainfall surface. International Journal of Climatology, 26, 2097-2115. Tait, A. & Woods, R. 2007. Spatial interpolation of daily potential evapotranspiration for New Zealand using a spline model. Journal of Hydrometeorology, 8, 430- 438. Taylor, B. R., Parkinson, D. & Parsons, W. F. 1989. Nitrogen and lignin content as predictors of litter decay rates: a microcosm test. Ecology, 70, 97-104. Team, R. C. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2013. ISBN 3-900051- 07-0. Ter Braak, C. J. & Smilauer, P. 2012. Canoco reference manual and user's guide: software for ordination, version 5.0. Thomas, R. J. & Asakawa, N. 1993. Decomposition of leaf litter from tropical forage grasses and legumes. Soil Biology and Biochemistry, 25, 1351-1361. Vesterdal, L., Schmidt, I. K., Callesen, I., Nilsson, L. O. & Gundersen, P. 2008. Carbon and nitrogen in forest floor and mineral soil under six common European tree species. Forest Ecology and Management, 255, 35-48. Waterman, P. G. & Mole, S. 1994. Analysis of phenolic plant metabolites, Blackwell Scientific. Wilkinson, L. 1992. SYSTAT for Windows: statistics, graphics, data, getting started, version 5, Systat. Wright, I. J., Reich, P. B., Westoby, M., Ackerly, D. D., Baruch, Z., Bongers, F., Cavender-Bares, J., Chapin, T., Cornelissen, J. H. & Diemer, M. 2004. The worldwide leaf economics spectrum. Nature, 428, 821-827. Yordanov, I., Velikova, V. & Tsonev, T. 2000. Plant responses to drought, acclimation, and stress tolerance. Photosynthetica, 38, 171-186. Young, R. G., Huryn, A. D. & Townsend, C. R. 1994. Effects of agricultural development on processing of tussock leaf litter in high country New Zealand streams. Freshwater Biology, 32, 413-427. 46 Zhang, D., Hui, D., Luo, Y. & Zhou, G. 2008. Rates of litter decomposition in terrestrial ecosystems: global patterns and controlling factors. Journal of Plant Ecology, 1, 85-93. 47 48 49 Chapter 3 Investigating relationships between environment, plant growth rate, and litter quality: Can litter quality be determined from plant growth rate? Chionochloa teretifolia, Mt Burns, Fiordland - M. Dickson 50 51 Introduction The sessile nature of plants and their inability to seek out or escape from certain abiotic and biotic conditions has serious consequences for plant growth and survival. This has generally led to the ability of plants to ameliorate the conditions they live in through physiological and morphological adaptations, as well as adaptations in resource use strategies (Ahuja et al., 2010). These adaptations and strategies result in tradeoffs between productivity and survival, with an increase in a plants ability to tolerate and survive in adverse conditions often resulting in a reduction in productivity (Chapin et al., 1987; Grime, 2006). The resulting changes in plant physiology, morphology, and resource use strategy not only affect plant growth rate, but also influence the chemistry of the plant material produced (Poorter and Villar, 1997). However, most studies investigating relationships between environment, plant growth rate, and litter quality include taxa from vastly different phylogenies and functional groups, resulting in genotypic differences that may have a greater influence than environment (Bradley and Pregitzer, 2007). To reduce the influence of genotype on plant growth rate and litter quality, here, a congeneric group of tussocks is used to test for relationships between environmental stresses and plant growth rate, and their consequent influence on plant leaf litter chemistry. Plant Productivity Productivity in vegetation is subject to a variety of environmental constraints, commonly including shortages and excesses in supply of solar energy, water ,and mineral nutrients (Grime, 2006). However, even when resources are in excess, plant growth rate is not always equal among different taxa (Chapin, 1991). The growth rate and productivity of a plant can broadly be defined by its resource use strategy adapted for its environment, as Grime (1988) proposed through expansion on the r/K selection model proposed by MacArthur and Wilson (1967). The r/K selection model suggests that where there are productive habitats with continuous resource replenishment, dominance is achieved by rapid rates of resource capture, which translates to rapid rates of growth and/or reproduction (r selection). Conversely, where there are systems with limited resources, high environmental stress, and low productivity, dominance is achieved through tolerance of these conditions and the protection of captured resources, translating to efficiency and longevity (K selection). Grimes C-S-R theory (Grime, 52 1988) expands on this by grouping plants in to three categories, competitors, stress- tolerators, and ruderals, based on the environmental conditions and disturbance regimes they experience. This study of long lived perennials tussocks in the genus Chionochloa focuses on the environmental stress gradient between competitors and stress-tolerators in the absence of site disturbance, aiming to investigate the relationship between environment, growth rate, and litter quality. The nature of competitors leads them to dominate in sites with low environmental stress and low disturbance. Ample resources in the absence of disturbance allow for plants with the highest growth rates to produce greater quantities of photosynthate. This allows competitors to have dominating traits such as tall structures, extensive lateral spread, build-up of large perennating organs, and rapid expansion of the surface areas of leaves and roots (Grime, 2006). Due to available resources, highly competitive plants are able to quickly replace and replenish any damaged leaf material, resulting in a high leaf-turnover rate (Coley, 1988). In addition, highly productive plants generally do not invest in the types of plant compounds and secondary metabolites that aid in leaf protection and longevity, as it is more cost effective to replace lost or damaged material than to produce protective compounds, the cost of which results in a reduction in growth rate (Coley et al., 1985). In contrast, the nature of stress-tolerators leads them to dominate sites with high environmental stress and low disturbance. High environmental stress favours traits that allow the retention and protection of plant resources, relative to fast-growing plants, due to the increased relative cost of damage owing to slow rates of recovery associated with slow growth rate and difficulty in replacing resources (Coley et al., 1985; Chapin, 1991). The production of secondary plant metabolites and associated compounds aids in leaf protection from physical environmental stress, herbivory, disease, as well as aiding in leaf longevity, which prevents resource loss through litter shedding (Coley et al., 1985). As a result, stress tolerators are comparatively long-lived, and tend to have features which relate to the capacity for endurance. Features include inherently slow growth rates, evergreen ha