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. Structure, function and quality development in apples A thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy III Plant Biology at Massey University, Palmerston North, New Zealand Lazar R. DraZeta 2002 11 Dedication This thesis is dedicated to my family in recognition of the suffering they endured and the love they invested in me. 111 Abstract Relationships between structure and function in apples were assessed in a series of experimental studies on different aspects of fruit quality development. A holistic model is presented that describes the maj or contributing elements. Special attention is directed to the mechanisms underlying the various physiological responses. The study of the relationship between seed set, fruit shape and fruit minerals showed that fruit growth was modular and was strongly influenced by seed distribution and seed growth. This pattern influences the spatial distribution of minerals (e.g. calcium) within the fruit and thus is likely to impact on fruit storage quality. The nature of the developmental stimulus was investigated by in situ applications of the auxin-transport inhibitor, N-( 1 -Naphthyl)phthalamic acid (NP A). NP A application reduced vessel differentiation in the stalk, enhanced fruit abscission, and led to a reduced seed and flesh growth. Because of the specificity of the action o f NP A, the identity of the developmental stimulus is most likely to be auxin. The slowing and eventual stoppage o f calcium import by apples is probably due to a decline in the functionality of its xylem. This decline in functionality was shown to be due to a physical disruption of the vascular bundles caused by flesh expansion. The extent of xylem dysfunction was governed by the structure and character of the tissue surrounding a particular bundle type. Flesh expansion also influences the textural properties of the fruit. The spatial distribution of intercellular air within the fruit and the shape and mutual disposition of the flesh cells were assessed using novel techniques. The radial pattern of intercellular air was not uniform indicating features that are likely to impact upon gas transfer within the fruit and thus on storage behaviour. The mechanistic understandings gained in this thesis permit the elucidation of complex interrelations between the processes of fruit quality development. The work offers new insights into the origins of some physiological defects and indicates new lines for future research. Acknowledgements I wish to thank: all people and organisations that have helped me to complete this thesis and have contributed to the well being of my family. IV Firstly, I thank: my supervisors, Dr Alexander (Sandy) Lang, Prof Paula E. Jameson and Dr Richard Volz for their patient guidance and selfless support throughout the course of this project. I have gained a lot from them. Hence, they gain all my respect and remembrance. The completion of this work would not have been possible without the assistance of Dr Alistair Hall, Dr Nihal De Silva, Dr Chiara Cappellini, Mr Michael Jarvis, Mr Doug Hopcroft, Mr Raymond Bennett, Ms Kathy Murray, Miss Andrea Leonard-Jones and Mrs Geraldine Wood. In addition, superb service and encouragement were provided by the HortResearch librarians Mr Steven Northover, Mrs Ann Ainscough and Ms Sarah Nation. I deeply appreciate help, inspiration and friendship I have received from many good people: Dr Zlatko KovaCic, Dr Dora Duarte de Carvalho, Dr Paul Austin, Dr Stefan Henton, Miss Geogina Milne, Miss Helen Brown, Mr Dusan Backulj a, Pro f Mico Olj aca, Mrs Catherine Kearins, Mr Justin Wilson, Dr Kate Maguire, Dr Markus Deurer, Dr Pieter Wolswinkel, Dr Bartolomeo Dichio, Mr Dave Adams, Mr Ben Anderson, Mr Murray Oliver, Mr Ian McIvor, families Clemens, Hayward, Vukcevic, Bizumic, Hotcenko, Fomin and Laurenson. Special thanks are due to Mrs Margaret Lawler and Mrs Sandra Simpson who were my first contacts in New Zealand and strongly supported my working towards a PhD. Finally, I wish to acknowledge the financial and professional support provided by the Agricultural and Marketing Research and Development Trust (AGMARDT), Work and Income New Zealand - Student Services, HortResearch, Massey University, and New Zealand Society of Plant Physiologists (NZSPP) that made this work possible. v Table of contents DEDICATION ................................................................................................................. ii ABSTRACT .................................................................................................................. iii ACKNOWLEDGEMENTS ............................................................................................ iv TABLE OF CONTENTS ................................................................................................. v LIST OF FIGURES .......................................................................................................... x LIST OF TABLES ........................................................................................................ xiv CHAPTER 1 . GENERAL INTRODUCTION 1 . 1 . Statement of the problem ...................................... ......... . ... ......... ... ... .. .......... .......... .. 1 1 .2 . Structure of apple fruit .............................................................................................. 3 1 .2 . 1 . Structural origins of apple fruit ........................................................... ....... 3 1 .2.2. Structure of the skin . . .... .... ......................................................................... 4 1 .2 . 3 . Structure of the extracarpeUary tissue ........................................................ 4 1 .2.4. Structure of the carpellary tissue ............................................................... . 4 1 .2 . 5 . Structure of the vascular tissue ................................................................... 4 1 .2 .6 . Structure of the seed ................................................................................... 6 1 .3 . The attributes of fruit quality ................................................................................ .... 6 1 .3 . 1 . Skin properties .......................................................................................... . 6 1 .3 .2 . Fruit size .................... . ................................................................................ 6 1 .3 .3 . Fruit shape ................................................................. ................................. 8 1 .3 .4. Flesh texture ............................................................................................... 9 1 .3 .4. 1 . Cell wall ..................................................................................... . 9 1 .3 .4.2. Cell shape .................................................................................. 1 0 1 .3 .4.3. Cell packing ............................................................................... 1 1 1 .3 .4.4. Tissue and air volume of the fruit ......................................... .... 1 2 1 .3 . 5 . Storage quality .......................................................................................... 1 3 1 .3 .5 . 1 . Vascular development - induction and growth pattern ............. 1 3 VI 1 .3 . 5 .2 . Xylem functionality diurnal and seasonal changes ..... ........... 1 5 1 .3 . 5 .3 . Xylem functionality - mineral accumulation and storage quality ..... ........................................ .............................. 1 6 1 .3 .6 . Eating quality ......... .......... ...... ......................................................... ......... 1 8 1 .3 .7 . Seed set .................................. ................................................................... 1 8 1 .3 . 7 . 1 . Ovule development ...................................................... ........ ... .. 1 9 1 .3 .7.2. Pollen development ............................. .................................... .. 1 9 1 .3 .7.3 . Seed development ........................................................ ............ . 2 1 1 .3 . 7 .4. Seeds in fruit development ........................................................ 2 1 1 .3 . 7 . 5 . Seeds as sites of auxin production .. ............... .. ..................... .... 22 1 .3 .7.6. Cellular mechanism of auxin transport and development of storage quality ................................................. 24 1 .4. Conceptual model of fruit quality development ...................................................... 2 5 1 .5 . Research objectives of the thesis . ................... ...... ................................................... 27 CHAPTER 2. SEED SET, FRUIT SHAPE AND FRUIT MINERALS 2. 1 . Introduction ................................................... .......................................................... 2 8 2 . 2 . Materials and methods ....... .. .... .. ............................................... . ..................... ..... ... 2 9 2.2. 1 . Plant material .... ............................................ ............... ....... ....... .............. 2 9 2 .2.2. Seed variability, seed asymmetry and fruit asymmetry ........................... 29 2.2.3 . Mineral analyses ........ .................................................................. ............. 32 2.3. Results ................................................................................................................. 32 2.3 . 1 . Seed variability ............... . . .................................................. ............ .......... 32 2.3.2. Seed asymmetry ............................................. ................. ..... .................... 34 2 . 3 . 3 . Fruit asymmetry . ............... . . ...... ........................ ........ . .............................. 3 5 2 . 3 .4. Fruit minerals . ............ ..................... ... ............. ........ . ........... ...... . .......... .... 3 7 2.4. Discussion ................................................. .......... ........ ........................... ................. 3 8 2.4. 1 . Seed variability ................................. . ....................................................... 3 8 2.4.2. Seed asymmetry . ............ .... .......................................... . ......... ......... ......... 40 2.4.3. Fruit asymmetry ................. ....................................... ..... .......................... 40 2.4.4. Fruit minerals .................................. .......... ............................................... 42 2 . 5 . Conclusions ..................................................................................................... .... .... 43 CHAPTER 3. XYLEM DEVELOP MENT IN TH E STALK OF APPLE AND TH E EFFECTS OF AUXI N TRAN SPORT INH IBITION vu 3.1. I ntroduction ...................... " ..................................................................................... 44 3.2. Materials and methods ............................................................................................ 45 3.2.1. Plant material ........................................................................................... 45 3.2.2. P reparation and application of auxin transport inhibitor (NPA) ............. . 45 3.2.3. Vessel measurements ... ......... ................................................................... 46 3.3. R esults ............................................. .................................................................... 49 3.3.1. S patial distribution of the vessels ..................... ... ............... ...... ............ .... 49 3.3.2. Frequency:size distribution of the vessels ........... ..................................... 50 3.3.3. Development of the stalk .......................................................................... 52 3.3.4. Effect ofNPA on fruit abscission . ................ ........................................... 54 3.3.5. Effect ofNPA on seed and fruit development ......................................... 55 3.3.6. Effect ofNP A on vessel differentiation ................................................... 56 3.4. Discussion ................................... ................ . ....... .................................................... 59 3.4.1. Vessel measurements and estimated conductance ....... ............................ 59 3.4.2. S patial distribution of the vessels ....... ........ .............................................. 59 3.4.3. Frequency:size distribution of the vessels ................................................ 60 3.4.4. Development of the stalk .......... ... . .......... . ................................................. 61 3.4.5. Effect ofNPA on fruit abscission .. ... ....................................................... 62 3.4.6. Effect of NP A on vessel differentiation ....................... ............................ 63 3.4.7. Effect ofNPA on seed and fruit development ......................................... 64 3.5. C onclusions ............................................................................................................. 66 CHAPTER 4. CAUS ES AND EFFECTS OF CHANGES IN XYLEM FUN CTIONALITY IN APPLE FR UIT 4.1. I ntroduction . ............................................................................................................ 67 4.2. Materials and methods ............. .. .............................................................. ............... 68 4.2.1. Plant material ....................... .......................... .......................................... 68 4.2.2. Temporal changes in xylem functionality ................................................ 68 Vlll 4.2.3. Exposure time ........................................................................................... 70 4.2.4. Spatial changes in xylem functionality .................................................... 70 4.2.5 . Anatomy of xylem breakage .................................................................... 7 1 4.2.6. Growth of the different fruit tissues ......................................................... 72 4.3. Results ................................................................................................................. 74 4.3. 1 . Exposure time ........................................................................................... 74 4.3.2. Temporal changes in xylem functionality ................................................ 75 4.3 . 3 . Spatial changes in xylem functionality .................................................... 79 4.3 .4. Anatomy of xylem breakage .................................................................... 8 1 4.3 . 5 . Growth of the different fruit tissues ......................................................... 84 4.3 .5 . 1 . Absolute tissue growth (ATG) .................................................. 84 4.3 .5 .2 . Relative tissue growth (RTG) ................................................... 8 5 4.3 . 5 . 3 . Spatial arrangement of petal and sepal bundles in the flesh (PIS) ........................................................................ 86 4.3 .5 .4. The layout of sepal bundle relative to the locule (S-L) ............. 87 4.4. Discussion ............................................................................................................... 88 4.4. 1 . Exposure time ........................................................................................... 88 4.4.2. Temporal changes in xylem functionality ................................................ 88 4.4. 3 . Spatial changes in xylem functionality .................................................... 92 4.4.4. Anatomy of xylem breakage .................................................................... 93 4.4.5 . Growth of the different fruit tissues ......................................................... 94 4.5. Conclusions ............................................................................................................. 96 CHAPTER 5 . AIR AND CELL MAPPING OF APPLE FRUIT 5 . 1 . Introduction ............................................... .............................................................. 98 5.2. Materials and methods ............................................................................................ 99 5 . 2 . 1 . Plant material ........................................................................................... 99 5 .2 .2 . Air volume measurement ....................................................................... 1 00 5 . 2 . 3 . Timings .................................................................................................. 1 02 5 . 2 .4. Spatial distribution of intercellular spaces ............................................. 1 03 5 . 2 . 5 . Cell measurements ................................................................................. 1 04 IX 5 .2 .5 . 1 . Staining ................................................................................... 1 04 5 .2 .5 .2 . Cell dimensions ....................................................................... 1 05 5 .2.6. Adjustment for air loss ........................................................................... 1 05 5 . 3 . Results ............................................................................................................... 1 07 5 .3 . 1 . Timings .................................................................................................. 1 07 5 . 3 .2 . Cell measurements .................................... ............................................. 1 08 5 . 3 . 3 . Adjustment for air loss ................................................... ........................ 1 09 5 . 3 A. Spatial distribution of intercellular spaces ............................................. 1 1 2 5 .4. Discussion ............................................................................................................. 1 1 3 5 .4. 1 . Timings .................................. ................................................................ 1 1 3 5 .4.2. Cell measurements ................................................................................. 1 14 5 .4.2. 1 . Staining '" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 4 5 .4.2.2. Cell dimensions ...... ............................................................... .. 1 1 4 504.2 .3 . Cell shape and cell packing ..................................................... 1 1 5 504.3 . Adjustment for air loss ........................................................................... 1 1 6 5 AA. Spatial distribution of intercellular spaces ............................................. 1 1 7 5 . 5 . Conclusions ........................................................................................................... 1 1 9 CHAPTER 6. GENERAL DISCUSSION AND CONCLUSIONS 6 . 1 . General discussion ................................................................................................ 120 6. 1 . 1 . Seed as a source of developmental stimuli ............................................ 1 20 6. 1 .2. Nature of the developmental stimulus .................................................... 1 23 6. 1 .3 . Character of fruit tissue growth as a detenninant of fruit quality development ...................................................................... 1 24 6. 1 04. Tissue air and cell alignments and fruit storage quality ......................... 1 26 6.2. General conclusions .............................................................................................. 1 27 REFERENCES ............................................................................................................. 1 29 x List of Figures Fig. 1 . 1 . General assessment of tissue growth and major quality attributes of apple fruit ...................................................................................................... 2 Fig. 1 .2. Transverse section of mature apple fruit (Ma/us domestica Borkh.) ................. 3 Fig. 1 .3 . Conceptual model of fruit quality development .............................................. 26 Fig. 2. 1 . Lopsided 'Granny Smith' apple ....................................................................... 30 Fig. 2.2. Transverse section of fruit showing the direction of the longitudinal cuts to separate the five sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1 Fig. 2.3 . Variability in seed weight ........ ........................................................................ 3 3 Fig. 2.4. Excised fruit with seeds ................................................................................... 3 4 Fig. 2 . 5 . Seed asymmetry vs the total number of seed per fruit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 5 Fig. 2.6. Fruit asymmetry in apple ................................................................................. 3 6 Fig. 2 . 7 . Spatial distribution of calcium in 20 lopsided fruit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 7 Fig. 2 . 8 . Spatial distribution of magnesium (A) and potassium (B) in 20 lopsided fruit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 8 Fig. 3 . 1 . Diagram showing the set-up for quantitative video-microscopic analysis ...... 48 Fig. 3 .2. Spatial distribution of the number of vessels (A) and estimated conductance (B) in the stalk of 'Granny Smith' ............................................. 50 Fig. 3 .3 . Fractional contribution by vessel size classes averaged from cross-sections of three excised stalks of 'Granny Smith': (A) for the mean number of Xl vessels; (B) transformed to reflect the mean conductance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1 Fig. 3 .4. Stalk extension growth in 'Royal Gala' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2 Fig. 3 . 5 . Vessel differentiation in the stalk of 'Royal Gala' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Fig. 3 .6. The effect of I mg NPA g-l lanolin (A), 10 mg NPA g-l lanolin (_) and control (e) on the number of vessels (A), and estimated conductance (B) in the stalks of 'Granny Smith' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Fig. 3 . 7 . Number of vessels and estimated conductance contributed by different size classes in the stalks of 'Granny Smith' treated with 1 0 mg NP A g-I lanolin (A, C) and control (B, D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8 Fig. 4. 1 . Dye infusion of 'Granny Smith' (left) and 'Braebum' (right) apples, c. 70 DAFB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Fig. 4.2. Transverse section of fruit showing radial dimensions measured from the fruit axis to the skin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Fig. 4.3. Dye accumulation in the vascular systems of 'Braebum' (A) and 'Granny Smith' (B) as a function of the period of exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Fig. 4.4. Transverse sections of dye-infused 'Braebum' (left) and 'Granny Smith' (right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Fig. 4.5. Dye level in the primary (cortical) bundles of ' Braebum ' (e) and 'Granny Smith' (0) as a function of fruit age (DAFB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Fig. 4.6. Dye level in the dorsal (A) and ventral (B) carpellary bundles of 'Braebum' (e) and 'Granny Smith' (0) as a function of fruit age XlI (DAFB) ............................................................................................................ 79 Fig. 4.7. Longitudinal changes in the functionality of xylem in 'Braebum' (A) and 'Granny Smith' (B), assessed at 1 60 and 1 63 DAFB respectively . . . . . .. . . . 80 Fig. 4.8. A dysfunctional primary bundle in 'Braebum' ( 1 4 1 DAFB) . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1 Fig. 4.9. Longitudinal sections of a primary (sepal) bundle in the fruit of 'Braebum', 85 DAFB (A-D) and 1 4 1 DAFB (E-F) respectively . . . . . . . . . . . . . . . . . . . . 83 Fig. 4. 1 0. Absolute growth of various tissues in the fruit of 'Braebum' (A) and 'Granny Smith' (B) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Fig. 4. 1 1 . Relative contribution of various tissues to the growth of the fruit in 'Braebum' (A) 'Granny Smith' (B) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Fig. 4. 1 2 . Comparison of the relationship between the relative movements of petal and sepal bundles in 'Braebum' (A) and 'Granny Smith' (B) during fruit growth . . . . .. . .... ............................ . ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Fig. 4. 1 3 . Relative increase in the distance between locule tip and the sepal bundle normalised with the corresponding fruit radii in 'Braebum' (e) and 'Granny Smith' (0) ................................................................................. 87 Fig. 4. 1 4. Dye distribution in 'Braebum' (left) and 'Granny Smith' (right) apples infiltrated with dye at 1 69 and 1 72 DAFB respectively . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . 88 Fig. 5 . 1 . Experimental arrangements for measuring fractional air volume . . . . . . . . . . . . . . . . . . 1 02 Fig. 5 .2 . Tissue sampling for air volume and cell measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 04 Xlll Fig. 5 . 3 . Relative air volume during water replacement (A) and air extraction (B) .... 1 07 Fig. 5 .4. Micrographs of the stained flesh of 'Braebum' fruit ..................................... 1 08 Fig. 5 . 5 . Average scores for relative air volume in blocks of different radial dimensions ............................................................................. , . . . . . . . . . . . . . . . . . . . . . . . 1 1 0 Fig. 5.6. The calculated adjustment for a 'trapezoidal' sample (B9 and BIO) . . . . . . . . . . . . . . 1 1 2 Fig. 5 . 7 . Changes in fractional air content along a radius in 'Braebum' fruit . . . . . . . . . . . . . 1 1 3 Fig. 6. 1 . Causative links between seed set and fruit calcium concentration (the mass of calcium in a particular volume of flesh) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 22 XIV List of Tables Table 3 . 1 . The effect of 1 0 mg NPA g-l lanolin on fruit abscission when applied to the distal (D), proximal (P), and distal and proximal ends of the stalk (D + P) .......................................................................................................... 54 Table 3 .2 . The effect of four NP A concentrations on fruit abscission when applied to the mid-point of the stalk ......................................................................... 5 5 Table 3 . 3 . The effect of 1 0 mg NPA g-l lanolin on fruit and seed development when applied to the distal (D), and proximal (P) ends of the stalk ........................ 55 Table 3 .4. The effect of four NP A concentrations on fruit and seed development when applied to the mid-point of the stalk ............................. ............................... 56 Table 5 . 1 . Air loss in the radial direction (tx) due to water invasion calculated using Eq. 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 1 Chapter J. General introduction Chapter 1. General introduction 1.1. STATEMENT OF THE PROBLEM World production of apples is about 60 million tonnes of which c. 8% (5 million tonnes) is traded internationally. New Zealand exports contribute almost 6% (286,000 tonnes) to the international market although the total production of apples in New Zealand is less than 1 % ( 486,000 tonnes) of world production (Anon., 200 1 ). This emphasises the importance of the quality image of the New Zealand apple in international marketing. Consumers demand consistent and uniform fruit quality whereas apple fruit are prone to a range of quality defects induced during the growth period or after the harvest. These defects are of serious concern to growers. Also, as market preferences can change quite quickly, the New Zealand apple grower must be able to respond. Because the international market is very competitive this imposes considerable pressure to develop strategies that accommodate these changes without impacting profit. 1 To develop the strategies required to efficiently manage a crop, it is necessary to elucidate the origins of quality defects. Knowledge of the factors that lead to the development of a particular quality defect might not only assist in adjusting management practices but might also help identify genotypes at risk. Market-selected quality attributes are, therefore, the primary target, and can only be optimised through an understanding of the key factors that affect fruit physiology. However, the factors associated with development of some fruit quality attributes are either poorly understood or completely neglected. Management practices are often unable to remedy negative impacts or are applied in the final stages of fruit quality development with often small effects but at large expense. Hence, an understanding of the mechanisms underlying physiological responses that affect final quality attributes of an apple is essential to optimising both selection during breeding programmes and management practices. Chapter 1. General introduction Generally, quality assessment of a fruit is based on particular market preferences for particular quality attributes. Hence, fruit quality cannot be simply defined, and further is affected by genetic predisposition, management practices and environment. However, quality criteria generally include major organoleptic and visual preferences such as flavour, texture, appearance, size, shape and storage characteristics (Janick et al., 1 996). Fruit development can be conceptualised as a series of physiological responses of constituent parts that establish the principal quality attributes of the apple fruit. The final characteristics of the fruit could, therefore, be aligned with development of a particular tissue such as proposed in Fig. 1 . 1 . + + Flower + + Skin development Flesh development Vascular development Seed development 1 � t � ~ 1 Skin Fruit Fruit Flesh Storage Eating Seed properties size shape texture quality quality set Fig. 1 . 1 . General assessment of tissue growth and major quality attributes o f apple fruit. 2 Fruit quality is dependent on many aspects that contribute to apple fruit development (Fig. 1 . 1 ). However, the emphasis in the following literature review is on those aspects that were investigated experimentally in this thesis. Of particular interest were the links between the elements of developmental quality paths towards fruit shape, flesh texture, storage quality and seed set. For completeness, other aspects of fruit quality development shown in Fig. 1 . 1 will also be considered. Chapter I. General introduction 3 1.2. STRUCTURE OF APPLE FRUIT 1.2.1. Structural origins of apple fruit The basic anatomy of the apple fruit is well established and in the present review Esau ( 1 977) was used as the general reference base. By definition, the fruit is a product of a matured ovary, although edible parts may also derive from extracarpellary tissues of the flower. An apple is a pome, which consists of two parts that originate from different floral structures but their interpretation initiates two hypothesis that favour different terminology (Pratt, 1 988). The origin of the extracarpellary tissue outside the core line is subject to debate with one view favouring an origin from the flower receptacle (receptacular interpretation), the other an origin from the fused floral appendages (apendicular interpretation). For the sake of convenience, rather than supporting a particular concept, in this thesis the carpellary tissue (ovary) inside the core line will be referred to as the 'core' with a fleshy 'pith' , whereas the extracarpellary tissue outside the core line will be referred to as the ' cortex'. The gross structure of apple fruit in transverse section is shown in Fig. 1 .2. a b (' h cl i f j !; c� k Fig. 1 .2. Transverse section of mature apple fruit (Malus domestica Borkh.). The structure of the fruit is based on 'Braebum' with carpel partitioning adapted from Esau ( 1 977). To clarify the details of some structures (e.g. the ventral bundles) the relative proportions of these tissues have been slightly adjusted. The symbols are: a, skin; b, extracarpellary tissue (,cortex'); c, carpeUary tissue (' core'); d, core line; e, carpel margins;f, cartilaginous endocarp (,locule'); g, seed; h, sepal (primary) bundle; i, petal (primary) bundle;j, dorsal (carpellary) bundle; and k, ventral (carpellary) bundle. - - ------ �------- Chapter J. General introduction 1.2.2. Structure of the skin The apple skin is a multi-layered tissue, comprising epidennal hairs (absent in mature fruit), cuticle, epidennis and hypodennis (Bell, 1 93 7). The cuticle is the non-cellular waxy coating of the epidennis (Miller, 1 982) which provides the maj or transport barrier for the diffusion of penneants (Riederer and Schreiber, 1 995). The epidennis and hypodennis are, however, cellular structures and their unmodified cellulose walls are penneable to water migration (Burton, 1 982). All these features make the skin the ultimate protective structure of the fruit. Furthennore, skin components influenced by environment and commercial practices can induce peculiar changes to the external features of the skin (Pratt, 1 988), that are used for cultivar identification (Bultitude, 1 983). 1.2.3. Structure of the extracarpellary tissue The extracarpellary tissue is made up of fleshy parenchyma penneated with vasculature and intercellular spaces. The major constituents, however, are the parenchyma cells that detennine the overall character of the tissue (Harker et aI, 1 997). Mature parenchyma cells are large, vacuolated, multifaceted fonns with thin, non-lignified walls under internal pressure (Pitt, 1 982). Depending on their position within the tissue, parenchyma cells vary in their wall thickness, size, shape and packing arrangements (Harker et al., 1 997). 1.2.4. Structure of the carpellary tissue 4 The syncarpous ovary is differentiated into a cartilaginous endocarp defining five locules that are embedded in the fleshy part of the pericarp. Each locule has 2 - 4 ovules that become seeds after fertilisation (Pratt, 1 988). The carpels of the ovary become fused along their margins in the later stages of growth (Esau, 1 977), which indicates a degree of physiological autonomy in carpel development. The ovary extends from the centre of the fruit to the core line, which is a distinct sheet of parenchyma cells that demarcates the line of fusion with the adnate extracarpellary tissue. 1.2.5. Structure of the vascular tissue The conducting framework of apple is made of vascular tissues, xylem and phloem. The xylem principally carries water and minerals while the phloem carries organic materials. Chapter J. General introduction 5 Both vascular tissues have highly specialised conduits, consisting of tube-like series of cells arranged at the ends to provide continuous flow. In the xylem, these are vessels, the files of finite length made of dead vessel elements with perforations at the end walls that enable free sap movement within the vessel. Intervascular pitting (bordered pits) on the walls enables sap movement between the vessels. In the phloem, however, sieve tubes are made of living sieve elements with porous plates that connect the modified protoplast of the cells. By analogy with the concept of the xylem path, the phloem source to sink path might be broken up into the number of sieve tubes of finite length (Lang, 1 979). The vasculature in apple is morphologically and physiologically organised into elongated strands known as vascular bundles, whose basic pattern of ramification has been assayed by many workers (McAlpine, 1 9 1 2; Kraus and Ralston, 1 9 1 6 ; MacArthur and Wetmore, 1 939; MacDaniels, 1 940). Vascular bundles are arranged within two distinct systems, ' cortical ' and ' carpellary' , in regard to their position and role within the fruit. The cortical vascular system has ten primary bundles that bend around the drupe-like carpels, giving off a skeleton of traces which ramifies across the flesh. Sepal bundles (opposite the locules) alternate with petal bundles in a ring, somewhat aligned with the core line. The carpellary vascular system is composed of ten ventral and five dorsal bundles that anastomose with each other. The paired ventral bundles emerge from the remaining stele after the separation of primary bundles to form a ring inside the locules, giving off lateral traces to the ovules. Dorsal bundles diverge from primary (sepal) bundles, swinging around the locules and terminating together with ventral bundles in the fused style of the pistil. The vascular system of the fruit includes the stalk (Pratt, 1 988). The stalk of an apple is a stem-like structure connecting the terminal fruit with a short spur. Vessels embedded in non-conducting structural components o f the stalk provide bi-directional sap flow and mechanical support essential for fruit development. However, other than describing the essential structural features (Kraus and Ralston, 1 9 1 6; MacArthur and Wetrnore, 1 939) or growth (Prive et al., 1 98 8 ; 1 989), relatively little work has been done to quantify the dynamics of tissue development in the apple stalk (Barden and Thompson, 1 96 3 ; Habdas et al., 1 982). Chapter 1. General introduction 6 1.2.6. Structure of the seed The seed is the primary organ of the apple as it provides the continuity of the species. The seed develops as the product of a fertilised ovule and differentiates into the embryo enclosed by stored nutrients (endosperm) and seed coat (testa). The embryo contained within the seed originates from a fertilised egg (zygote) and develops into the seedling at the expense of endosperm. The testa is derived from the integuments enclosing the nucellus of the ovule and protects the embryo (mechanically and biochemically) until germination. 1.3. THE ATTRIBUTES OF FRUIT QUALITY 1.3.1. Skin properties The appearance of the skin strongly influences consumer preference while the actual area and structure of the skin are essential for gas exchange. The resistance o f the skin to gas diffusion tends to exceed flesh resistance by around 1 0- to 20-fold, which makes the skin a primary barrier to the permeance of the gases in the apple fruit (Solomos, 1 987). However, during the growth period, apple skin is meristematically active (Tetley, 1 930; Harker and Ferguson, 1 98 8a) and skin permeance to gases (02) increases as the fruit matures (Andrich et al., 1 989). In addition, the permeance of the skin to water vapour regulates the weight loss of apples which is mainly due to water loss (Wilkinson, 1 965; Maguire et al., 2000). This appears to be an important aspect of quality development because inadequate weight loss during fruit ripening can lead to either physiological disorders or skin blemishes (Wilkinson, 1 965; 1 970; Martin et al., 1 967; Scott and Roberts, 1 967, 1 968; Wills and Scott, 1 972; Johnson, 1 976; Hatfield and Knee, 1 988; Maguire et al., 2000). 1.3.2. Fruit size Fruit size is one of the physical growth attributes of a fruit and an important market preference (Opara, 2000). Growth towards final size is accompanied by significant changes in cellular structure that involve increase in cell number (cell division), cell size (cell expansion) and intercellular spaces (Pratt, 1 988). The latter is indicated by a greater ---- ------ Chapter I. General introduction increase in fruit volume than in fruit weight and a decrease in fruit density (Bain and Robertson, 1 95 1; Skene, 1966; Westwood et al., 1 967). Larger fruit are usually softer (Harker et al., 1 997), having higher intercellular air volumes than smaller fruit (Reeve, 1 953; Ruess and St5sser, 1 993; Volz et al., 2002). 7 Since the growth pattern is generally characterised by a short initial period of cell division followed by a long period of cell expansion (Bollard, 1 970), there has been much discussion about the contribution of each stage to fruit size. A number of studies that often involved thinning and crop load effects on fruit size (including alternate bearing) showed high degrees of contradiction. In one view, cell number is the primary determinant of fruit size (Smith, 1950; Bain and Robertson, 1 95 1; Pearson and Robertson, 1 953; Sharples, 1968; Bergh, 1 985; 1990; Goffinet et al., 1 995). Other reports favour cell enlargement as the maj or contributor to fruit size, (Martin and Lewis, 1 952; Sharples, 1 964). However, a combination of both processes prevails under certain conditions or treatments (Smith, 1940; Denne, 1 960; Westwood et al., 1 967; Blanpied and Wilde, 1 968). The significance of air volume on fruit size is also a matter of controversy. Tetley ( 1930), MacArthur and Wetmore ( 1 941 ) and Reeve ( 1 953) claimed that fruit enlargement at the end of the season was mainly due to an increase in the size of intercellular spaces. In contrast, Skene ( 1 966) and Goffinet et al. ( 1 995) suggested that development of intercellular spaces alone was not sufficient to account for the majority of late fruit growth. Nevertheless, in both opinions, the air fraction of the flesh was a significant contributor to the volume of the fruit. It appears that the growth responses of different cellular structures to different environmental and management stimuli are not well understood. Furthermore, the evidence cited above points to the physiological balance that can trigger a separate developmental process in regard to the nature, time period and strength of the stimuli. The assessment of the relationship between fruit size and character of growth is necessary to establish the timing and the nature of cultural practices that will balance size preferences with storage potential. Chapter I. General introduction 8 1 .3.3. Fruit shape Fruit shape is one of the key determinants of quality in apples. It is geometrically defined by a heart-shaped outline extended between two points of arrest (Thompson, 1 966). Shape is a cultivar characteristic (Bultitude, 1 983) and requires visual uniformity within a pack (ENZA, 1 995). A significant effort was made to graphically visualise and identify principal shape traits from a broad population of genotypes, so as to determine the origins of shape defects (Currie et al. , 2000). Among five independent shape traits, ' asymmetric-sides' (lopsidedness) was determined to be mainly due to non-genetic effects as the heritability of the trait was close to zero. Indeed, standard shape defects in lopsided apples includes dropped shoulder at the calyx end (ENZA, 1 995), which was related to seed development (Brookfield et al., 1996). The physiological basis of shape defects can, therefore, include the difference in the distribution of the growth within the developing apple. Skene ( 1 966) showed that longitudinal growth was more pronounced around the calyx end which indicates why the most conspicuous change in tissue growth occurred at this part of lopsided fruit. The asymmetry of lopsided apples indicates specific morphological and physiological features underlying this phenomenon. The ovary consists of five carpels that fuse with one another at a later stage of fruit development (Esau, 1 977). This structural feature could be of great importance in elucidating the nature of shape defects because it implies that the growth of the ovary is modular in nature. This could change the holistic approach to assessing shape defects because the apple could than be viewed as an assemblage whose asymmetry is associated with uneven development of replicate parts. Spatial asymmetry confined to a particular part of the fruit (Heinicke, 1 9 1 7; Simons, 1 965; Way, 1 978) conforms to the modular concept of the fruit. In this thesis, the effect of the modular nature of growth of the ovary on extracarpellary growth and mineral composition is assessed in regard to seed set. Chapter 1. General introduction 9 1.3.4. Flesh texture Harker et al. ( 1 997) concluded that the textural properties of apple are determined by the physical characteristics of the cells and the associated patterns of cell packing. Flesh is comprised of parenchyma cells with non-rigid walls so that tissue strength is the resultant of internal pressure (turgor) that develops in each cell, the properties of cell walls including the nature of bonds between adjacent cells and the network of intercellular spaces (Pitt, 1 982). 1.3.4.1. Cell wall. The primary cell wall is a network of cellulose micro fibrils impregnated by polysaccharides and protein (Knee and Bartley, 1 98 1 ). The structural components of the cell wall can be grouped into the inner and the outer part, the latter, known as the middle lamella, is a gel-like structure where pectic substances provide the contact with adjacent cells (Demarty et al., 1 984). This composition enables cell expansion in rapidly growing fruit (through pectic and hemicellulosic components) and confers strength (through cellulose), which determines the shape of the cell (Harker et al., 1997). The mechanical properties of cell walls (elasticity and rigidity) thus appear essential in determining the textural properties of the fruit. Furthermore, the cell walls are permeated with numerous pores radiating between pectic polymers that enable transport of solutes across the walls (Read and Bacic, 1996). An important structural basis of texture includes the nature of cell-to-cell contact in apple parenchyma. Over half of the cell wall constituents are pectic substances (Jona and Foa, 1 979), mainly galacturan chains with free carboxyl groups available for linkages (Ferguson, 1 984). Calcium is thought to form the cross-links between the carboxyl groups of pectin molecules, both in the middle lamella (Knee and Bartley, 1 98 1 ; Ferguson, 1 984) and the inner part of the cell wall (Demarty et al., 1 984). Reversing the loss of firmness during ripening by infiltrating fruit with calcium confirms the essential role of calcium ions in the structural integrity of the cell wall (Stow, 1 989; Glenn and Poovaiah, 1 990; Siddiqui and Bangerth, 1 996). Fruit ripening is characterised by two major chemical changes in cell wall structure: the fractional increase of soluble pectin, presumably originating from the pectin depositions of the middle lamella (Knee and Bartley, 198 1 ), and the loss of non-cellulosic neutral Chapter 1. General introduction 1 0 sugar components of pectic polysaccharides, namely galactose (Gross and Sams, 1 984). In addition, Ferguson ( 1 984) suggested that calcium could affect postharvest changes in texture either through the structural stability (binding) to the cell wall or through functional interactions with components other than the cell wall. Electron microscopy of ripe apple fruit, however, showed that cell wall breakdown was principally encountered in the middle lamella (Fuller, 1 976; Ben-Arie et al., 1 979; Glenn and Poovaiah, 1 990; Siddiqui and Bangerth, 1 996). These observations suggest that structural changes occurring in the region of cell-to-cell contact are likely to influence the cohesive strength of the parenchyma. Indeed, the loss of cohesion in stored apples was associated with low adhesion between adj acent cells (Glenn and Poovaiah, 1 990; Harker and Hallett, 1 992) such that early season apples usually have lower cell adhesion and become mealy sooner than late season apples of the same age (Khan and Vincent, 1 993). Furthermore, the lower stiffness of the apple parenchyma could be related to the reduced amount of cell wall material in early season cultivars (Khan and Vincent, 1 993). Considering that the walls of adjacent cells are held together by the calcium cross-links with pectic polymers (Knee and Bartley, 1 98 1 ), it seems likely that calcium accumulation could promote the strength of the cell wall network. However, Siddiqui and Bangerth ( 1 996) found no relationship between flesh firmness and the amounts of calcium associated with pectin fractions of purified cell walls. This raised the question of the mode of action of calcium in maintaining the structure and functioning of the cell walls. Saftner et al. ( 1 998) found that the calcium binding capacity increased as the fruit softened which implied that only the ratio of wall-bound calcium to the total binding capacity of the cell wall could be taken to explain the maintenance of fruit firmness. 1 .3.4.2. Cell shape. The shape of cells differs with respect to their position within the flesh and appears closely related to the increase in tissue organisation towards the centre of the fruit. Subepidermal cells are small and periclinally flattened. A marked increase in cell size with depth also brought change to the shape of cells that become spherical in the outer cortex. However, the increase in cell orientation towards the interior is followed by ------------- -- -- --- Chapter I. General introduction 1 1 the gradual change in aspect ratio and the cells become increasingly elongated (Bain and Robertson, 1 95 1 ; Reeve, 1 95 3 ; Khan and Vincent, 1 990; Ruess and Stosser, 1 993). Parenchyma cells tend to be of a distinct 1 4-sided shape, known as tetrakaidecahedra (Lewis, 1 923) that enables the most efficient packing (Korn, 1 984). By analogy with plant cells, physical models approaching this form were established by compressing spherical lead shots in a cylinder (Marvin, 1 939; Matzke, 1 93 9). These assessments showed the importance of contact area and pressure of growth in multi-cellular tissues, although the intercellular spaces that normally would have developed in the tissue were abundant after the compression. However, parenchyma cells in the mature apple are loosely arranged and the area of cell-to-cell contact was estimated to be less than 40% of the overall cell wall area (Vincent, 1 989), which signifies the importance of tissue growth dynamics in cell shape determination. There is no progressive change in the aspect ratio of cortical cells in apple (Lang, unpublished) although the cells became more angular as the fruit mature (Blanpied and Wilde, 1 968). The stability of the cell aspect ratio suggests that uniform physical force acts on the walls of rapidly growing parenchyma cells so that cell edges expand at a constant rate (Korn, 1 980). 1 .3.4.3. Cell packing. Subepidermal cells of apple consist of compact layers of cells, which change gradually towards the interior of the fruit. Progressing inwards, cortical cells initially appear randomly distributed and loosely packed that clearly indicates the presence of intercellular spaces. However, a transition from a random to a radial arrangement occurs deeper within the flesh. The cells become more elongated and organised into radial tiers with a network of markedly elongated intercellular spaces lying in between, which form air channels up to 3 ,000 J..l.m long (Reeve, 1 95 3 ; Khan and Vincent, 1 990). The mechanical properties of apple parenchyma closely reflect the patterns of cell packing. Apple parenchyma is highly anisotropic and heterogeneous (Vincent, 1 989; Khan and Vincent, 1 990) such that the specimen collapses in a single line when compressed radially but fails in shear when compressed tangentially (Khan and Vincent, Chapter 1. General introduction 1 993). Mechanical tests shows that parenchyma stiffness is a function of cell-to-cell contact. Hence, radially orientated cell tiers and intercellular spaces of the tissue are thought to influence the path of the crack (Vincent, 1 989; Khan and Vincent, 1 993). 1 2 I t was proposed that the expanding cells tend to reduce mutual adhesion b y turgor-driven cell rounding (Vincent, 1 992). The same mode of action was thought to increase the volume of intercellular spaces in ripening apples (Hatfield and Knee, 1 988; Harker and Hallett, 1 992). However, the concept of cell rounding requires either progressive increases in cell turgor or decreases in contact area between the cells. The evidence suggests that cells maintain their osmotic properties and burst rapidly when incubated in water but leakage becomes more extensive as fruit mature (Simon, 1 977). Cell shape appears constant during growth while the osmotic concentration of the cell wall free space increases (Lang, unpublished). The alternative view is that the cell-to-cell contact areas do not increase as fast as cell-to-air areas (Lang, unpublished). This may have the result that the intercellular space fraction increases relative to the cell fraction without changes in the pattern of cell packing. 1 .3.4.4. Tissue and air volume of the fruit. The ratio of tissue volume to air volume changes as a result of the collective dynanlics of the physical properties and developmental patterns that take place during fruit growth. It is well established that the volume of intercellular spaces increases as fruit mature (Bain and Robertson, 1 95 1 ; Skene, 1 966; Westwood et al., 1 967; Soudain and Phan Phuc, 1 979; Goffinet and Maloney, 1 987; Harker and Ferguson, 1 988b; Yamaki and Ino, 1 992; Ruess and Stasser, 1993) including during postharvest storage (Soudain and Phan Phuc, 1 979; Hatfield and Knee, 1 988; Harker and Hallett, 1 992; Ruess and Stasser, 1 993). This has the result of increased tissue aeration of developing fruit while the character of intercellular spaces influences the mechanical properties o f the flesh (Vincent, 1 989; Khan and Vincent, 1 993). The intercellular spaces of apples, like the cells, are distributed anisotropically. Progressing inwards, the intercellular spaces showed increased radial orientation that results in formations of elongated air channels, greater than the actual size of voids between the adj acent cells (Reeve, 1 953; Khan and Vincent, 1 990). The change in aspect --------------- - - - Chapter J. General introduction 1 3 ratio from the randomly orientated and roughly isodiametric intercellular spaces o f the outer flesh to the spatially arranged air channels of the interior lying in between the columns of cells may have general implications for the character of gas exchange. These air channels are enriched with C02 emanating from respiring cells and mainly serve to exchange gases with the surrounding atmosphere, either solely (Solomos, 1987) or in a combination with fluid/solid matrix of the flesh (Rajapakse et al., 1 990). This signifies the critical role of the intercellular network in fruit development because the volume, distribution and degree of communication between intercellular spaces determine the physical limits of respiration across the apple flesh. Vincent ( 1 989) claimed that the radial anisotropy of apple prompted specific patterns in communication between intercellular spaces that mainly occurred along the radius with negligible lateral contacts. Hence, a tissue with a small air volume is likely to develop steeper gradients in the concentration of gases along the radius of the fruit. This could affect both the respiratory metabolism of the fruit as well as oral perception of flesh texture. In this thesis, the spatial character of air and cell alignments was assessed in stored apples. A gravimetric technique based on Archimedes ' principle with calculated adjustment for air lost along the sample edges was used for accurate measurements of tissue air content. 1.3.5. Storage quality A major contributing factor to the storage quality of fruit is mineral composition (Ferguson and Watkins, 1 989). In this section, emphasis is given to the relationship between the structure and function of the xylem and how this influences the postharvest condition of the fruit. 1.3.5.1. Vascular development - induction and growth pattern. There is evidence that all of the classical plant hormones may be involved to some extent in complex patterns of vascular differentiation (Aloni, 1 995). However, the emphasis will be given to the role of auxin in controlling xylem differentiation (xylogenesis). -�- -� --� ----�--------- - -- Chapter 1. General introduction 14 The fact that vascular tissues are organised as distinctive longitudinal strands stretching from developing organs to roots points to a possible inductive signal that enables vascular differentiation along a file of cells. Jacobs ( 1 952) demonstrated that auxin (lAA) produced by the young leaves or applied as a synthetic substitute after defoliation o fthe stem promoted basipetal xylem regeneration in Coleus plants. These results implied that xylogenesis was induced and limited by polar auxin movement. Sachs ( 1 98 1 ) concluded that basipetal auxin diffusion through undifferentiated cells resulted in vascular strands that become the preferred path for further auxin flow. Furthermore, the increase in vessel size and decrease in vessel number along the plant axis was associated with a gradual decline in auxin concentration (Aloni and Zimmermann, 1 983). The induction of the two sorts of vascular tissues takes place at different auxin concentrations. Phloem differentiates at low auxin levels, often in the absence of xylem formation (Aloni, 1 980; Aloni and Bamett, 1996), while xylogenesis occurs only at higher auxin levels (Aloni, 1980). Anatomical similarities in the structure of plants grown in similar environments fit with the proposed polar pattern of xylem differentiation. The short growth period characteristic of extreme habitats (desert, arctic and alpine) was thought to evoke vigorous auxin production that results in rapid differentiation of numerous but narrow vessels. However, longer duration of growth (humid tropics) favours lower auxin levels that, over the extended period, produce fewer and wider (therefore more efficient) vessels (Aloni, 1 987). These effects are consistent with the patterns o f vessel differentiation observed in transgenic plants with altered production of endogenous auxin (Klee et al., 1 987; Romano et al., 1 99 1 ). There are two developmental stages of the vascular meristem: the procambium that initiates primary (proto- and meta-) formations and the cambium that initiates secondary formations of vascular tissue. The different forms of wall thickenings characterise specific ontogenetic series. In vessels, annular (ring) thickening occurs during initial development, followed by helical (spiral) thickenings, reticulate (netlike) thickenings and finally pitted elements. The unlignified walls of sieve tubes lack characteristic patterns of thickening. The proto-formations become dysfunctional early in development due to excessive stretching generated by the growth of surrounding tissues. The meta- --------------� � � Chapter 1. General introduction formations remain functional after primary growth but cease to function with the commencement of secondary development (Esau, 1 977). 1 5 1 .3.5.2. Xylem functionality - diurnal and seasonal changes. Xylem function exhibits specific diurnal and seasonal changes. The diurnal growth rhythm of apple fruit occurs as the result of daily cycles (Tukey, 1 964; Jones and Riggs, 1 982; 1 985; Riggs and Jones, 1 984; Tromp, 1 984) where xylem inflow from the tree at night replaces xylem outflow to the tree during the previous day (Lang, 1 990; Lang and Volz, 1 998). Fruit volume not only varies with time of day but depends upon the severity of weather conditions (Jones and Riggs, 1 982, 1 985; Riggs and Jones, 1 984; Tromp, 1 984; Lang, 1 990; Lang and Volz, 1 998). The seasonal pattern revealed an imbalance between the contributions of the xylem and phloem to apple growth. Lang ( 1 990) demonstrated that sap flow into the fruit changed progressively in favour of the phloem and the change was more marked in the cultivar susceptible to the calcium-related physiological disorder, bitter pit. Subsequently, Lang and Ryan ( 1994) concluded that the progressive decline in xylem conductance was due to dysfunction of the vessels within the fruit, and that the decline in xylem conductance was again more severe in bitter pit susceptible cultivar. These findings agree with changes of xylem function observed in fruit of other species (Findlay et al., 1 987; DUring et al., 1 987; Creasy et al., 1 993; Wolswinkel et al., 1 999). A plausible mechanistic explanation predicts the growth-induced rupture of fruit xylem (Findlay et al. , 1 987), even the fragmentation of whole bundles (During et al., 1 987; Wolswinkel et al., 1 999). The physical disruption of the xylem coincided with the sudden expansion of the grape berry (Findlay et al. , 1 987; DUring et ai., 1 987) while the phloem remained functional (Findlay et ai., 1 987). The cellular properties of the berry changed subsequently (Lang and During, 1 99 1 ). These lines of evidence indicate that all inter­ linked effects are serving to adjust for high sink demand in advanced maturity so that the change in fruit mineral composition is the consequence o f the mechanism that controls assimilate partitioning. Chapter 1. General introduction 1 6 In this thesis, temporal and spatial patterns of the xylem function i n apple were assessed by infusing the fruit with dye. In addition, the dynamics of tissue growth and microscopic examination of dysfunctional bundles were carried out to determine the character of changes during the development of the fruit. 1 .3.5.3. Xylem functionality - mineral accumulation and storage quality. The dynamics of growth and function of the conduits appear to determine the mineral status of the fruit, which strongly influences fruit storage quality. This relationship is of particular importance in long-distance transport of calcium and the occurrence of bitter pit. Calcium nutrition is characterised by the limited ability of apple fruit to accumulate and distribute calcium. The xylem is considered to be the principal pathway for calcium transport. This transport is thought to occur with the convective flow of water in this tissue (Ferguson and Watkins, 1 989) although ion exchange between the flowing sap (the mobile phase) and the cell walls (the stationary phase) through which it is flowing will slow the flow of calcium relative to that of water (Marschner, 1 986). Hence, the accumulation of calcium in the fruit will be governed by the hydraulic conductance o f the xylem vessels, the calcium concentration of the sap and the driving force for flow. The potential conductance of a xylem vessel is proportional to the fourth power of the vessel radius, so that wide vessels have a disproportionately large effect on flow (Zimmermann, 1 983). The progressive dysfunction of the xylem in apple (Lang, 1 990; Lang and Ryan, 1 994) will reduce the hydraulic capacity of this tissue to the cross­ sectional area of just the functional vessels and thus reduce the amount of xylem sap entering the fruit. Indeed, most of the calcium is imported early on in fruit development (Wilkinson, 1 968; Jones et al., 1 983), because water flow entering the fruit is diverted via the phloem during the later part of the season (Lang, 1 990). The calcium concentration of the sap is the measure of available calcium for fruit uptake. The common method of mineral assessment includes extraction of the sap from a stem segment (Bollard, 1 953 ; Bradfield, 1 976; Ferguson and Bollard, 1 976; Tromp, 1 979; Jones et al., 1 983; Lang and Volz, 1 998). Xylem sap extracted from the shoot showed the highest calcium concentrations early in the season, before calcium reached maximum Chapter J. General introduction 1 7 concentrations in the fruit, but with a declining trend from then on (Tromp, 1 979; Jones et al., 1 983). More importantly, around half of the calcium present in the shoot sap was in ionic form (Ca2+), while the rest was uncharged or in negatively charged complexes with organic acids (Bradfield, 1 976). These complexes may expedite the mobility of calcium in the xylem (Ferguson and Bollard, 1 976) probably by reducing the degree of adsorption at negatively charged exchange sites in the wall (van de Geij n et al., 1 979). Diurnal cycling becomes more pronounced when the weather favours a high transpirational loss of water (Jones and Higgs, 1 982, 1 98 5 ; Higgs and Jones, 1 984; Tromp, 1 984; Lang, 1 990). This may affect calcium accumulation in the fruit because spur-wood sap carries markedly more calcium than fruit sap while the volume of xylem inflow exceeds that of xylem outflow (Lang and Volz, 1 998). Hence, defoliation of subtending spur leaves reduces calcium accumulation in the fruit (Feree and Palmer, 1 982; Jones and Samuelson, 1 98 3 ; Proctor and Palmer, 1 99 1 ; Volz et al., 1 994; 1 996a), presumably as the consequence of diminished xylem sap cycling (Lang and Volz, 1 998). The significance of calcium movement into the fruit has been extensively studied in relation to the occurrence of storage disorders, such as bitter pit, that has serious consequences for the economics of apple production. Ferguson and Watkins ( 1 989) provided arguably the most comprehensive review on the role of calcium in postharvest physiology o f apple fruit. Poor mineral nutrition and, in particular, low fruit calcium was related to the incidence of bitter pit. The disorder is characterised by discrete necroses in the subepidermal zone of the cortex, and is more frequently located at the calyx-end of the fruit. The localised nature of the disorder complies with uneven calcium distribution in radial and longitudinal directions of the fruit. A number of different factors are thought to affect fruit calcium and development of bitter pit, such as transport routes, diurnal growth, canopy foliage, and fruit growth. Other factors involved in the incidence of bitter pit could be related to genetic predisposition (cultivar susceptibility), climate and soils (growing regions) or crop maintenance (management practices, postharvest conditions). Chapter J. General introduction 1 8 1 .3.6. Eating quality Increases in fruit dry matter content reflect the continued import of assimilates throughout growth. Although phloem-driven, high fruit dry matter content is profoundly influenced by changes in xylem conductivity during growth. The mechanism proposed for grapes states that the xylem breakdown in the berry around the onset of ripening prevents an outflow of sugary sap while the loss of membrane integrity in the berry (compartmentation breakdown) reduces the apoplastic osmotic potential (viz the osmotic potential becomes more negative). This attracts more assimilate into the fruit by reducing sieve tube turgor in the sink region (Lang and Thorpe, 1 989; Lang and DUring, 1 991) . The imbalance between xylem and phloem sap flow i n apple may, i n a similar way, serve to stimulate the flow of assimilates into the fruit. In young apple fruit, the osmotic concentration of the cell wall free space of the flesh (apoplast) is not insignificant and this concentration increases strongly during the season (Lang and Diack, unpublished). The very high values toward maturity must decrease cell turgor and thus will increase phloem sap flow into the fruit (sieve tube sap flow is proportional to the source:sink pressure difference). Although a progressive decline of xylem conductance appears to be an integral part of fruit development (Lang and DUring, 1 99 1 ), through physical disruption of the xylem strand (Findlay et al., 1 987; DUring et al., 1 987; Lang and Ryan, 1 994; Wolswinkel et al., 1 999), there is no evidence that the phloem is able to withstand growth-induced stretching. However, during the latter part of the season, fruit growth is predominantly phloem driven (Lang, 1 990), suggesting unchanged integrity of the phloem structures. The effect of stretching on a vascular bundle will depend on its stress:strain properties and its continued ability to grow. 1 .3.7. Seed set Seed set relies on many internal and external factors. In apple fruit, the number of ovules formed during flower development sets an upper limit for seed number (potential). However, irregularities in ovule and pollen development and unfavourable conditions for pollination can alter the number of seeds (actual). Chapter J. General introduction 1 9 1.3.7.1. Ovule development. The genetically detennined limit for the number and biological characteristics of ovules is affected by various physiological and environmental factors that will impact on fruit setting. Ovule number, however, is more closely related to the dynamics of floral development that appears to be detennined in the previous season (Williams, 1 970). The important characteristics of ovule development are the functional longevity and the fertility potential (Williams, 1 970). Ovule longevity is essential for pollen transfer to stigmas and pollen tube growth in the style to be effective (Stott, 1 972). Extended ovule longevity and the presence of more than one embryo sac in the nucellus of triploid ovules compensate for high ovule sterility and enables heavy cropping in triploid apple cultivars (Williams, 1 970). In contrast, some diploid cultivars show a high proportion of ovule sterility resulting from irregularities in ovule development that are related to erratic fruit set (Howlett, 1 92 8 ; 1 93 8 ; Hartman and Howlett, 1 954; Williams, 1 970; Williams et al., 1 984; Costa Tura and MacKenzie, 1 990). The establishment of vascular tissues in the apple gynoecium is also related to ovule development. Inadequate vascularisation between ovule and placental tissue is associated with open calyx of the fruit, and favoured ovule abortion (Simons and Chu, 1 968). The fact that early xylem lignification in the ventral carpellary bundles coincides with nucellus differentiation in the ovule (Costa Tura and MacKenzie, 1 990; MacKenzie and Costa Tura, 1 99 1 ) indicates a causative link between these events. 1.3.7.2. Pollen development. Commercial apple cultivars are generally self-incompatible and require cross-pollination to set fruit. Insect vectors, such as honey bees, usually carry pollen released from the anthers of the donor tree to the stigmas of the receptor tree (Janick et al., 1 996) so that close proximity of pollen source or hand pollination markedly improves seed and fruit set (Free, 1 962; Free and Spencer-Booth, 1 964; Williams and Smith, 1 967; Maggs et al., 1 97 1 ; Way, 1 978; DeGrandi-Hoffman et al., 1 990; Milutinovic et al., 1 996; Schneider et al., 200 1 ). Similar patterns of distance­ dependant pollen dispersal were also confinned using marked pollen donors (Wertheim, 1 99 1 ; Kron et al., 200 1 a; 2001 b) . Hence, the main characteristics of pollen development that could affect cropping are availability and effectiveness of pollen (Williams, 1 970). Chapter 1. General introduction ------------ - -� - ---- --- --- 20 Pollen availability is determined by the amount of pollen released from the anthers, and strongly depends upon climatic conditions, specifically temperature and humidity (Williams, 1 970). In some cases, the effect of flower morphology on pollination is a matter of controversy as the presence of wide basal gaps between stamens permits honey bees to extract nectar from flowers without touching the stigmas (Roberts, 1 945; Robinson, 1 979; Robinson and Fell; 1 98 1 ; Kuhn and Ambrose, 1 982; DeGrandi­ Hoffinan et al., 1985). This outcome is of particular importance for adequate pollination that appears more dependant on insect than wind-borne dispersal of pollen (Williams, 1 970). The effectiveness of released pollen is characterised by viability, compatibility and growth rate of the pollen tube (Williams, 1970). Viability denotes the germinating capacity of pollen, which is usually high in diploid cultivars (Stott, 1 972; Deckers and Porreye, 1 984). On the basis of pollen tube compatibility, Williams ( 1 970) divided apple cultivars into three groups : invariably self-sterile, invariably self-fertile and variably self­ fertile. Kobel et al. ( 1 939) had assigned compatibility to a single multi-allelic gametophytic locus S, which regulates cropping by preventing inbreeding if the pollen carries the same alleles as the pistil. This phenomenon has been extensively studied since, and many apple genotypes characterised for the presence of S alleles (Manganaris and Alston, 1 987; Sassa et al., 1994; Batlle et al., 1 995; Janssens et al., 1 995; Broothaerts et al., 1 995; 1 996; Sakurai et al., 1 997; 2000; Matsumoto et al., 1 999; Boskovic and Tobutt, 1 999; Goldway et al., 1 999; 2001; Schneider et al., 2001 ). The growth rate of the pollen tube is, therefore, principally governed by the genetically­ controlled compatibility between donor and receptor tissues (Modlibowska, 1 945; Haasbrook et aI., 1 967; Stott, 1972), although temperature may influence this process (Modlibowska, 1 945; Child 1966; Williams, 1970; Stott, 1 972; Visser and Marcucci, 1 983). Furthermore, two successive pollinations where the first 'pioneer pollen' promotes the growth of the second pollen, can produce more seeds in apple (Visser and Verhaegh, 1 980; Visser and Marcucci, 1983; Visser et al., 1 983). This implies that the contribution of later (second) pollen is dominant to seed set. Chapter 1. General introduction 2 1 1 .3.7.3. Seed development. The dynamics of seed development in apple are described in a comprehensive review by Pratt ( 1 988). The development of the seed generally undergoes three stages and closely accompanies that o f the fruit. During the initial stage, the embryo grows slowly while the other fruit and seed structures develop rapidly. At the same time, the first wave of abscission (drop) of fruitlets takes place as a consequence of irregularities during fertilisation. The second stage commences with the change from a free-nuclear to a cellular endosperm (cytokinensis), which enables the rapid growth of the embryo at the expense of the primary endosperm. Secondary endosperm, developed and constantly regenerated by a peripheral meristem, replaces digested primary endosperm and enables continuous growth of the embryo until it attains its full size. Concurrently, the maturing apple also attains its maximum size. The cessation of meristematic activity in the secondary endosperm marks the completion of the third stage of embryo development. The testa hardens and changes colour and the seed enters a period of dormancy in the mature fruit. In many pome fruit, there is a range of seed classes at maturity that are associated with irregularities in seed development. Generally, seeds that have undergone normal development can be categorised as 'viable' and seeds in which abortion has occurred at some stage can be categorised as 'aberrant' (Luckwill, 1 95 3 ; Rohitha and Klinac, 1 990; Brault and de Oliveira, 1 995; Brookfield et al., 1 996; Broom et al., 1 998). 'Aberrant' seeds vary in size and appearance having small and deformed locular traces (rudiments) and expanded but flattened seeds with empty interiors (Rohitha and Klinac, 1 990). Morphological discrepancies among 'aberrant' seeds indicate that changes may have occurred at various stages of growth, such that smaller seeds represent early aberrations in seed development, as noted in grapes (Barritt, 1 970; Ebadi et al. , 1 996). 1 .3.7.4. Seeds in fruit development. Nitsch ( 1 970) emphasised that overall fruit development is primarily dependent on pollination and the resultant seed formation. Stimuli for fruit growth commence with the germination of pollen and continue throughout fertilisation and development of seed structures, which in turn govern the development of the fruit. Several lines of evidence support the principal role of seeds in fruit development. Chapter 1. General introduction 22 The growth of apple fruit is dependent upon the continuous presence of seeds, such that physical damage to the ovules induces severe fruit shedding (Abbott, 1 959; Marini and Byers, 1 988; Ward et al., 1 999), while increases the incidence of fruit deformity and reduces fruit size (Proctor and Schechter, 1992). Hence, an over-seeded apple (more than ten seeds in the fruit) is usually heavy bearing (Simons, 1 974). The final size and shape of apple fruit are, therefore, directly related to the number and distribution of the seeds it contains (Heinicke, 1 9 1 7; Alderman, 1 9 1 8; Murneek and Schowengerdt, 1 935; Latimer, 1 93 1; 1 937; Einset, 1 939; Roberts, 1 946; Denne, 1963; Way, 1 978; Tomala and Dilley, 1 989; 1 990; Brault and de Oliveira, 1995; Brookfield et al., 1 996; Broom et al., 1 998). It appears that seed set must precede fruit growth. Because the stimulus for growth commences with the onset of pollination (Nitsch, 1 970), any fertilised ovule will exert an influence upon fruit growth. Hence, aberrations in seed development at different stages during fruit development will impact differently on the fruit (e.g. grapes, Ebadi et al., 1 996). Indeed, the relationship between seed set and size and shape of mature apple fruit are better explained when ' aberrant' seeds are included (Einset, 1 939; Denne, 1 963; Brault and de Oliveira, 1995). Furthermore, seed development and storage quality appear closely related. Poor seed set or artificially induced parthenocarpy are associated with lower fruit calcium and/or higher incidence of physiological disorders (Heinicke, 1 920; Bangerth, 1 976; Greene et aI., 1 982; Tomala and Dilley, 1989; 1 990; Bramlage et al., 1 990; Brookfield et al., 1 996; Volz et al., 1996b; Broom et al., 1 998). 1 .3.7.5. Seeds as sites of auxin production. Since the biosynthesis of auxin is thought to occur in meristematic tissues (Swamp et al., 2000), an obvious site for auxin production would be developing seeds. Several lines of evidence indicate that this is the case: seeds generally contain higher auxin levels than other fruit tissues (Gustafson, 1 939; Nitsch, 1 950; Varga and Bruinsma, 1 976), increase in seed number stimulates the amount of auxin exported from apple (Bangerth et al., 1 989; Callej as and Bangerth, 1 997; Bangerth, 2000), and terminal ( 'king') fruit which set first export more diffusible auxin than subordinate (lateral) fruit (Gruber and Bangerth, 1990). The earlier appearance and faster growth of terminal-borne fruit is likely to parallel that of seed development, which could Chapter 1. General introduction be taken to explain why 'king' fruit on average produce twice as much auxin while having only a slightly higher seed number than lateral-borne fruit of the same cluster (Bangerth, 2000). 23 It also appears that fruit dominance within the cluster could be mediated by the amount of auxin emanating from the fruit. The concept of 'primigenic dominance' proposed by Bangerth ( 1 989) refers to inter-fruit competition in which the earlier developed fruit is dominating by means of stronger polar auxin export that suppresses the weaker stimuli coming from later setting fruit. Furthermore, the removal of 'king' flowers at the time of full bloom led to a significant increase in diffusible auxin from the remaining lateral fruit and vice versa (Grub er and Bangerth, 1 990). This evidence is supported by the fact that external auxin application can mimic the dominating effect ofthe 'king' fruit in a cluster (Bangerth, 1 993; 1 997). The altered dominance and, therefore, fruit set is a self­ regulating process, which depends upon the strength of the polar auxin signal (Bangerth, 2000). This view is supported by the results from Quinlan and Preston ( 1 97 1 ) that showed an increase in fruit set after the removal of shoot tips. In addition, the importance of polar auxin movement for calcium inflow to the fruit is well established. Banuelos et al. ( 1 987) demonstrated the dependence of ascending calcium transport on a steady auxin flow from the opposite direction. The decreased supply of calcium in fruit treated with auxin transport inhibitors was independent of the transpiration stream and the cation exchange capacity of the fruit. In a number of other studies, the relationship between application of auxin transport inhibitors and calcium deficiencies or the incidence of physiological disorders of the fruit has been observed (Bangerth and Firuzeh, 1 97 1 ; Bangerth, 1 976; Himelrick and Ingle, 1 98 1 ; Stahly and Benson, 1 982; Stahly, 1 986; Bangerth et al., 1 989; Tomala and Dilley, 1 989; 1 990). In some cases, an inhibition of auxin transport markedly suppressed seed development and promoted a physiological disorder (Hamamoto et al., 1 998). Although the physiological mechanisms for the regulation of long-distance transport of nutrients are complex, it appears that auxin production associated with seed development influences calcium uptake by the fruit. - - - - - ------------- - - - Chapter 1. General introduction 24 1 .3.7.6. Cellular mechanism of auxin transport and development of storage quality. Indole-3-acetic acid (IAA) is considered to be the principal fonn of auxin in higher plants that is synthesised in young apical tissues and transported in a polar fashion towards the basal target tissues (Swarup et al., 2000). Auxin has been implicated as a key factor involved in a range of developmental processes including cell division and expansion, vascular differentiation, root initiation, tropistic responses, apical dominance, leaf senescence, leaf and fruit abscission, flower growth, fruit set, growth and ripening and assimilate partitioning (Davies, 1 995). Understanding of the mechanisms regUlating these processes is essential to govern and adjust plant responses influenced by interactions between the tissues and environmental (cultural) growing conditions. The long-standing "chemiosmotic hypothesis" of polar auxin transport proposes a carrier-mediated and energy-dependent process that occurs in a cell-to-cell fashion. The difference in pH is believed to move undissociated lipophilic auxin molecules across the cell membrane (by diffusion) or, to some extent, via specific uptake symport carriers distributed around the membrane. Inside the cytoplasm, auxin (weak acid) dissociates as the result of higher internal pH. A specific efflux carrier drives auxin anions across the cytoplasm by excretion at the basal end of the cell, which provides the polar nature of auxin transport (Lomax et al., 1 995). Because auxin transport in the cell is mediated by influx and efflux carriers, the existence of carrier-specific inhibitors provide a useful tool in understanding the nature of events that takes place during polar auxin movement. ManipUlations with phytotropins, synthetic inhibitors of auxin transport, suggest a high affinity with the efflux carrier (Lomax et aI., 1 995). However, the knowledge of influx carrier function is still scarce, and the recent discovery of a novel class of carrier-specific inhibitors will provide useful tools to investigate the function and properties of the influx carrier (Imhoff et al., 2000). Manipulations with auxin transport inhibitors have shown the relationship between basipetal transport of auxin and acropetal calcium uptake of the fruit appears insensitive to either transpiration or exchange adsorption mechanisms of transport (Banuelos et al., 1 987). This infers that auxin and calcium transport pathways are directly linked, possibly through the mutually dependent interaction that also occurs in the opposite direction Chapter J. General introduction (sunflower, del a Fuente and Leopold, 1 973; sunflower and corn, dela Fuente, 1 984; de Guzman and del a Fuente, 1 984; 1 986; zucchini, Allan and Rubery, 1 99 1 ). 25 Furthermore, the nature of interaction between auxin and calcium could be involved in the character of dominance between the organs. The influx of calcium may facilitate the reciprocal release of auxin that sustains the dominating position of the organ. This will further reduce the supply of calcium and the efflux of auxin from the dominated organs, thus affecting the overall calcium status of these organs (Tamas, 1 995). Hence, the export of auxin from the maj or sites of biosynthesis within the fruit could be essential to the calcium status and thus the storage quality of the fruit. In this thesis, the nature of developmental stimuli is assessed through the effect of auxin transport inhibition on vessel differentiation in the fruit stalk, abscission, and seed and fruit development during the early stages of fruit growth. 1 .4. CONCEPTUAL MODEL OF FRUIT QUALITY DEVELOPMENT The review of the literature provides a solid insight into the origins and the paths of fruit quality development. Based on this evidence, the initial assessment shown in Fig. 1 . 1 was developed into the conceptual model shown in Fig. 1 .3 . The path towards each quality attribute was structured to accommodate physiological processes (text) and factors (ovals) that influence various tissue properties (boxes) along the path. Every element o f the model is arranged in time and function such that some tissue properties can became ' externality factors' across the paths of quality attributes. The potential usefulness of this model is to provide a broad framework for conceptualising the major aspects of fruit growth and development. The model allows for mUltiple interactions at any level of path development, which describe the mutual dependence between the quality attributes. In that way, each element of the model can be assessed for the hierarchical position within the path, which is time dependent. The model could, therefore, be used to indicate the origins of physiological defects in apple fruit and provide mechanistic links between the elements involved. Chapter J. General introduction 26 11 � � � " '" � S Q. " 0 <> I:b � !i. i � .a lj � -i " 5: � 'Q � .::: � :;: r;;:; ] Ii ... '!/j G &. .� '" '" � i ] i:: '" .." � " :.!il I/) Fig. 1 .3. C onceptual model of fruit quality development. The symbols are connecting the interactions between processes (text), factors (ovals) and properties ( boxes) (+), and hormonal influences on tissue development (u). Chapter I. General introduction 27 1 .5. RESEARCH OBJECTIVES OF THE THESIS The aim of this thesis was to investigate the relationship between structure and function in different fruit tissues that impact on fruit quality development. The view was taken that some elements of developmental quality paths have not been fully examined in the light of current knowledge and their implications for both fruit physiology and apple industry have not been considered. More specifically, the research obj ectives were: • To determine the impact of seed set on fruit shape and fruit storage potential by recording its impact on modular flesh development and mineral distribution; • To determine the nature and impact of developmental stimuli on fruit structure by manipulating auxin efflux from the fruit and recording vessel differentiation in the stalk; • To determine the impact of different aspects of flesh growth and vascular morphology on xylem functionality during the season; and • To develop an anatomical basis for the gas-exchange properties of apple tissue using a novel technique to determine the amount and disposition of intercellular air and, also, the shape and packing of the flesh-cells. The primary research aims outlined above are the focus of the experimental studies presented in Chapters 2-5. As the experimental studies presented in this thesis differed from one another, a conceptual model of apple fruit growth was developed based on the literature (Fig. 1 .3). The model identifies and links the various developmental and environmental elements that contribute to the overall development of fruit quality. In this way, each study falls within a path of the model which links to a particular quality attribute of the fruit. Chapter 2. Seed set, fruit shape and fruit minerals 28 Chapter 2. Seed set, fruit shape and fruit minerals 2.1. INTRODUCTION The apple (Malus domestica Borkh.) is one of New Zealand' s most important fruit crops with annual production of 486,000 tonnes of which c. 60% were exported and valued at NZ $339 million (fob), (Anon. , 200 1 ). International markets demand quality and consistency including uniform size, shape and blemish-free fruit. Significant quantities of fruit fail to meet critical shape criteria but the magnitude of these losses is unknown. Some calculations based upon losses in the packhouse suggest that c. 2 .4% of the annual crop is rejected (Palmer, pers. comm.). This figure does not include lopsided fruit rejected by pickers in orchards. The development of lopsided apple fruit has been attributed to poor and asymmetrical seed set (Heinicke, 1 9 1 7 ; Latimer, 1 93 1 , 1 937; Roberts, 1 946; Way, 1 978; Brault and de 0liveira, 1 995; Brookfield et al., 1 996). Furthermore, a number of studies indicate a relationship between seed set and fruit calcium (Bramlage et al. , 1 990; Tomala and Dilley, 1 989; 1 990; Brookfield et al., 1 996; Volz et al., 1 996b; Broom et al., 1 998), and the occurrence of physiological disorders (Heinicke, 1 920; Bramlage et al., 1 990; Brookfield et al., 1 996). However, the rational relationship between seed set, fruit shape and fruit calcium (as distinct from a simple correlation) has not been properly examined either mechanistically or quantitatively. Since the pomaceous ovary is made up of five fused carpels (Esau, 1 977), the apple fruit can be viewed as an assemblage of replicate parts - fruit shape asymmetry suggests that each carpel tends to develop as a unit. If this is the case it is likely that the developing seed may have a localised effect on fruit development so that each carpel contributes somewhat autonomously to the overall fruit shape and calcium concentration. Therefore, the obj ectives of this study were to quantify seed variability, quantify seed asymmetry, relate seed set to fruit asymmetry, and relate fruit asymmetry to the spatial distribution of minerals. Preliminary data and analyses have been published (Drazeta et al., 2000). Chapter 2. Seed set, fruit shape and fruit minerals 29 2.2. MATERIALS AND METHODS 2.2.1 . Plant material The study used mature trees of ' Granny Smith' apples grown at the Massey University Fruit Crops Unit, Palmerston North, New Zealand. Trees were planted on MM 1 06 rootstock and managed according to standard commercial practice including the usual calcium sprays. A total of 1 80 fruit were randomly collected at commercial harvest. Fruit were stored in a cool-store (0.5°C, c. 95% RH) for up to 8 months pending analysis. Day­ to-day storage was in a cold room (2°C, no humidity control). 2.2.2. Seed variability, seed asymmetry and fruit asymmetry To assess variability in seed weight, a total of 1 20 fruit without blemishes were processed. Fruit were weighed, then halved and the number and individual weights of the seeds were recorded. In addition, spatial variability of seed weights in the locules was obtained from a sample of 60 lopsided apples used for the assessment of fruit asymmetry. In each locule, seeds can arise through fertilisation of either or both o f two ovules. giving four possibilities - locules can contain (0,0), ( 1 ,0), (0, 1 ), or ( 1 , 1 ) seeds. With five locules, there are, therefore, 45 or 1 ,024 possible seed arrangements. To quantify the asymmetry associated with uneven seed distribution an index of seed asymmetry (1) was calculated for each arrangement. A value was calculated for I, for all possible seed arrangements amongst the five locules. To do this the fruit was treated as if it were a five-spoked wheel held in the horizontal plane with a number of uniform weights (corresponding to the seeds) variously distributed around its margin in line with the spokes. The asymmetry of any particular seed arrangement can be measured by the extent to which the centre of gravity of the wheel falls away from the axis. The index of asymmetry I is then quantified as the radial displacement of this imaginary centre of gravity from the geometrical centre. By analogy with the wheel : ( 1 ) and (2) Chapter 2. Seed set, fruit shape and fruit minerals 30 from which, by Pythagorus, / = vf"x2 + /) (3) where x and y are the orthogonal displacements of the centre of gravity from the geometric centre, being calculated by taking moments for seed masses mj at radial distances Ij arranged at angles ai. M is the total seed mass, or 'f.mj. Assuming h = 1 while m[ is equal for all seeds, than 1 can range from 0 (perfect symmetry) to 1 (maximum asymmetry) . To quantify fruit asymmetry and spatial mineral distribution, 60 fruit (previously used to assess spatial variability of seed weights in the locules) were processed. The effects of uneven seed distribution were maximised by selecting only lopsided fruit. The most common form of fruit asymmetry was expressed as a severely dropped shoulder at the calyx end to make an angle greater than 1 5° (Fig. 2 . 1 ) . Irregular enlargement of the base at the stem end with oblate fruit shape was also apparent. Fig. 2. 1 . Lopsided 'Granny Smith' apple. Asymmetry exceeds 1 5° divergence from the vertical axis at the calyx end. The fruit does not meet the 'Granny Smith' grade standard for class 1 apple (ENZA, 1 995). Chapter 2. Seed set. fruit shape and fruit minerals 3 1 Apples were weighed and cut in half transversely near the equator to expose the locules without damaging the seed. The fruit halves were each carefully sliced longitudinally into five sectors along the carpel margins passing through the primary (petal) bundles and adnate (extracarpellary) cortex (Fig 2.2). The weight of each individual sector (recombining the two portions of each sector previously separated by the transverse cut and comprising of a locule with seeds, flesh with skin as shaded in Fig. 2 .2) and of the seeds alone were recorded. Sub-samples for mineral analysis were taken from the middle of each sector using a cork borer, weighed, promptly sealed in polyethylene bags and frozen at -2oDe pending analysis. Fig. 2.2. Transverse section of fruit showing the direction of the longitudinal cuts to separate the five sectors. The weight of the sector (shaded) was related to the weight of all of the seeds as combined by a three-order model. To relate actual seed weight to the sector weight, a three-order model was developed. In this model, the weight of a sector is related to the weight of the seed within it (as a first­ order effect), to the weight of the seed in the two adj oining sectors (second-order) and to the weight of the seed in the two distant sectors (third-order), (Fig. 2.2). The model is expressed as: S = a + b(wJ + C(Wi-l + Wi+l) + d(wj-2 + Wi+z}, (4) where S is sector weight, a is the intercept and b, c and d are the coefficients for the seed weights Wj in the various sectors. To test the data for the three-order model, analysis was � - - - �-------� Chapter 2. Seed set, fruit shape and fruit minerals carried out using SAS Proc Mixed with a Toeplitz correlation structure (SAS, 1 996), to allow for the expected correlation between data from different sectors in the same fruit. Separate intercepts were fitted for each fruit. 2.2.3. Mineral analyses 32 For each lopsided fruit, goodness of fit to the model (eq. 4) was estimated by calculating the ratio of the sum of the squared deviations about the model estimates to the sum of the squared deviations about the model sector weight. A subset of 20 fruit ( 1 00 samples) was selected for mineral analyses by choosing those fruit for which the sector weights best fit the model but omitting those fruit where the diversity of sector weights was too small . Mineral analyses were carried out by atomic absorption spectrophotometry and spatial distribution of calcium, magnesium and potassium concentrations were evaluated using regression analysis. However, as the size of analysed fruit differed markedly ( 1 34- 1 96 g), fruit effects, as determined by the analysis of variance, were removed before carrying out the regression. 2.3. RESULTS 2.3.1. Seed variability Seeds were described as ' rudimentary' (� 1 0 mg), ' empty' (c. 1 0-30 mg) or 'viable' (c. 40- 1 00 mg), (Fig. 2.3A). 'Rudimentary' and ' empty' seeds together were described as ' aberrant' . ' Rudimentary' seeds were characterised by completely arrested development, shrivelled testa and deformed shape, and consisted mainly of very small seeds (� 2 mg). 'Empty' seeds were bigger than 'rudimentary' seeds but the interior was resorbed giving the seeds a flattened appearance with testa curving inwards. Such seeds notably varied in their size. In contrast, 'viable' seeds were fully developed with a plump interior and weighed considerably more. The frequency of 'viable' seeds showed a normal distribution with a wider range of weights than for ' aberrant' seeds. Chapter 2. Seed set, fruit shape and fruit minerals 33 The frequency distribution of seed weights in the locules of lopsided ' Granny Smith' showed marked variability (Fig. 2.3B). A relatively small number of locules contained only 'aberrant' seeds (� 20 mg) and these were prevalent in the underdeveloped parts of lopsided fruit. The total seed weight in a locule within the range of c. 30- 1 00 mg usually represented one 'viable' and one ' aberrant' seed, while locules with total seed weight of over 1 00 mg usually contained two 'viable' seeds and these were associated with larger sector development (Fig. 2.4). 1 50 30 A . 1 25 25 1 00 20 :» C) !::i 75 1 5 � ;:I er � � 50 1 0 25 5 0 0 0 20 40 60 80 1 00 1 20 0 25 50 75 1 00 1 25 1 50 1 75 200 Individual seed weight (mg) Seed weight per locule (mg) Fig. 2.3. Variability in seed weight. (A) Frequency distribution of individual seed weights from 1 20 fruit. Each bar represents 2 mg increments in seed weight. (B) Frequency distribution of combined seed weights in an individual locule from 60 lopsided fruit. Each bar represents 5 mg increments in seed weight. Chapter 2. Seed set, fruit shape andfruit minerals 34 • , , , . • • Fig. 2.4. Excised fruit with seeds. The size of individual sectors relates to the seed size. 2.3.2. Seed asymmetry The theoretical asymmetry of seed arrangements in the fruit is shown in Figure 2 . 5 . The line was obtained by joining the means of I for each seed arrangement for a number of seeds in a fruit, weighting each data point by the number of seed arrangements that gave rise to it. The fewer seeds the higher the value of I. This reached a maximum (I = 1 ) in fruit with one seed, and in one arrangement in fruit with two seeds (both seeds in the same locule). In contrast, a perfectly symmetrical seed distribution (J = 0) occurred only for fruit with no seeds or with ten seeds and one arrangement in fruit with five seeds (having one seed in each locule). For all other seed arrangements asymmetry generally increased as seed number decreased. Chapter 2. Seed set, fruit shape and fruit minerals 1 .0 0.9 0.8 S 0.7 B I 0.6 � 0.5 ] w 0.4 '-M o >< 0.3 ] 0.2 0.1 0.0 • • • • 0\" ; : • • • ."'-0 ......... • • e • • . .� . • • "0 ......... · . . . . \ o 1 2 3 4 5 6 7 8 9 1 0 Number o f seeds 3 5 Fig. 2 . 5 . Seed asymmetry vs the total number o f seed per fruit. Results calculated for all possible combinations of seed number and seed arrangement. The line connects means (0) from the population of I (e). 2.3.3. Fruit asymmetry Analysis of the data partially confinned the three-order model. Sector weight was found to be principally influenced by its own seeds (b = 75 .36; p � 0.00 1 ), less so, but still significantly, by the seeds in the two flanking locules (c = 1 8 .97 ; p � 0.05) but not significantly by the seeds in the two distant locules (d = 0.2 5 ; p = 0.97). The contribution of fruit effects to the variance was found to be highly significant (p � 0.00 1 ), but the Toeplitz variance components were not significant. Chapter 2. Seed set, fruit shape and fruit minerals 36 A refined seed weight model with d = 0 was therefore fitted with separate intercepts for each fruit, but no Toeplitz structure. This seed weight model gave very similar values of b and c (75 .56 and 1 8.86 respectively). Regression analysis shows a strong relationship between the combined seed weights (for the sectors own seeds plus those from the two flanking locules) and individual sector weight adjusted to the same average intercept (Fig. 2.6). The slope is significant (p :s; 0.00 1 ) while the fit (r = 0.42) suggests that the variation is partially explained by the model remaining variability is probably the result of a range o f factors such as vigour, crop load, position in the canopy. 55 50 45 40 35 -0/) '-' � 30 ..... Cl) � 25 .... El <) Cl) 20 r.f) • • 1 5 1 0 5 0 0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 Seed weight model Fig. 2.6. Fruit asymmetry in apple. The influence of seed weight on sector weight as combined in the seed weight model (from the sectors own and the two flanking locules). Each data point represents an individual sector weight. Chapter 2. Seed set,jruit shape andjruit minerals 2.3.4. Fruit minerals The mineral composition of lopsided 'Granny Smith' showed different distribution patterns for different minerals. The concentration of calcium declined with increasing sector weight (p � 0.05), (Fig. 2.7), while the concentrations of magnesium and potassium were not significantly affected (Fig. 2.8). 2.8 2.7 i 2.6 /;)l) 0 0 2.5 - /;)l) S '-' s:: 2.4 0 '':;:: tl s:: 0 u 2.3 s:: 0 u S =' 2.2 'u � U 2.1 2.0 20 25 30 35 40 45 Sector weight (g) Fig. 2.7. Spatial distribution of calcium in 20 lopsided fruit. For clarity, each data point represents the mean of a successive grouping of ten values. Vertical bars represent SE. 37 Chapter 2. Seed set, fruit shape and/ruU minerals 2. 3 :i 2.2 ' 1:iIJ 8 2. 1 - I:iIJ E 2.0 '-" � .g Cd 1 .9 g g 1 .8 o U 1 .7 A . 2 0 2 5 30 35 40 45 Sector weight (g) 82 80 78 76 74 72 70 68 B. 20 25 3 0 3 5 40 45 Sector weight (g) Fig. 2.8 . Spatial distribution of magnesium (A) and potassium (B) in 20 lopsided fruit. For clarity, each data point represents the mean of a successive grouping of ten values. Vertical bars represent SE. These regressions showed no significant slope at p � 0.05 . 2.4. DISCUSSION 2.4.1 . Seed variability 38 The three types of seeds described for ' Granny Smith' are similar to those described for other apple cultivars (Brault and de Oliveira, 1995; Brookfield et al., 1 996; Broom et al., 1998). The variability in the size of ' aberrant' seeds probably indicates that seed abortion occurred at different developmental stages, as suggested for grapes (Barritt, 1970; Ebadi et al., 1996). For the collapsed ' rudimentary' seeds abortion presumably occurred at the earliest stage of development. However, in this study, no histological analysis was carried out to determine whether the trace formations in lopsided ' Granny Smith' represent aborted (unfertilised) ovules or aborted (fertilised) seed. Locular traces were regarded as seed rudiments based on their microscopic appearance with the desiccated Chapter 2. Seed set, fruit shape and fruit minerals 39 testa-like surface indicating that fertilisation of the ovule may have take place even though the testa was incomplete (Ebadi et al. , 1 996). From a botanical standpoint, a true seed is a fertilised ovule with a developed embryo, stored nutrients and testa (Abercrombie et al., 1 98 1 ). Hence, ovular structures that abort after fertilisation can be still regarded as seeds. Abortion at a later stage of development presumably leads to the formation of ' empty' seeds since the main body of the seed (as measured by growth in length) is attained within 4-5 weeks after petal fall (Luckwill, 1 953; Sato and Kanbe, 1 984). On the other hand ' empty' and 'viable' seeds could attain similar size but the variability in weight is considerable. The difference is the result of a degenerated interior in an 'empty' seed - mainly the lack of an embryo and endosperm (Olmo, 1 934; Stout, 1 936; Luckwill, 1 953; Ebadi et al., 1 996) that occupy the largest volume of the seed (Esau, 1 977). Because most of the locules in lopsided ' Granny Smith' apples were fully seeded (containing two seeds), any aberrations in seed development create a considerable discrepancy in seed weights. A locule having a single large seed can have a similar weight of seed as one with two small seeds. As a result, only a seed weight distribution of over 1 00 mg is indicative of normal seed development in a locule (Fig. 2 .3B). The remaining locules with lesser seed weights were likely associated (to a different degree) with irregularities in seed and flesh development (Fig. 2.4). Seed number indicates the degree of fertilisation in the fruit but seed weight reflects the dynamics of post-fertilisation development. Indeed, seed weight rather than seed number was better correlated with fruit shape in nashi pear (Rohitha and Klinac, 1 990) while the array of seed classes (including aberrant seeds) influenced apple fruit shape (Brault and de Oliveira, 1 995). Hence, the effect of seed on fruit growth should be assessed within the same seed class based on the morphological appearance of the seed. Considering the role of seeds in sector development of ' Granny Smith' (Fig. 2.6) it seems likely that larger seeds generate a stronger developmental signal. Chapter 2. Seed set, fruit shape and fruit minerals 40 2.4.2. Seed asymmetry The index of seed asymmetry (1) quantifies the deviation of a particular seed arrangement from a perfectly symmetrical distribution. Fruit with fewer seeds will generally have a more asymmetrical seed arrangement, which is reflected as a shape defect because flesh development is less near locules with 'aberrant' seed (Fig. 2.4). In practice, Latimer ( 1 937) showed that apple fruit with one or two seeds were lopsided whereas fruit with no seeds developed a symmetrical fruit shape. This conforms to the predicted trend of seed asymmetry where I increase steadily with decreasing seed number, but suddenly drops when fruit contained no seed (Fig. 2.5). Hence, the mathematically generated I indicates the actual effect of seed set on fruit shape. The theoretical model assumes identical seed masses (mi) distributed around the five spokes (locules) of the light wheel (fruit) but a wide range of seed number. However, in practise, seed weights varied considerably (Fig. 2.3B) but the average number of seeds in lopsided 'Granny Smith' was fairly high (8.63 ± 0.2). However, 'aberrant' seeds are little more than aborted ovules and if omitted from seed asymmetry assessment on the basis of having a small influence on fruit development, the omission could create the range of seed distribution seen in Fig. 2.5. In applying the theoretical model, an ' aberrant' seed should be scored as 'no seed' and all 'viable' seeds should be assumed to be of equal weight. Since variability in seed weight has not been included in the ' light-wheel' model, the distribution of I represents solely the property of uneven seed arrangement. 2.4.3. Fruit asymmetry Sector weight was influenced by the weight of the associated seeds as combined in the seed weight model (Fig. 2.6). A significant physiological effect on the growth of a particular sector occurred not only from the seeds in the sector's own locule (first order) but also from those in the two flanking locules (second order). Clearly, fruit asymmetry is the result of unbalanced sector growth associated with an asymmetrical seed weight distribution. This indicates that the course of seed development, as shown by the variability in seed weight, and seed distribution are the principal determinants of fruit asymmetry. Failure to form a symmetrical shape early in the season has previously been attributed to 'ovule' abortion (Simons, 1965). Chapter 2. Seed set, fruit shape and fruit minerals 4 1 For many years the asymmetry of pome fruit was simply correlated with uneven seed set (Heinicke, 1 9 1 7; Latimer, 193 1 , 1 937; Roberts, 1 946; Way, 1 978; Brookfield et al. , 1 996). Other attempts were made to quantify fruit asymmetry by assessing fruit sector dimensions through the carpel (Rohitha and Klinac, 1990) or by calculating an asymmetry index from radial measurements taken along the carpeUary margins (Brault and de Oliveira, 1 995). In both studies the relationship between fruit asymmetry and seed set was investigated at the carpel level although they did not view the fruit from a modular standpoint. Furthermore, their results were based on one- or two-dimensional measurements, whereas the tissue weight assessed in this study is a more integrated parameter that effectively gives a three-dimensional perspective (i.e. tissue volume) of tissue development. An additional benefit of using weight as the measured variable (rather than radius or height) is that weight can be recorded with very high accuracy. Based on the seed weight model (Fig. 2.6), seed set influences fruit shape in apple via an influence on fruit sector development. This relationship supports a modular view of an apple as an assemblage of five replicate sectors (extended carpels), where each develops as a unit although influenced to some extent by the immediately adjacent units. The morphology of the apple fruit itself supports such a modular view. The ovary (core) is composed of five carpels fused along the margins to include a locule with seeds and a fleshy pericarp while the extracerpeUary tissue (cortex) encloses the ovary and forms the majority of the fruit flesh (Esau, 1977). Clearly, the core structure implies autonomous growth at the level of the single carpel. However, the fruit sector margins were extrapolated through the outer cortex on the basis of the positions of the carpeUary partitions (Fig. 2.2). They do not represent a strict morphological division of the extracarpellary tissue. Seeds clearly influenced the development of the two flanking sectors indicating that their principle affect is in the centrifugal direction (towards the skin) a first-order effect. Their radial influence is somewhat weaker - a second order effect. ------------- �. �- Chapter 2. Seed setJruil shape and/ruil minerals 2.4.4. Fruit minerals 42 The distribution of minerals within the fruit was affected by the distribution of fruit growth. The stronger growth of the larger sectors was associated with lower concentrations of minerals (Figs. 2.7 and 2.8), but only that of calcium was significantly affected. The consequence of such a pattern of mineral composition is a changing ratio of calcium to magnesium and potassium that could promote the occurrence of calcium­ related physiological disorders, such as bitter pit (Ferguson and Watkins, 1 989). The data provide a plausible physiological link to increased frequency of bitter pit lesions in larger and well seeded parts of lopsided apple fruit as first suggested by Heinicke ( 1920). Spatial trends in fruit mineral status and bitter pit development somewhat contradict the mechanism underlying the relationship between seed set and fruit calcium. If seed set directly influences calcium uptake into fruit (Tomala and Diley, 1 989; 1 990; Bramlage et al. , 1 990; Broom et al., 1 998), it follows that fruit mineral concentration could be maximised simply by improving seed set. However, the overly simplified explanation (more seed -+ more calcium) does not fully account for the significance of other factors involved in fruit mineral nutrition. For example, the larger sectors usually have a full complement of seed, yet calcium concentration declines with increased sector weight (Fig. 2.7). Bangerth ( 1 973) considered fruit size was one of the factors secondary to fruit calcium in the development of bitter pit. In apples, the xylem sap, as the principal carrier of calcium (Marschner, 1 986), undergoes a progressive decline in flow rate relative to the phloem (Lang, 1 990). The changing balance in transport pathways favours the accumulation of phloem-borne minerals, while reducing the amount of available calcium in the tissue. In such a way, the ratio of calcium to magnesium and potassium becomes less favourable in rapidly growing or larger fruit (Ferguson and Watkins, 1 989). By analogy, the uneven spatial development in lopsided fruit could lead to relatively lower calcium concentrations in the most vigorously expanding tissues, which in turn may lead to a localised occurrence of bitter pit, presumably more often confined to the bigger sectors. Chapter 2. Seed set, fruit shape and fruit minerals 2.5. CONCLUSIONS Seed weight was variable and the distribution of seed weights tended to follow their morphological appearance, presumably as the result of aborted seed development. Because seed formation precedes fruit set, seed variability is reflected in two important attributes of quality, namely fruit shape and fruit mineral composition. 43 The nature of fruit asymmetry itself is closely related to the morphology of the apple fruit so that each carpel contributes somewhat autonomously to overall fruit growth. Sector growth is principally influenced by the complement of seed in its own locule (first order) and to a lesser extent by the seed in the two flanking locules (second order). The distribution of seed weight appears to be a key in defining fruit symmetry with 'aberrant' seeds generally confined to the poorly developed sectors, while fully formed seeds prevail in the bigger sectors. Consequently, as seed weight affects sector weight, seed arrangement affects fruit shape. Asymmetry in seed distribution increases when seeds are few. Variations in spatial fruit development result in variable mineral composition. A change in the ratio of essential fruit minerals was associated with uneven growth. Calcium concentrations in the larger fruit sectors were lower than the calcium concentrations in the smaller fruit sectors while magnesium and potassium concentrations were largely unaffected. This pattern favours calcium deficiency and the occurrence of physiological disorders in the fruit. Seed set, either directly or indirectly, is thus one of the key factors affecting final mineral status of the fruit. Although fruit quality depends on various growth factors, improved seed set (attained through good pollination) provides a relatively inexpensive and efficient practice to improve fruit quality. However, seediness could be less responsive to pollination in apples prone to irregularities in pollen and ovule developments. This suggests that a more comprehensive study would be worthwhile to quantify possible genetic differences between apple cultivars. - - -- - ----- - Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 44 Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 3.1. INTRODUCTION The fact that xylem tissue differentiates as a strand of elongated cells, with a gradual change in diameter and density of vessels along the plant axis, suggests that an inducting stimulus moves in a polar fashion between source and sink. Since Jacobs ( 1 952) demonstrated that the auxin from developing leaves in Coleus was the limiting factor for xylem regeneration in the stem internode, it has been generally accepted that auxin plays a maj or role in controlling xylogenesis. The young developing organs are regarded as the primary sites of auxin synthesis and it is the steady polar flow of auxin towards the roots which is thought to promote development of conductive tissues (Aloni et al., 2000). The maj or operating conduits of the xylem are vessels (Esau, 1 977) which, due to the existence of open perforation plates at the end walls of the vessel elements, provide a low-resistance pathway for xylem sap (Nobel, 1 999). The measured hydraulic conductances in several dicotyledons with different perforation plates were close to estimated values based on Hagen-Poiseuille 's law (Schulte et al., 1 989). For this reason, the relative flow rate (efficiency of conductance) depends upon the cross-sectional area of the xylem capillaries (Zimmermann, 1 983). The pattern of vascular differentiation, which determines relative frequency:size distribution of vessels, is thus a critical factor in the hydraulic architecture of a particular organ. Cross-sectional area and the length of time when the vessels are functioning could be of particular importance to mineral nutrition, and thus keeping quality, of fruit such as apples (Lang and Ryan, 1 994). This study was undertaken to determine the structural changes of vessels in the developing stalk, and the effects of auxin transport inhibition imposed after bloom. For this, N-( 1 -Naphthyl)phthalamic acid (NP A) was used, based on its specific ability to inhibit cellular carrier-mediated auxin transfer (Lomax et al., 1 995), without itself being transported in a polar manner (Thomson et al., 1 973). The hypothesis Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 45 being tested was that i f auxin export from the fruit was the detennining factor in xylogenesis, then NP A applied to the fruit stalk just after bloom would suppress proximal post-bloom vessel differentiation in the stalk. Fonnations of primary xylem laid down during the initial growth of the apple flower stalk were not expected to be affected, but the secondary differentiation that occurs largely after anthesis (Barden and Thompson, 1 963) was expected to be affected. 3.2. MATERIALS AND METHODS 3.2.1 . Plant material Measurements to assess stalk development were carried out by Dr Alexander Lang on mature 'Royal Gala' apple trees grown at the Massey University Fruit Crops Unit, Palmerston North, New Zealand. The trees were planted on MM 1 06 rootstock and subjected to standard commercial management. The NPA experiments were carried out on ' Granny Smith ' apple trees. Six mature trees planted on MM 1 06 rootstock (three trees for each of the two NP A experiments) were chosen from a private organic orchard on the outskirts o f Palmerston North, New Zealand. Clusters were randomly selected around the canopy at 1 -2 m height, mainly on 2-year old spurs. To minimise possible effects of uneven seed set on vessel differentiation in the stalk, clusters were hand-thinned to one lateral flower and hand­ pollinated with 'Red Delicious' pollen collected the same season. 3.2.2. Preparation and application of auxin transport inhibitor (NP A) N-( l -Naphthyl)phthalamic acid (NPA) was purchased from Tokyo Kasei Kogyo Co Ltd, Tokyo, Japan. In both experiments anhydrous lanolin was used as a carrier for exogenous application. Since NP A powder may not dissolve completely in lanolin resulting in uneven application (Haga and lino, 1 998), NP A was dissolved in a mixture of ethanol (99%) and 1 M NaOH before being added to the lanolin. The NP Nlanolin pastes were prepared in four logarithmically scaled concentrations: 0.01 mg g-I , 0. 1 mg g-I , 1 mg g-l and 1 0 mg g-l . To obtain 1 0 mg NPA g-l lanolin, 200 1l1 1 M NaOH and 400 III ethanol were mixed with 50 mg NP A. Solution was slowly added to 4950 mg of melted lanolin Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 46 resting on a hot plate to avoid an explosive reaction of ethanol and stirred with a glass rod until homogenised. Similar amounts of solvent added to lanolin served as the control. Prepared pastes were kept in a refrigerator for a short period (several days) until their application in the orchard. Two different applications ofNP A were carried out in the earliest stage of fruit development, 2 - 4 d after petal fall. In the first experiment, 1 0 mg NP A g-l lanolin was applied in a horizontal ring (about 2 mm wide) to the surface of the stalk using the tip of a wooden toothpick. Two distinct positions along the stalk were chosen - proximal, at the base of the stalk adj acent to the spur and distal, at the base of the fruit j unction with the stalk. This created four different treatments: (a) NPAllanolin paste to the proximal end, (b) NP Allanolin paste to the distal end, (c) NP Allanolin paste to both proximal and distal ends, and (d) lanolin paste (with solvent only) to both proximal and distal ends. There were 60 samples for each treatment, giving a total of 240 samples. In the second experiment, four different NPAllanolin pastes (0. 0 1 mg g-l to 1 0 mg g- I) and control (with solvent only) were placed with a wooden toothpick as a horizontal half-ring (again some 2 mm wide) at the midpoint of the stalk. For each treatment there were 40 samples, giving a total of 200 samples. 3.2.3. Vessel measurements To monitor extension growth and vessel differentiation in ' Royal Gala' stalks, 22 periodic measurements were taken between bud break and maturity. At each sampling time, five average sized fruit were randomly selected from the canopy, which provided a total of 1 1 0 fruit to be analysed. To assess the effects ofNP A on vessel differentiation, all treated ' Granny Smith' fruit were collected 26 and 30 d after petal fall . Fruit were picked from the orchard in the morning and held in promptly sealed polyethylene bags pending measurements on the same or following day. In the laboratory, fruit and seeds were weighed. Seeds were classified into two broad categories: ' viable' and ' aberrant' and their number recorded. The parameter for separating seeds was the overall appearance, with the 'viable' seeds being fully formed and fleshy while the ' aberrant ' seeds were considerably smaller and often flattened with empty interiors. - --------- Chapter 3. Xylem development in the stalk of apple and the ef ects of auxin transport inhibition 47 Stalks were removed with a small amount of fruit flesh still attached at the distal end and immediately placed in Formol-Acetic-Alcohol (FAA) fixative (5mI 3 8 % w/w aqueous formaldehyde and 5 ml glacial acetic acid in 90 ml 70% v/v aqueous ethyl alcohol) pending microscopic examination. PROC GLM (SAS, 2000) was carried out to evaluate the effects of NPA treatments from both experiments through a) fruit abscission and b) seed set (number and weight) and fruit weight. Tree and branch effects were tested for significance (as class variables). In both studies, sectioning was done by hand using a sharp single-edged blade (Leica Model 8 1 8) to produce transverse sections suitable for examination after staining. For microscopy, excised transverse sections were soaked on a glass slide with phloroglucinol solution (2. 5 g in 50 ml of 70% v/v aqueous ethyl alcohol) for about 60 s and then flooded with concentrated HCl. Once the lignified walls had stained red the sections were mounted in 80% (w/w) glycerol. The size and frequency of vessels within any size class were assessed immediately after slide preparation because the stain used to highlight the lignified cell walls fades after a few hours. The image from a compound light microscope (Kyowa Microlux- l 1 coupled to a Panasonic GP-KR222 video camera) was presented on a 1 7" flat VDU screen (Sony Trinitron Multiscan G400) at x475 magnification using an image-grabber card and Video Master™ software (Fig. 3 . 1 ). Based on the method described by Lang and Ryan ( 1 994), the internal diameters of vessels were evaluated by eye using custom mouse-controlled software that presented a moveable bitmap image on the screen. This image comprised a set of five graded rings (sizes 1 , 3 , 5 , 7 and 9) allowing estimates to be made of four intermediate sizes by interpolation (sizes 2, 4, 6 and 8). That arrangement allowed nine vessel-size classes to be recorded using custom software that converted the PC's numeric keypad into a nine-channel tally counter. Each ring size was calibrated to a known diameter on the slide (3, 5 , 7, 9, 1 2, 1 4, 1 6, 1 8 and 2 1 �m). Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 48 Fig. 3 .1 . Diagram showing the set-up for quantitative video-microscopic analysis. Sizing was by comparison with the set of five mouse-controlled reference images (visible at bottom right of the screen image) and their four (estimated) half sizes. The data were directly stored through the proprietary software to disc. The hydraulic conductance (C) was estimated using the Hagen-Poiseuille law (Nobel, 1999) as the product of the sum of vessels (n) and the fourth power of radius (r) for a particular vessel size class (C ex:; n r4). Because of the nature of the NP A experiments, it was essential to determine the precise positions from which transverse sections were to be taken. Three relatively long stalks (� 30 mm in length) of control fruit taken from the first NPA experiment were hand­ sectioned every 2 mm along the stalk, starting from the spur end. This provided 14 positional cuts per stalk. Immediately after the completion of serial vessel area ----------- - - - - Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 49 measurements, the data were analysed for the spatial distribution of vessels. The result determined two regions with similar hydraulic conductance (c. 20% o f the stalk length in from either end) from which proximal and distal sections were obtained in the second NP A experiment. The analysis showed that stalks treated with 1 mg NP A g-l lanolin revealed no significant effect on vessel differentiation. Therefore, sectioning of the 0.0 1 and of 0. 1 NPA g-l lanolin samples was not done and that o f the 1 mg NPA g-I lanolin samples was discontinued after just five fruit stalks had been assessed. All of the thirteen available samples treated with 1 0 mg NP A g-l lanolin were analysed whilst for the control the analysis o f ten samples (including three samples used for serial sectioning) was considered sufficient. PROC GLM (SAS, 2000) was carried out to evaluate the significance of differences between the two positions on the stalk. The data were normalised by taking the distal:proximal ratio to remove fruit-to-fruit variability and tested to determine if the ratio differed: • between NP A-treated fruit and control, and • from 1 for each treatment. Tree and branch effects were tested for significance (as class variables). 3.3 RESULTS 3.3.1. Spatial distribution of the vessels Serial sectioning indicated an even pattern of vessel frequency along the fruit stalk (Fig. 3 .2A). However, estimated hydraulic conductance exhibited a conspicuous bell-shaped distribution (Fig. 3 .2B), producing regions with higher hydraulic resistance closer to the two ends of the stalk. The extent of xylem development at the distal end was reduced compared to that in the middle of the stalk, but there was a more significant reduction in the conductance at the proximal (spur) end (c. 3-fold decrease). Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 50 200 I- en '0 en � 1 50 :> ....... o I-< 1 00 Zi 50 A . 2 4 6 8 1 0 1 2 1 4 1 6 1 8 20 22 24 2 6 28 I I I -- 200 I- B . . � § • • • • I I - .f 1 50 I- • .... O-�-� - g 1 00 �. /- • • • • ��_; � . . Q 50 "8 8 • • 2 4 6 8 1 0 1 2 14 1 6 1 8 20 22 24 26 28 Distance from spur end (mm) Fig. 3 .2 . Spatial distribution ofthe number of vessels (A) and estimated conductance (B) in the stalk of ' Granny Smith' . For each position along the stalk, the mean value (0) was derived from three measurements (.). 3.3.2. Frequency:size distribution of the vessels The frequency of a particular size class of vessel showed a distinct pattern in the stalk of ' Granny Smith' (Fig. 3 . 3A). The 'narrow' vessels (:S; 9 !-lm in diameter) dominated in the cross-sectional area o f the xylem tissue, comprising over two-thirds o f the existing conduits with the most frequent diameters being 5 and 7 !-lm. The drop in frequency o f the 9 !-lm size class, relative t o the 7 and 1 2 !-lm classes, was taken arbitrarily a s the boundary between vessels of different derivation. The 'wide' vessels (� 9 !-lm in diameter) contributed relatively little to the overall vessel frequency. The majority of -� - ----------------- Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 5 1 these vessels fell into the 1 2 /lm and 1 4 /lm classes with a progressive decrease in the frequency of vessels with larger diameters. The mean number of vessels contributed by each size class in the stalk was used to calculate their fractional mean conductance (Fig. 3 . 3B). Contrary to the previous trend, most of the conductance in the stalk was attributed to a relatively small number of 'wide' vessels. The largest proportion of the conductance (37%) occurred through the vessels in the 1 4 /lm size class. However, the larger vessels ( 1 6, 1 8 and 2 1 /lm diameter) were relatively few, and together contributed less (33%) than the single 1 4 /lm size class, while 'narrow' vessels were too limited in diameter to contribute significantly. 40 A . ,-.., � '-' rJl 30 -Q) rJl rJl Q) :> 20 <..,.. 0 I-< � 1 0 :;:::I Z 0 3 5 7 9 1 2 1 4 1 6 1 8 21 40 B. ,..-... � 30 '-' Q) u Q .a 20 u :;:::I '"0 § 1 0 U 0 3 5 7 9 1 2 14 16 1 8 21 Diameter of vessels (J-lm) Fig. 3 . 3 . Fractional contribution by vessel size classes averaged from cross-sections of three excised stalks of ' Granny Smith' : (A) for the mean number of vessels; (B) transformed to reflect the mean conductance. Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 52 3.3.3. Development of the stalk The apple stalk extended rapidly over a brief period at the beginning of the season. Extension started pre-bloom and had largely ceased by 1 2 days after full bloom (DAFB), (Fig. 3 .4). Consequently, extension growth of the stalk of ' Royal Gala' was completed within 4-5 weeks after extension growth started at bud break. 40 r- ,..-, § 30 '-" t 5 20 - � S if) 1 0 -20 o 20 40 60 80 Time (DAFB) 1 00 1 20 1 40 Fig. 3 .4. Stalk extension growth in ' Royal Gala' . For each time period, the point (e) represents the mean of five individual measurements (unpublished data, A. Lang). Vessels differentiated over a relatively short period of time during the extension of the stalk (Fig. 3 .5A). The differentiation of 'narrow' vessels (� 1 0 /-Lm in diameter) ceased at -8 DAFB by which time ' wide' vessels had emerged in the xylem. During the period of bloom, there were no new ' wide' vessels (a lag phase) but soon after vessel differentiation proceeded in an exponential fashion. Their differentiation started from 4 DAFB and continued rapidly for several weeks during which time ' wide' vessels fully differentiated in size. By 28 DAFB vessel differentiation in the stalk was effectively completed. Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 5 3 When assessed for conductance (Fig. 3 . 5B), 'narrow ' vessels exhibited an initial increase but contributed very little to the flow due to their limited cross-sectional area. They reached a peak contribution around -4 DAFB. At that time, a small number of larger vessels had differentiated but no further differentiation occurred during bloom. From 4 DAFB vessel differentiation increased rapidly and by 2 8 DAFB the flow capacity in the stalk was essentially established. Conductance continued to increase after that but at a much-reduced rate due to the very limited additional vessel differentiation. 250 A . 200 Vl -0 Vl Vl 1 50 0 :> 4-< 0 I-< 1 00 0 � Z 50 0 -20 0 75 ,.-.., B. .l!l . .... 60 § C � 45 . .... 12 ('(j ';" 30 () � () 1 5 .g § U 0 -20 0 Y'" ---" ... -AA-... __ ._ ..... ... ... . ..... _.-.-...... --......... -. 20 -0-00-0-0 __ 0-0 -O-o-cro-Cl -0-00-0..0-0-0-0-0-0.0-0 40 60 80 1 00 1 20 �-.-.'II-. ��':Y-�"-T -•• -... ..--y �-� :'If'r ..... . . . .�...........--.� / ... -&&-.-.--.... -.-..... . -. � _.-.-..... --.... ... ... o-o-CJo-o-o--o -0 20 40 60 80 1 00 1 20 Time (DAFB) 1 40 1 40 Fig. 3 . 5 . Vessel differentiation in the stalk of ' Royal Gala' . Cumulative distributions of vessel frequency (A), and estimated conductance (B), contributed by different size classes: S 7 . 5 /lm (0), S 1 0 /lm (0), S 1 2 /lm (e), S 1 4 /lm (.�), S 1 6 /lm (+), S 1 8. 5 /lm (T) and S 2 1 /lm (.). For each time period the points represent the means of five individual measurements (unpublished data, A. Lang). � - --�����-�---------- Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 54 3.3.4. Effect of NP A on fruit abscission Post-bloom observations in the orchard revealed a severe drop o f NP A-treated fruitlets from the first experiment only two weeks after application of NP A. The remaining fruitlets were harvested after the estimated completion of vessel formations in the stalk, between 26 and 3 0 days after petal fall. A small number of fruitlets from both experiments whose size was markedly reduced were held weakly in the canopy with obvious signs of partial disconnection at the abscission zone. Byers et al. ( 1 99 1 ) and Ward and Marini ( 1 999) observed cessation of fruit growth prior to abscission, while Avanzi et al. ( 1 988) noted that the ovule structures of abscised and abscising apples showed no histological differences. On this basis, abscising fruit were excluded from further analyses. The number of abscised fruitlets from the control treatment was very low and differed significantly from all NPA treatments (p � 0.00 1), (Table 3 . 1 ). However, the comparison between NP A treatments showed that distal applications caused significantly less abscission (p � 0.0 1 ), while the proximal application bore no significant difference from 1 0 mg NPA g-l lanolin applied to both distal and proximal ends of the stalk. Table 3 . 1 . The effect of 1 0 mg NPA il lanolin on fruit abscission when applied to the distal (D), proximal (P), and distal and proximal ends of the stalk (D + P). There were 60 fruit per treatment. Numbers followed by the same letter are not significantly different from one another. Position D P D + P Control Fraction abscised (%) 83.4 b 9 1 .6 c 1 00.0 c 1 8.4 a Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 5 5 Drop was not as severe i n the second experiment, but again the highest concentration of NP A resulted in a significant increase in fruitlet abscission (p s 0.01 ), while the other treatments with lower NP A concentrations were no different from the control (Table 3 .2). Table 3 .2 . The effect of four NP A concentrations on fruit abscission when applied to the mid-point of the stalk. Numbers followed by the same letter are not significantly different from the control. NPA (mg g-l lanolin) 0.01 0. 1 1 1 0 Control Fraction abscised (%) 32.5 a 3 5 .0 a 2 5 . 0 a 67.5 b 3 5 .0 a 3.3.5. Effect of NP A on seed and fruit development Data showed no difference in either 'viable' seed number or total seed number between the NP A-treatments and controls from both experiments. However, the proximal application of 1 0 mg NPA i' lanolin in the first NPA-experiment caused a significant reduction in seed weight (p s 0.01 ) and fruit weight (p s 0.001 ), (Table 3.3). Table 3.3. The effect of 10 mg NPA g- l lanolin on fruit and seed development when applied to the distal (D), and proximal (P) ends of the stalk. Numbers o f fruit per treatment are in brackets and are those from the first experiment shown in Table 3 . 1 . Numbers followed by the same letter are not significantly different from the control. Position Fruit weight Seed weight Seed number Seed number (g) (mg) ( , viable') (total) D ( 1 0) 3 . 1 a 8 1 a 8 . 6 a 1 0.2 a P (5) 1 . 8 b 5 9 b 9.2 a 1 0.2 a Control (49) 3 . 8 a 97 a 8.7 a 1 0. 1 a Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 56 The analysis of data from the second experiment showed that 10 mg NP A g-I lanolin effectively suppressed seed weight and fruit weight (p � 0.00 1 ), (Table 3 .4). It should be noted, however, that 0.0 1 and I mg NPA g- l lanolin treatments stimulated fruit growth (p � 0.05) but had no effect on seed growth (except for the slightly higher total seed number from 1 mg NP A g- I lanolin treatment, p :::; 0.05). The overall picture was essentially the same when ready-to-abscise fruit were added to the data, only the increased fruit-to-fruit variability amongst the treatments changed the p-values (data not shown). In addition, comparison of controls from the two NP A­ experiments showed uniform seed number but a highly significant reduction in the weight of seed and fruit picked only 4 days apart (p :::; 0.00 1 ) . Whenever fruit or seed weight effects were significant, so too were tree or branch effects (p � 0.05). Hence, these class variables were retained in all analyses. Table 3 .4. The effect of four NPA concentrations on fruit and seed development when applied to the mid-point of the stalk. Numbers of fruit per treatment are in brackets and are those from the second experiment shown in Table 3 . 2 . Numbers followed by the same letter are not significantly different from the control. NPA Fruit Seed Seed number Seed number (mg g-l lanolin) weight (g) weight (mg) ('viable') (total) 0.01 (27) 6.6 b 1 45 a 8.2 a 1 0 . 1 a 0. 1 (26) 5 . 8 a 1 27 a 7.9 a 1 0 . 1 a 1 (30) 6.3 b 1 43 a 8.4 a 1 0.3 b 1 0 ( 1 3) 3 .4 b 1 00 b 8 . 5 a 1 0.0 a Control (26) 5 . 5 a 1 32 a 8.7 a 1 0.0 a 3.3.6. Effect of NP A on vessel differentiation When the frequency of different vessel sizes was computed proximally and distally to the point ofNP A application in the second experiment, two peaks were seen, representing populations of 'narrow' and 'wide' vessels in the stalk. This is apparent both for the number of vessels (Fig. 3 .6A) and for the conductance (Fig. 3 .6B) contributed by each Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 5 7 size class o f vessel. The effect o f auxin transport inhibition was statistically assessed in 'wide' vessels only, since it was assumed that the ' narrow' vessels had differentiated pre­ bloom, as in the stalk of 'Royal Gala' (see Fig. 3 . 5A). In addition, branch and tree effects were not significant and so were omitted from the analyses. Assessment of the data showed that 1 mg NPA i1 lanolin caused a slight increase in the number of vessels and in conductance (Fig. 3 .6) with changed distai:proximal (dip) ratio (0.84 and 0.85 respectively) indicating promotion of growth in the proximal part of the stalk. While analysis showed these trends to be statistically non-significant, there was an effect on subsequent seed and fruit growth (Table 3 .4). 1 00 A . 00 80 -0) 00 00 0) 60 ;;. ...... 0 � 0) 40 � Z 20 0 3 5 7 9 1 2 1 4 1 6 1 8 2 1 -. 1 00 B. 00 .... ...... $:I :::I 80 � 60 ...... -e e<:S '-' 0) 40 (.) B (.) 20 :::I '"0 $:I 0 u 0 3 5 7 9 1 2 1 4 1 6 1 8 2 1 Diameter o f vessels (J..lm) Fig. 3 . 6. The effect of 1 mg NP A g-l lanolin (.A.), 1 0 mg NP A g-l lanolin (_) and control (e) on the number of vessels (A), and estimated conductance (B) in the stalks of ' Granny Smith ' . Each point represents the average of proximal and distal sections. Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 5 8 The application o f 1 0 m g NP A g- l lanolin showed a significant increase i n dip ratio for vessel number relative to the control (p � 0.0 1 ), (Figs. 3 . 7A and 3 . 7B). The dip ratio in NPA-treated stalks (2.3 1 ) differed from 1 (p � 0.00 1 ) suggesting greater suppression of vessel differentiation proximal to the point ofNP A application. In comparison, the dip ratio in the control stalks ( 1 .05) was not significantly different from 1 . When expressed as conductance, 1 0 mg NPA g-I lanolin also showed a significant increase in dip ratio relative to the control (p � 0.05 ), (Figs. 3 . 7 C and 3 . 7D). The dip ratio in NPA-treated stalks (2.83) differed from 1 .0 (p � 0.00 1 ) but not in control stalks ( 1 .25). The most pronounced differences occurred in 1 2, 14 and 1 6 /-lm vessel size classes that contribute over two thirds of the flow capacity in the stalk (see Fig. 3 .3B). 60 50 '" '& 40 [G .;: 30 0 � 20 :::3 Z 1 0 ,.-.. . .;J 50 § C 40 g :e 30 (IS '-' 11) � 20 1 10 A . B . /;'" , , cS 0' '0, c. D . U o���������������������������� 3 5 7 9 1 2 14 16 18 21 3 5 7 9 1 2 14 16 18 Diameter o f vessels (Ilm) Diameter of vessels (Ilm) Fig. 3 . 7 . Number of vessels and estimated conductance contributed by different size classes in the stalks of ' Granny Smith' treated with 1 0 mg NP A g-J lanolin (A, C) and control (B, D). The points represent the mean values from distal (e) and proximal (0) sections. --------------- --- -- - - Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 59 3.4. DISCUSSION 3.4.1 . Vessel measurements and estimated conductance The method used to measure the numbers and internal diameters of vessels proved simple in concept and accurate. The cut vessels were clearly seen on the PC screen, as bright perforations with distinguishably stained walls. Oblique or over-thick sections were easily rej ected because light only illuminated part of the vessel areas, and vessel margins were indistinct. Measurements were quick and straightforward using a mouse­ controlled selective template with direct data acquisition to disc. Hence, every vessel in transverse section was processed only once without risk of repeating the same measurement and with minimal operator error because the eye quickly adopts an accurate and consistent pattern during evaluation. Vessels were rarely circular in their cross-section, and the inner and the outer limits were matched by superimposed rings making allowance for irregularities in a perfect circle. This non-ideal geometry of conduit lumen may contribute to the differences between theoretical (estimated) and actual conductivity as noted by some authors (Tyree and Zimmermann, 1 97 1 ; Calkin et al., 1 986; Tyree and Ewers, 1 99 1 ). The agreement between the theoretical and actual conductance could be improved by using a correction factor for reduced conductance through non-cylindrical cross-sections of the xylem conduits (Calkin et al., 1 986). However, when such a correction to the Hagen-Poiseuille equation for laminar flow through ideal (perfectly cylindrical) capillaries was applied to the elliptical shape of the xylem conduits in the internodes and petioles of the grape it resulted in a very small reduction of the predicted conductance ranging from only 1 .5 to 5 .4% (Schultz and Matthews, 1 993). Since the difference appears small, the conductance in the stalk of the apples used in this study was estimated by way of the original Hagen­ Poiseuille equation (Nobel, 1 999). 3.4.2. Spatial distribution of the vessels The stalk seems to be an obvious site of increased resistance to water flow due to the limited cross-sectional area of conduits branching from the more developed fruit-bearing spur beneath and transporting sap into the much larger fruit above. While the number of vessels along the length of the stalk was consistent (Fig. 3 . 2A) their cross-sectional areas -- --�-----------� Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 60 were less developed towards the ends producing structural 'bottleneck' regions with markedly reduced conductance. Moreover, vessel development was more arrested at the proximal end of the stalk adjacent to the abscission zone (Fig. 3 .2B). This is in agreement with the existence of distinct xylem constrictions at or near the abscission zone of the fruit (MacDaniels, 1 937; Habdas et al., 1 982; MacKenzie, 1 988; Lee, 1 989) with increased hydraulic resistance of the bottleneck regions just distal to the branch points (During and Lang, 1 993). Whether vessel formations terminate at the junction and develop as adj acent vessels from the other side (arranged end-to-end) or simply reduce in frequency across the zone, it certainly indicates a natural cleavage site for organ separation. Furthermore, restrictions of xylem water flow could play an important role in the hydraulic architecture of the plant. It has been argued that hydraulic bottlenecks at branch points may provide a safety feature against xylem cavitation of the parent body after the abscission of terminal organs (Zimmermann and Milburn, 1982); or thus may lead to the development of a water potential gradient between the stem and the fruit to maintain the ascent of sap (Lee, 1 989); or thus may cause the diversion of xylem water flow into the phloem (Lee, 1 989; During and Lang, 1 993). The change of water flow carrier across the bottleneck regions could be of particular importance for fruit. Apple fruit growth is mainly phloem dominated (Lang, 1990) and this pathway is known to be calcium immobile (Marschner, 1 986). Assuming that hydraulic constrictions along the xylem strand do promote a shift in the balance in favour of the phloem-borne transport, this may contribute to the relatively limited capability of plants to regulate calcium uptake by developing fruit. However, the convoluted pattern of xylem and phloem flow in regard to fruit development requires further research to confirm this assumption. 3.4.3. Frequency:size distribution of the vessels The distribution of vessel size classes in the stalk of ' Granny Smith' (Fig. 3.3A) suggests two stages in vessel differentiation. The frequency o f size classes correlated with the arbitrary division between 'narrow ' and ' wide' vessels, the former representing the greater number of vessels in the stalk. In contrast, Lang and Ryan ( 1 994) reported a bell- Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 6 1 shaped vessel distribution in two different apple cuItivars but their distinctive profiles indicate that the course of vessel differentiation is a particular feature of a given cultivar. When the frequency data of Fig. 3 .3A were transformed to represent the conductance contributed by each vessel size class, it could be seen that the large number of 'narrow ' vessels could carry only a small proportion of the sap (c. 1 0%); most of the conductance was contributed by the relatively small number of 'wide' vessels (Fig. 3 .3B). However, the calculations assume that all the vessels in the stalk are functional. This is probably not the case. Actual flow is certainly smaller than estimated conductance because the earliest xylem formed is stretched and broken during the rapid early growth (Barden and Thompson, 1 963; Esau, 1 977; Mapfumo et al., 1 993), and because some large vessels could retain degenerated protoplasm which would restrict the effective vessel diameter (Mapfumo et al., 1 993). Xylem sap flow through the stalk will thus depend upon size and frequency of large and functional vessels, as suggested by other authors (Zimmermann, 1 98 3 ; Tyree and Ewers, 1 99 1 ; Lang and Ryan, 1 994). 3.4.4. Development of the stalk Stalk extension in 'Royal Gala' is effectively linear (Fig. 3 .4) and matches that observed in other apple cultivars (Prive et al., 1 988). Final stalk length is apparently a cultivar characteristic and although it can be manipulated by the application of exogenous hormones (paclobutrazol and gibberellins), the dynamics of growth do not change (Prive et al. , 1 989). Vessel formation, and thus the hydraulic conductance of the stalk, start before flowering (Fig. 3 . 5A) as also shown in previous studies of apple (Barden and Thompson, 1 963), pear (Nii, 1 980) and grape (Theiler and Coombe, 1 985). The physiologically significant increase in the cross-sectional area o f the vessels was due to a relatively short post-bloom development of 'wide' vessels, which determines the hydraulic capacity o fthe stalk by around four weeks after bloom (Fig. 3 . 5B). It should be noted that the rapid post-bloom vessel differentiation starts to plateau when the stalk has achieved final length ( 1 2 DAFB) suggesting a correlative link between extension and structural development of the stalk. Furthermore, the timing and dynamics of vessel differentiation is in line with the rate of auxin transport from developing apple fruit (Gruber and Bangerth, 1 990). Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 62 The growth dynamics of the vessels and corresponding conductance in the stalk indicate distinct stages in vessel differentiation. Vessels laid down after the initiation of flower development (� 1 0 Jlm in diameter) clearly mature before bloom (Fig. 3 . 5 A) and could develop only as early formations of the primary xylem (protoxylem). The remaining vessels that start differentiating before bloom but largely developed after bloom, therefore, represent later stages of the primary xylem (metaxylem) as well as the secondary xylem formations in the stalk (Barden and Thompson, 1 963). This bears close resemblance with the frequency distribution of vessels size classes in ' Granny Smith' apple (Fig. 3 .3A). 3.4.5. Effect of NP A on fruit abscission Young developing apple fruit are a source of diffusible auxin (Grub er and Bangerth, 1 990), such that seed number greatly influences polar auxin efflux (Bangerth et al., 1 989; Callej as and Bangerth, 1 997; Bangerth, 2000). The synthetic phytotropin NPA is thought to inhibit transport of auxin by blocking the sites of auxin efflux at the basal end of the cells (Lomax et al., 1 995) or even disrupting the influx carriers at higher concentrations (Delbarre et al., 1 996). Hence, sufficient NP A concentration that inhibits carrier­ mediated auxin transport at the site of application will deplete the level of auxin at the abscission zone of the stalk. Indeed, the highest NPA concentration ( 1 0 mg g- l ) induced significant abscission, which was more pronounced when a greater amount ofNP A-paste was applied to the stalk as a complete ring (Tables 3 . 1 and 3 .2). Although the abscission was complete when NPA was applied to both ends of the stalk (Table 3 . 1 ) all 1 0 mg NPA g-l lanolin treatments caused significantly greater drop than the controls, suggesting that disruption of auxin transport distal to the abscission zone promoted shedding of the young fruit. The results support the view that changes in the normal auxin status of a tissue initiate fruit abscission (Sexton and Roberts, 1 982; Brown, 1 997; Bangerth, 2000) at a pre­ determined position (Osbome, 1 989). Auxin has been shown to antagonise abscission (del Campillo and Bennett, 1 996) by inhibiting the accumulation o f the hydrolytic enzyme cellulase (Tucker et al., 1 988) in the abscission zone of the apple fruit stalk (Pandita and Jindal, 1 99 1 ; Ward et al., 1 999; Ward and Marini, 1 999). Hence, the lower - �. --------------- Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 63 NPA concentrations (0.0 1 , 0. 1 and 1 mg g.J ) were presumably insufficient to counter the endogenous auxin transport at the site of efflux (Table 3 .2). Seeds appear critical in detennining fruit retention on the tree. Abscised fruit may show degenerated ovule structures (Forino et. al. , 1 987; Avanzi et al., 1 988), reduced seed set (Heinicke, 1 9 1 7) or marked inhibition of seed growth (Wertheim, 1 97 1 ; Abruzzese et al., 1 995; Knoche, et al., 2000). Furthennore, damage to ovules leads to severe fruit drop in apple (Abbott, 1 959; Marini and Byers, 1 988; Ward et al., 1 999) that increases if injury is imposed earlier (Proctor and Schechter, 1 992). This emphasises the significance o f dynamics o f 'honnone' production b y developing seeds that coincides with the rate o f fruit abscission (Luckwill, 1 948 ; 1 953). 3.4.6. Effect of NP A on vessel differentiation Post-bloom application of 1 0 mg NP A g'! lanolin induced significant changes in the xylem but only the ongoing differentiation of 'wide ' vessels was affected (Fig. 3 . 6). Because ' wide' vessels principally govern the conductance in the stalk (Fig. 3 . 3 B), any change in their frequency will amplify the difference between the treatments (Fig. 3 .7). With respect to the point ofNP A application on the stalk, the xylem at the proximal position developed fewer vessels indicating that the critical signal was travelling from the developing fruit along the stalk and not coming from the other direction. This result points to the auxin-like signal, which can regulate the induction and fonnation of the vascular tissues (Aloni and Zimmennann, 1 983). A good body of evidence has shown the close relationship between growth dynamics of the fruit and vascular development in the stalk (Nii, 1 980; Habdas et al., 1 982; Bustan et a l. ; 1 995). Hence, the amount of vascular fonnation in the stalk relies on the ability of the fruit to supply a growth-promoting stimulus. The pattern of vessel differentiation signifies the mechanism underlying vascular changes in NP A-affected stalks. Because auxin movement is preferentially basipetal from the apple fruit (Bangerth, 2000), NP A was expected to alter vessel fonnations proximal to the point of the blockage. However, post-bloom vessel differentiation appears affected (to a different degree) along the entire stalk when compared to the control (Fig. 3 . 7). In - - �------------ Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 64 other words, 1 0 mg NP A g- l lanolin suppressed developing formations of 'wide ' vessels at both proximal and distal ends ofthe stalk although the nature of NP A was likely to inhibit polar auxin movement only at the site of action (Thomson et al., 1 973). This could arise because the application ofNP A was exogenous, in the orchard, and the lanolin paste may have migrated along the exterior surface of the stalk (both lanolin and the stalk surface are lipophilic). However, a more likely scenario involves inhibition of auxin biosynthesis due to severe disruption of auxin efflux at a point along the stalk. Since vessels resume rapid differentiation in the short space of time after anthesis (Fig. 3 .5A), NP A principally affected ongoing development of ' wide' vessels at the proximal end of the stalk deprived of the continuous auxin signal (Fig. 3 . 7). This, in turn, could have stimulated auxin accumulation upstream (c! membrane vesicles isolated from Cucurbita, Michalke et al., 1 992). Once the blockage of polar auxin transport was completed and/or auxin levels in the tissue exceeded an upper concentration threshold it seems likely that auxin biosynthesis in the fruit would have ceased, which then affected vessel formations at the distal end of the stalk. The time lag between NPA-induced auxin inhibition at the mid-point of the stalk and negative feedback of auxin biosynthesis could explain positional differences in vessel differentiation on the same stalk. In addition, the stimulating effect of the lower NP A concentration on vessel differentiation (Fig. 3 . 6) and fruit growth (Table 3 .4) conform with the idea that subtle rebalancing o f hormones determines the nature of the physiological response. 3.4.7. Effect of NP A on seed and fruit development Hand-pollination caused a high seed complement in all fruit (Tables 3 .3 and 3 .4). Analysis of variance showed fairly uniform seed number between NP A-treatments and controls suggesting that changes in the dynamics of fruit abscission and post-bloom vessel differentiation in the stalk were not influenced by inadequate or uneven seed formation. This is in line with evidence that pollination, as a precursor of seed formation in the fruit, controls the extent of the ' first drop' shortly after flowering (Wertheim, 1 9 7 1 ) The most significant effect o n seed and fruit growth occurred when 1 0 m g NP A g- I lanolin was applied either at the proximal end of the stalk (Table 3 .3 ) or mid-stalk Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 65 (Table 3 .4). Basipetal auxin export is obviously an initial step in the physiological mechanism that controls growth. This suggests that a continuous growth-promoting signal from the actively growing fruit structures is critical not only for vessel differentiation but also for the normal development of the fruit itself. It has been shown that an inhibition of auxin transport from fruit caused substantial reduction in seed weight and partially seed number (Hamamoto et al., 1 998). However, the assessment of 1 0 mg NPA g-l lanolin treatments between two experiments was not possible because of the substantial differences in seed and fruit weight of the control treatments, which had slightly different durations of growth. These observations point to the interaction between hormones and assimilates in developing fruit. During initial growth of apple, the seed undergoes a prominent increase in size (Harley, 1 942; Luckwill, 1 95 3 ; Sato and Kanbe, 1 984), but changes in this period are not restricted to seed set only. Arrested seed development leads to inadequate hormonal production (Bangerth et al., 1 989; Callej as and B angerth, 1 997; Bangerth, 2000) and diminished sink capacity (Weinbaum et al., 200 1 ), which could affect fruit growth. The NP A-induced reduction in the size of seed and fruit suggests that movement of nutrients into the developing fruit is under hormonal control . The physiological significance of these effects is probably to promote the redistribution of nutrients into the dominant fruit. However, the exact mode of action or possible nature o fthe self­ inhibiting signal remains uncertain. Whether early seasonal vessel differentiation is simply a result of a continuous hormonal stream that supports fruit growth by expanding the conductive structures or as an auxiliary outcome of actively developing fruit structures is not yet clear. On the other hand, an increase in fruit size requires a suitable vascular capacity to meet the increasing nutritional demands of the sink. Limited development of conducting tissues due to reduced supplies of growth promoting substances could in return limit the rate of ascent of sap into the growing fruit and alter its sink capacity. As a consequence, weaker fruit may lose the ability to compete for nutrients supplied from the tree, which might lead to shedding. It seems obvious that further studies are needed to elucidate the mechanisms of hormonal action in the control of the physiological response of the vascular tissues and its possible influence on the nutritional status of the fruit. - - _._--------------- Chapter 3. Xylem development in the stalk of apple and the effects of auxin transport inhibition 66 3.S. CONCLUSIONS Growth dynamics of vessel fonnations were detennined periodically throughout the season in the stalks of 'Royal Gala' . Quantitative microscopic analysis showed that a maj ority of vessels are differentiated pre-bloom with the commencement o f stalk extension but that the full conducting capacity of the xylem is attained shortly after bloom. These results suggest two distinct stages in vessel differentiation in the stalk. The influence of auxin transport inhibition in post-bloom vessel differentiation of ' Granny Smith' apple stalk was investigated. Vessel-size frequency (that yields a potential conductance) was assessed using a quantitative microscopic analysis o f the stalk of young fruit harvested around four weeks after petal fall. The evidence is presented that an auxin-like signal emanating from the young fruit not only stimulates vessel differentiation in the stalk but also controls the abscission and development of seed and fruit. The effects were discussed through the honnonal and nutritional aspects of fruit development and a proposed physiological link was related to the dynamics of vessel differentiation by auxin. In addition, vessel constrictions were identified near the j unctions of the stalk suggesting structural modifications of the xylem strand in the plant. Physiological consequences were discussed based on reports from the literature. Although the physiological mechanism for the regulation of long-distance transport is more complex, hydraulic adaptations along the xylem strand will contribute to the efficiency of the mineral uptake of the fruit. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 4.1 . INTRODUCTION A number of fruit species undergoes a progressive xylem dysfunction (Findlay et al., 1987; During et al., 1 987; Lang, 1 990; Creasy et al., 1 993; Lang and Ryan, 1994; Wolswinkel et al., 1 999; Drazeta et al. , 200 1 ). For example, in wine grapes, vascular dysfunction coincided with the sudden resumption of berry growth after the onset of ripening (Findlay et al., 1987; DUring et al., 1987). Viewed in these terms, the progress of xylem dysfunction is determined by a growth-induced damage of the bundles as the fruit expands. 67 Progressive breakdown of conducting tissue in apple has been attributed to the failure of a number of vessels in the xylem (Lang and Ryan, 1994). This is thought to result in the reduction in rates of xylem inflow to the fruit and a shift in the relative contributions of xylem and phloem saps to fruit growth (Lang, 1 990). This may have consequential effects on the balance and amounts of key fruit minerals and thus on fruit quality. Calcium, a xylem-mobile element (Marschner, 1 986), shows a progressive decline in its rate of accumulation during the season (Wilkinson, 1968; Jones et al., 1 983). A cultivar that is more susceptible to bitter pit (a low calcium induced disorder) showed an earlier slowing of xylem flow rate than a less susceptible one (Lang, 1 990; Lang and Ryan, 1994). A more recent analysis of 1 0 apple cultivars showed a significant relationship between the susceptibility of a cultivar to bitter pit and the extent of xylem dysfunction taken at a single time point (Drazeta et al. , 2001 ). The aim of this study was to elucidate the dynamics and the nature o f xylem failure in developing apple fruit using a water-soluble dye as an indicator of functionality coupled with a conventional light microscopic examination of dysfunctional bundles. Two apple cultivars were tested, these having different susceptibilities to bitter pit. It was hypothesised that the accumulation of vessel damage in the fruit might explain a Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 68 progressive reduction in xylem conductance. In addition, by recording the growth of the internal structures of the fruit, it may be possible to relate aspects of their growth to particular functional outcomes. These hypotheses may together be able to explain cultivar differences in bitter pit susceptibility in terms of differences in xylem functionality, and these in terms of differences in the growth of the internal structures of the fruit. The study should lead to an improved understanding of the mechanisms of bitter pit susceptibility in apple. 4.2. MATERIALS AND METHODS 4.2.1 . Plant material The study used mature trees of ' Braeburn' and ' Granny Smith' apple cultivars from a private organic orchard on the outskirts of Palmerston North, New Zealand. The trees were planted on MM 1 06 rootstock in three-row blocks alternating with pollinator blocks. The trees had not been managed according to standard commercial practices and the fruit was not selectively thinned. The flowering period was briefer than usual, with full bloom starting about three days earlier in ' Granny Smith' than in ' Braeburn' . 4.2.2. Temporal changes in xylem functionality Functionality of fruit xylem was recorded at weekly intervals until mid-season ( 1 1 3 and 1 1 6 days after full bloom, DAFB) and then fortnightly until harvest ( 1 83 and 200 DAFB). Average-sized fruit without visible blemishes were selected. Whole fruit-bearing spurs were collected from the orchard in the early morning (transpiration negligible, therefore, tree water potential close to soil water potential, - zero) and held in closed polyethylene bags pending investigation later the same day. On each occasion, fifteen fruit of each cultivar were assessed for xylem functionality. A total of 5 5 5 fruit were assessed throughout the season. Xylem function in the fruit was examined using a simple dye-infusion technique to stain the vasculature along the path of dye movement (Findlay et al., 1 987; DUring et al., 1 987; Creasy et al., 1 993; Wolswinkel et al., 1 999; Drazeta et al., 200 1 ). To avoid air embolism of the xylem, each fruit was cut from the spur at the base of the stalk. This was Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 69 carried out under distilled water prior to immersion in a small vial (a 2 cm3 Ependorf tube) containing a dye solution. The apoplastic dye ( 1 % aqueous acid fuchsin) was drawn up into the fruit through the stalk for 2 h under standard transpiration conditions (c. 22 °C and 65% RH) with a brisk airflow from two fans to remove boundary layer effects (Fig. 4. 1 ). Fruit were then sectioned equatorially using a microtome blade to minimise tissue damage and to produce a smooth cut surface. The vascular bundles were clearly visible in these transverse sections and were scored for dye intensity (D) as 0 = nil, 1 = trace, 2 = medium and 3 = high. Because developmental changes in xylem functionality may differ between the classes of vascular bundles, the presence of dye was assayed separately for the ten primary, five dorsal and ten ventral bundles. An averaged D value was computed for each fruit from the number of scores obtained for a particular class of bundle at any time. Fig. 4. 1 . Dye infusion of 'Granny Smith' (left) and ' Braeburn ' (right) apples, c. 70 DAFB. Stalks were immersed in vials containing 1 % aqueous acid fuchsin and exposed to the draught of two small fans. Some patchy dye accumulation can be seen at the fruit surfaces. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 70 Dye was assessed in four categories with each being assigned an integer value (i.e. 0, 1 , 2 , and 3). For such data, that is constrained between two extreme values (0 and 3), a logistic transfonnation can be applied to both linearise the data and to make the residuals more unifonn In( Y J = a + bx Yo - Y ( 1 ) where Yo i s the upper asymptote for D and was assessed as 3 . 1 (arbitrary units) for all data. Following this procedure, however, the transfonned data failed to summarise D from the sigmoid-like curves. Therefore, for each bundle type and cultivar, a model was chosen which described the approximate shape of the curve of changing D with time. Each model was fitted by non-linear, least-squares, using PROC NLIN (SAS, 2000). For each model, F-tests were applied to detennine whether selected features of the shape of the fitted curve were significant, and whether differences between the fitted curves for each cultivar could be adequately described by a simple difference in time scales. 4.2.3. Exposure time Preliminary staining trials conducted early in the season showed satisfactory dye distribution in the bundles of the fruit after c. 2 h. However, it was feared that changes in xylem functionality might alter the amount of dye appearing later in the season and so change the optimum period of exposure. Hence, it was worthwhile to assess dye accumulation as a function of time. For this, some 50 fruit of each cultivar were chosen in the age range o f 93 - 1 04 DAFB and infiltrated with dye for 3 0, 60, 1 20, 240 and 480 min. The three classes of vascular bundles were assessed for D and averaged separately for each fruit as described previously. 4.2.4. Spatial changes in xylem functionality Longitudinal dye accumulation in the fruit was investigated in ten fruit of each cultivar at a more advanced stage ( 1 60 1 63 DAFB). Dye-infiltrated fruit were sliced transversally using a domestic food slicer to fonn a contiguous series o f c. 2 . 5 mm thick tissue disks, starting from the calyx end. Due to a slight variability in fruit size, different numbers of ---------------- - -- - - - Chapter 4. Causes and effects of changes in xylem functionality in apple fruit disks were obtained from each fruit. D was assessed in each disk (not all classes of bundles were visible in the disks obtained near the ends of the fruit). PROC GLM (SAS, 2 000) was used to evaluate the significance of any longitudinal gradient in D for each bundle class and cultivar then, secondly, to test whether the gradients differed: • between bundle classes for each cultivar, and • between cultivars for each bundle class The high natural variability between fruit was taken into account by including a fruit effect within each cultivar. 4.2.5. Anatomy of xylem breakage 7 1 S everal fruit were collected at 8 5 and 1 4 1 DAFB for microscopic study. Fruit from both collection dates were infiltrated with dye and cut transversally into three, approximately equal, disks (using a sharp knife): disc 'A' was from near the stalk-end of the fruit, ' B ' was from the middle and ' C ' from near the calyx-end. I n each disc, randomly chosen dysfunctional (unstained) bundle was marked on the skin along assumed path and excised with a small amount of attached flesh so as to produce an elongated sample. Fruit picked at 85 DAFB were used to identify structural changes in the xylem. Each sample was soaked for several minutes in an excess of phloroglucin solution (2. 5 g in 5 0ml of 70% v/v aqueous ethyl alcohol) and flooded afterwards with concentrated HCI to visualise lignified vascular elements. This procedure yields distinctively red-stained xylem strands embedded within a reasonably transparent parenchyma. To avoid rupture of xylem during the final tissue preparation, the bulk of the flesh was carefully dissected away with a microtome blade and fine forceps working under a dissecting microscope (Wild M3Z). The excised bundle was mounted in 80% w/w glycerol on a glass slide and examined with a compound light microscope (Microlux- l l ) attached to a video-camera system. Fruit picked at 1 4 1 DAFB were used to investigate the nature of xylem disruption by a more sophisticated light microscopic procedure. To prepare dysfunctional bundle for fixation, intercellular air was first removed by evacuation. To do this the sample was Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 72 placed in a 50 cm3 vial containing water and wetting agent (0. 1 % w/w Silwet L-77), and exposed to reduced atmospheric pressure (c. 1 8 mmHg at 20°C) for 2 min. After the vacuum had been released, water flowed in to replace the extracted air over a 2 min period. This treatment rendered the parenchyma reasonably transparent and suitable for dissection. Samples were trimmed with a sharp blade under the dissecting microscope (Wild M3Z) until the bundle was exposed with a very thin layer of surrounding flesh. It was then cut transversally every 4 - 5 mm. A total of five, four and three specimens were obtained from disks A, B and C respectively. Once ordered, (AI to As) the specimens were preserved separately in sealed glass vials containing half-strength Karnovsky' s fixative (3% v/v glutaraldehyde and 2% w/v formaldehyde in 0. 1 M phosphate buffer at pH 7.2). Secondary fixation was in 1 % osmium tetroxide using the same buffer. Specimens were dehydrated in a graded acetone series, and infiltrated and embedded in Procure 8 1 2 epoxy resin. Cured resin blocks were trimmed and longitudinal sections ( l /-tm thick) were cut from the middle o f the bundle, heat mounted on glass microscope slides and stained with 0. 1 % toludine blue. Tissues were prepared for microscopy and sections by Mr Douglas Hopcroft, EM Unit, HortResearch, Palmerston North. Slides were examined and photographed using a Zeiss Axioplan compound light microscope with attached MC 1 00 camera system. A view of xylem breakage in primary bundles was obtained from the same fruit used for photomicrograph examination ( 1 4 1 DAFB). Fruit were cut transversally (with a sharp knife) into three uniform slices and each was re-cut longitudinally (with a microtome blade) along the path of unstained bundles so as to expose regions where dye abruptly stopped. Specimens were photographed using a Wild M3Z dissecting microscope with attached 3 5 mm Leica-Wild MPS5 2 camera system. 4.2.6. Growth of the different fruit tissues To evaluate the growth of the different fruit tissues, dimensional measurements relative to the axis were made in the fruit of both cultivars (Fig. 4.2). These were measured in the same fruit used for dye infusion. One extra fruit of each cultivar at each time was cut equatorially using a microtome blade. Measurements were made using digital callipers (precision 0.01 mm). Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 73 Radial distances were taken from the fruit axis to the skin through the primary (sepal and petal) bundles. These gave a measure of the absolute tissue growth (ATG) of the core and cortex (assuming that the vascular ring delineates the boundary between two fleshy portions ofthe fruit) and of the fruit itself. The length of the cartilaginous walls defining locules was also recorded. The dimension of the fleshy region between the tip of the locule and the adj acent sepal bundle was obtained by difference. Relative tissue growth (RTG) was obtained by normalising the dimensional tissue measurements with the corresponding fruit radii to determine how the various structural components contributed to the growth of the whole. S l ----+---_Q; ".-."��-t;l---� P 1 L 1 ----f--�.,4.--- Fig. 4.2. Transverse section of fruit showing radial dimensions measured from the fruit axis to the skin. Here S I refers to sepal bundle whereas P I refers to petal bundle. Hence, distance from the fruit axis to L l gives a measure of locule, to S 1 and P I gives a measure of core (ovary) and to S l ' and P I ' gives a measure of fruit dimensions respectively. The difference between fruit and core dimensions quantifies the size of the cortex. Following this approach, five locule dimensions and ten core, cortex and fruit dimensions were obtained for each fruit. In addition, the distance between the locule tip (Ll ) and the sepal bundle was assessed. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 74 Observations of vascular functionality made about midway through the season revealed that the sepal bundles of the cortical vascular system of ' Granny Smith' retained functionality for longer than the petal bundles. This distinction was not apparent in ' Braeburn ' . The spatial arrangements of the petal and sepal bundles were assessed in both cultivars. Each point was defined by the ratio of the distances of the petal and sepal bundles to the fruit axis (ten individual measurements) and so it was necessary to carry out analysis by assessing the mean ratio for a single fruit. For each cultivar PROC GLM (SAS, 2000) was used to evaluate the significance of the slope of the mean ratio of the radial distances to petal and sepal bundles in the fruit (PIS) vs DAFB. It seemed reasonable to assume that the degree of sepal bundle drift outward was related to the magnitude of ovary growth in the same direction. This emphasised the importance of the proximity of the sepal bundle to the neighbouring locule (S-L) normalised with the corresponding fruit radii. The significance of cultivar differences in the time course for the mean ratio was compared using PROC GLM (SAS, 2000). 4.3. RESULTS 4.3.1 . Exposure time Dye infusion induced a rapid spread of dye throughout the functional portions of the vascular network enabling an assessment of functionality for the range of exposure times used. In all bundle classes D initially rose but reached a plateau after 1 20 minutes (Fig. 4.3). However, different bundle classes accumulated different amounts of dye. In both cultivars, the dorsal bundles appeared more functional than the primary or ventral bundles. Furthermore, the intensity of dye in the primary bundles of ' Braebum' was less than in ' Granny Smith' Chapter 4. Causes and effects of changes in xylem functionality in apple fruit I � � 0.5 -I 0.0 I I 0.0 L....>.. .J......I ...I.-ooi.-I.....o..... ........ .J......I ...I.-ooi.-I.....o.....J..,.,J o 60 1 20 1 80 240 300 360 420 480 0 60 1 20 1 80 240 300 360 420 480 Period of exposure (min) Period of exposure (min) 75 Fig. 4.3. Dye accumulation in the vascular systems of ' Braeburn' (A) and ' Granny Smith' (B) as a function of the period of exposure. The points are mean values (n = 1 0) for the primary (.), dorsal (.) and ventral (.) bundles of the fruit with the corresponding SE indicated by the vertical bars. Fruit were assessed at 93 DAFB ( 'Braeburn') and 1 04 DAFB (' Granny Smith') . 4.3.2. Temporal changes in xylem functionality Changes in dye accumulation were observed initially in the anastomosing vasculature, which ramifies into numerous branches to form a dense network of fine bundles that scatters across the flesh and terminates just beneath the skin. Early in the season, dye spread freely through the numerous fine traces leading to a somewhat diffuse appearance. Later, the fine vascular network was less strongly coloured. The reduction in colour took place before any noticeable change in the functionality of the primary bundles and followed a similar pattern in both cultivars, although it was more pronounced in ' Braeburn' (Fig. 4.4). Periodic observations of the stalk apex revealed no signs of dysfunction with the bundles appearing completely stained all through the growth period. This strongly suggests that xylem dysfunction occurs exclusively in the fruit. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 76 Fig. 4.4. Transverse sections of dye-infused 'Braeburn' (left) and 'Granny Smith' (right). Fruit were assessed at 64 and 67 DAFB respectively. The degree of coloration in fine vascular anastomoses indicates a difference in vascular function between the two fruit. The dye level in the primary bundles was at the maximum measurable value of 3.0 early in the season. After this colouration declined exponentially with time (Fig. 4.5), so that a model was chosen of the form { 3, Dprim = 3 exp(-k(t - to )) (2) where to (the time, DAFB, when D first drops below 3.0) and k are parameters to be fitted. The exponential decay parameter k can be related to a 'half-life' (ty,), the time interval for D to be halved from any point on an exponential curve by log(2) tx = k (2a) For both cultivars, the decrease in D with time was highly significant (p � 0.00 1), i.e. k "# 0, and by harvest fruit have lost most of their xylem function. Differences in the extent of xylem dysfunction between the two cultivars were well described by a single time­ scaling factor, with the difference being highly significant (p � 0.00 1 ), and with 'Braeburn ' taking just 0.67 times as long as ' Granny Smith' to decay at any given D. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 77 Using this scaling factor, fitted parameter values were estimated at to = 67 DAFB and tv, = 6 1 days for ' Granny Smith', and to = 45 DAFB and tv, = 4 1 days for 'Braeburn' . The sudden collapse in xylem function occurs earliest in ' Braeburn' , whereas in ' Granny Smith' function continues for longer and the fruit still retains some functionality at harvest. 3.0 ,-. 2.5 � '§ � 2.0 .!::l :.e 1 5 � . i .£ 1 .0 � Cl 0.5 0.0 I..--......... -.L..-"'_I..--......... -.L..--'-_I---'---J...--'-_I---'---J... .... _I...--.J o 25 50 75 1 00 1 25 Time (DAFB) 1 50 1 75 200 Fig. 4 . 5 . Dye level in the primary (cortical) bundles of 'Braeburn' (e) and ' Granny Smith' (0) as a function of fruit age (DAFB). The mean dye level for each time period was obtained by assessing 1 5 fruit per cultivar. Vertical bars represent SE. The dye level in the dorsal bundles (Fig. 4.6A) rose initially to a maximum value (A), but then followed a somewhat complex path. The conductance in both cultivars oscillated markedly but the xylem remained fairly functional throughout the season. As there was no obvious mechanism by which bundle conductance, and therefore D, could increase around day 1 25, this fluctuation was ignored and the path was described by a declining logistic curve - � - --------- Chapter 4. Causes and effects of changes in xylem functionality in apple fruit Dd = A' { A + m(t - to ), ors 1 + exp(k(t - t h » where A, to (the time, DAFB, at which D starts to decay), m (the slope up to the 78 (3) maximum D), th (the time, DAFB, when D drops to half of A ,), and k are parameters to be fitted. To ensure continuity, the upper asymptote of the logistic curve At= 1 + exp(k(to - th ) . For a logistic, parameter k can be related to the times (DAFB) at which D drops to two thirds or one third of A I ( t � or t}:j ), by log(2) t 2/ , t I/ = t h ±--'O. ___ /3 73 k (3a) As was the case for the primary bundles, the drop in D was significant (p s 0.00 1 ), and using separate values for every parameter in each cultivar provided no significant improvement over using a simple scaling factor. The difference in time scales was significant (p s; 0.05), with the fitted scale factor of 0.8 5 . With this scaling factor, fitted parameter values for ' Granny Smith' were m = 0. 1 9, to = 42 DAFB, and th = 1 72 DAFB (with t� and t}:j 25 days before and after this), and for Braebum were m = 0.23, to = 3 6 DAFB, and th = 1 46 DAFB ( t� , t}:j ± 2 1 days). For both cultivars, A = 2.2. The dye level for ventral bundles (Fig. 4.6B) also rose initially to a maximum value (A), but dropped roughly exponentially to a non-zero final value (B). Hence, a model to describe these characteristics was { A + m(t - to )' Dvellt = B + (A - B) exp(-k(t - to» (4) Parameters to be fitted are A, B, m, to, and the exponential decay parameter k, which is related to a 'half-life ' (ty,), the time interval for D to halve along exponential curve relative to the final level B by (2a). For both cultivars, the drop in D with time following to was highly significant (p s 0.00 1 ), and the final level reached (B) is significantly greater than zero (p s; 0.00 1 ). Nevertheless, there were no significant differences between parameters for the two cultivars, with the fitted scaling factor (0.97) not differing significantly from one. For Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 79 both cultivars, fitted parameter values were approximately A = 2.3, B = 0.6, m = 0. 1 , 10 = 43 DAFB, and Iv, = 1 2 days. 3.0 ,-.., 00 ."t:: 2 5 c: . ::l � 2.0 -:e 1 . 5 C':I '-' � 1 .0 ;;. � � 0.5 A 0.0 3.0 ,-.., 00 ·S 2.5 ::l C 2.0 e: ...-:e 1 .5 C':I '-' '0 1 .0 ;;. � � 0.5 Cl 0.0 0 0 A. • 2 2 25 50 75 1 00 1 25 1 50 1 75 200 B. 0 25 50 75 1 00 1 25 1 50 1 75 200 Time (DAFB) Fig. 4.6. Dye level in the dorsal (A) and ventral (B) carpellary bundles of 'Braeburn ' (e) and 'Granny Smith' (0) as a function of fruit age (DAFB). The mean dye level for each time period was obtained by assessing 1 5 fruit per cultivar. Vertical bars represent SE. 4.3.3. Spatial cbanges in xylem functionality Serial (lengthwise) sectioning of dye-infused fruit showed a strong spatial gradient in D with increasing distance from the stalk end of the fruit (p � 0.01 ), except for ventral bundles in ' Braeburn ' (Fig. 4.7). Mean dye level is plotted against longitudinal position in the fruit, where ' 0 ' refers to the point near the stalk end where the vascular bundles were first apparent, and ' 1 0 ' refers to the last such point close to the calyx end. However, Chapter 4. Causes and effects of changes in xylem functionality in apple fruit the ventral bundles in ' Braebum' were not clearly distinguishable at the ' 0 ' position in the fruit (missing value, Fig. 4.7 A) while some data points in other vascular bundles of the same cultivar were determined only by a single value and lack error bars. 1 .8 1 .6 -'"' !!l 1 .4 .;:: ::l 1 .2 C £ 1 .0 � 0.8 '-' "0 0.6 ;. � 0.4 IU >-.. Cl 0.2 ,.-, 0.0 1 .8 1 .6 !!l 1 .4 .;:: ::l 1 .2 C £ 1 .0 � 0 8 --- "0 0.6 a> - 0.4 � Cl 0.2 0.0 A. A . A. T ���V'H t-�t-+-O-f-� B. lr B. B. � j �-t 1 . l \tH�� ~ o 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 1 0 0 1 2 3 4 5 6 7 8 9 1 0 Position (arbitrary units) Position (arbitrary units) Position (arbitrary units) 80 Fig. 4.7. Longitudinal changes in the functionality of xylem in 'Braebum' (A) and ' Granny Smith' CB), assessed at 1 60 and 1 63 DAFB respectively. Each point represents the means dye level for the primary (e), dorsal (A), and ventral bundles C-) falling within the same fractional distance down the fruit. Vertical bars represent SE. For both cultivars, there were significant differences among the gradients for the bundle classes (p $; 0.0 1 ), with dorsal bundles having the steepest gradient for both. The gradient for dorsal bundles in ' Granny Smith' was significantly greater than the gradient for 'Braebum' (p $; 0.0 1 ), but there were no significant differences in the gradients between cultivars for primary or ventral bundles. Chapter 4. Causes and effects of changes in xylem fonctionality in apple fruit 8 1 4.3.4. Anatomy of xylem breakage Longitudinal image of the dysfunctional primary bundle of 'Braeburn' shows interruption of dye movement close to the stem-end of the fruit (Fig. 4.8). The collapse of xylem function was evident through the sudden cessation of dye import beyond the point of disruption leaving the bundle completely unstained (dysfunctional) when assessed in transverse section. Diffuse dye in the flesh around the distinctively stained part of the bundle confirms that a substantial amount of apoplastic dye enters the fruit but that dye passage in the bundle can be blocked. Fig. 4.8. A dysfunctional primary bundle in 'Braeburn' ( 1 4 1 DAFB). Discontinuity of dye distribution through the bundle indicates the point of dysfunction. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 82 Microscopic examination of dysfunctional primary bundles confirmed a localised area of damage in the xylem tissue (Fig. 4.9). The conducting elements of the xylem consist mainly of vessels organised into parallel columns with spiral thickening of the walls, although occasional annular thickening is also apparent. The most conspicuous damage appears as an irregular spacing and deformation of the vessel files. Annular structures sometimes collapse entirely leaving occasional rounded elements completely isolated. At other times, spirals appear to be overly stretched and partially fragmented (Fig. 4.9A-B). That this characteristic was an artefact of sample preparation is j udged unlikely due to the existence of vessels with intact walls adjacent to the deformed and fragmented structures. Furthermore, many vessels were seriously damaged with helices very stretched (Fig. 4.9C-D). Additional structural changes were apparent when longitudinal sections of dysfunctional primary bundle were assessed for xylem damage in the advanced stage of fruit growth. The evidence showed that complete breakage of the xylem strand occurred randomly along the bundle. Cells of parenchyma filled the spaces between files of vessels indicating total discontinuity and collapse in the function (Fig. 4.9E). Within a xylem strand, stretched vessels with spiral thickening were frequent, often sustained in a broken line, which reflects the high degree of strain (Fig 4.9F). Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 83 Fig. 4.9. Longitudinal sections of a primary (sepal) bundle in the fruit of ' Braebum', 85 DAFB (A-D) and 1 4 1 DAFB (E-F) respectively. (A-B) Irregular spacing of the annular and spiral wall thickenings exists alongside undamaged cell structures. (C-D) Deformed and fragmented vessels. Arrangement of helices in the broken segments indicates stretching. (E) Breakage of the xylem strand. Shattered parts are displaced so that parenchyma cells occupy the resulting void. (F) Ruptured string of vessels. Multiple breaks signify excessive change in length of the stressed walls. The bar in A-F represents 5 0 /-lm. Arrows indicate the structural changes being described. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 84 4.3.5. Growth of the different fruit tissues 4.3.5.1 . Absolute Tissue Growth (ATG). Fruit structures in both cultivars exhibited a sigmoidal growth pattern (Fig. 4. 1 0). A linear increase in the earlier stages was followed by a gradual decline in growth rate as the various dimensions of the tissues approach their maximum values. The curves of the various fruit tissues i llustrate the difference in their growth dynamics. Increments in the locule dimensions flatten earlier in comparison with other tissues. Locule growth, therefore, makes the least contribution to overall fruit growth. The growth of the core continues for some time after the slope of the locule growth flattens as a result of prolonged flesh (pith) enlargement within the core. Overall fruit growth continues almost until harvest. The two cultivars exhibit similar growth patterns, although the slopes of 'Braeburn' growth are somewhat steeper. 40 A. � 40 B . 35 35 30 30 � � 25 25 '-" (1) N 20 20 ...... r.t> (1) ;::1 r.t> 1 5 1 5 r.t> ...... f-; 1 0 1 0 5 5 0 0 0 25 50 75 1 00 1 25 1 50 1 75 200 0 25 50 75 1 00 1 25 1 50 1 75 200 Time (DAFB) Time (DAFB) Fig. 4. 1 0. Absolute growth of various tissues in the fruit of 'Braeburn' (A) and ' Granny Smith' (B). The points are mean values for a single fruit at each time period derived from five locule (.), or ten core ( .... ) and fruit dimensions (.). The difference between core and fruit curves represents the cortical tissue. Vertical bars represent within-fruit SE. ------------------- - � - - - - Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 85 4.3.5.2. Relative Tissue Growth (RTG). The data on ATG were transformed to represent the contributions of a particular tissue to the growth of the whole. Although absolute fruit growth involves rapid increases in the core structures, it is the cortical tissue that accounts for most of the enlargement of the fruit in both cultivars (Fig. 4. 1 1 ). The difference between relative locule and core sizes shows the magnitude of changes in the size of the fleshy pith. Nevertheless, the declining trend in the relative size of the core, compared with the steady increase in the relative size of the cortex, provides key evidence of the relative contribution of these tissues to overall fruit growth. 90 A. 80 � 70 Cl) .!::l 60 fI) g 50 � '':; 40 Cl) .� 30 t ·c 25 '"0 (l.) ...... '"0 I=: ;::$ 20 ,!:) a (l.) 1 5 .S .� 5 0 ..... .s � 0 0 25 50 75 1 00 1 25 1 50 1 75 200 Time (DAFB) Fig. 4. 1 3 . Relative increase in the distance between locule tip and the sepal bundle normalised with the corresponding fruit radii in 'Braeburn' (e) and ' Granny Smith' (0). The points are mean values for a single fruit at each time period. Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 88 Fig. 4. 1 4. Dye distribution in 'Braebum' (left) and ' Granny Smith' (right) apples infiltrated with dye at 1 69 and 1 72 DAFB respectively. Note that more sepal bundles in ' Granny Smith' contain more dye and hence presumably remain functional in this fairly advanced stage of maturity. 4.4. DISCUSSION 4.4.1. Exposure time Because the dye level in the vascular bundles appeared stable after 1 20 minutes (Fig. 4.3) any differences in D can be taken to reflect a physical property of the bundle itself rather than of the measuring technique. 4.4.2. Temporal changes in xylem functionality For the primary and ventral bundles in the fruit, the model developed provided a good fit to the mean curves for both cultivars (Figs. 4.5 and 4.6B). For dorsal bundles, the model could not capture the later rise in D, but otherwise fitted reasonably well (Fig. 4.6A). The decline in D was significant for all three bundle-classes. Nevertheless, dysfunction was irregular, with some fruit losing the greater part of their xylem conductance relatively early on. The majority of bundles exhibited reduced dye movement, which indicated random damage to the xylem at any one time. Hence, the trend in D at any stage of the - - --------------- Chapter 4. Causes and ef ects of changes in xylem fimctionality in apple fruit 89 growth period reflected averaged xylem conductance based on the prevailing condition of individual bundles. Consequentially, the few operational bundles close to harvest often masked the actual extent of xylem collapse in nearly all bundles of the fruit. The most marked change in xylem conductance occurs in the primary bundles (Fig. 4.5) that are the principal nutrient suppliers of the flesh (MacDaniels, 1 940). The highest xylem flow of these bundles was obtained at the beginning of the season. The early­ season plateau in dye accumulation could be seen as an artefact of the arbitrary dye assessment that limits the maximum to which D can rise to an integer value of 3 . Nevertheless, D i n both cultivars exhibited identical trends indicating that the primary bundles were fully functional at the beginning of fruit development. A similar pattern in functionality of the fruit xylem was obtained in different apple cultivars examined using a pressure-bomb technique (Lang and Ryan, 1 994; Drazeta et al., 200 1 ). As the season progressed, an increasing proportion of primary bundles failed to transport apoplastic dye. For 'Braeburn' , changes in D were more extreme, and xylem conductance became negligible over the final weeks of fruit development. A fitted time-scaling factor of 0.67 suggests that xylem conductance in 'Braeburn' breaks down in roughly 2/3 o f the time it takes for conductance breakdown in ' Granny Smith ' . The sharp decline in D from around 45 DAFB ( 'Braeburn') and 67 DAFB ( ' Granny Smith ' ) closely matches the completion o f vessel differentiation (c 5 0 DAFB) in the primary bundles of ' York Imperial' apple (Barden and Thompson, 1 963). This indicates that cessation of cambial activity in the bundles of the fruit coincides with the onset of xylem dysfunction. A mechanistic explanation for the reduction in xylem conductance proposes growth­ induced damage to the xylem strand in the bundle (Findlay et al., 1 987; DUring et al., 1 987; Lang and Ryan, 1 994; Drazeta et al., 200 1 ). However, the function of the phloem remains unchanged relative to the xylem flow (Lang, 1 990). The hypothesis that fruit expansion stretches and ruptures the xylem provides no elucidation o f the mechanism that maintains phloem function in the bundle. Because phloem tissue is composed o f living conducting elements (Esau, 1 977), the persistence of functional phloem could b e attributed to; Chapter 4. Causes and effects of changes in xylem functionality in apple fruit • persistent sieve tube differentiation that bypasses the broken elements, since the phloem differentiates in the absence of the xylem and at lower auxin levels (Aloni, 1 980; Aloni and Barnett, 1 996), • the more extensible cell walls of the sieve tubes (Lee, 1 98 1 ) with a more prolonged synthesis of cell wall depositions enabling the sieve tubes to accommodate more strain, and • continuous symplastic pathway through parenchyma cells that occupy the voids between the broken structures. 90 If fruit growth principally creates stress on conductive tissues it is likely that any one, or some combination, of these features allows continuous phloem uptake into the fruit. Cortical and carpellary vascular systems vary in the dynamics of the xylem breakdown. Dye accumulation was somewhat less disrupted in the carpellary bundles that are seemingly less conductive in regard to their principal function in the fruit (Figs. 4.5 and 4.6). Comparing with primary bundles, it is apparent that dorsal bundles are less affected with a progressive but slower loss of xylem function (Fig. 4.6A) but, similarly, the conductance appears to drop without a lower limit. Changes in D once again occur more rapidly for 'Braebum' and although statistically significant, cultivar differences were smaller and less regular. Fitted values suggest early (36-42 DAFB) but rather gradual decay in D that takes much longer to reach half its maximum value ( 1 42- 1 72 DAFB). Ventral bundles exhibited distinctively lower D at most stages throughout the growth period (Fig. 4.6B). After the initial rise, xylem functionality declined sharply but proceeded with a rather low, yet steady flow following an exponential decay. Maximum D was reached at about 43 DAFB, than dropped exponentially with a 'half-life' of only 1 2 days (ty,). For most of the fruit evaluated, the ventral bundles accumulated substantially less dye and often appeared unstained. The lower D in the dorsal and ventral bundles could result from naturally reduced dye uptake through the carpeUary network, which develops well below the skin and appears less affected by transpiration. Dorsal bundles that follow the outlines of the carpel bases may undergo seasonal changes in xylem function closely related to the growth rates of adjacent tissue. Hence, the xylem in the dorsal bundles may have been stretched beyond Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 9 1 its breaking strain due to a strong increase in the size o f the carpellary tissue. Nevertheless, considering that the dorsal bundles branch from the primary bundles at the base of the fruit, the late season decline of flow may indicate blockage beyond the point of separation. If flesh expansion towards the stem-end upsets primary bundle function near the stalk that, in return, could affect the conductance of the branching dorsal bundles. In contrast, the ventral bundles principally serve the seeds (MacArthur and Wetmore, 1 939), which presumably require less nutrient (diminished water influx) than the flesh. This, and the fact that the ventral bundles develop below the carpel bases, may affect dye accumulation in a way that dye spreads slowly throughout the ventral network, which results in consistently lower D when assessed in transverse sections. Simultaneously, the establishment of reproductive structures determines the course of vascular function such that ovule abortion causes degeneration of the supporting ventral bundles (MacArthur and Wetmore, 1 939). Considering that seed abortion in developing apples is continuous throughout the season (Lang, unpublished), it seems reasonable that functionality of the ventral bundles is somewhat hampered compared with the other bundles. However, the origin of the functional disturbance at the beginning of the season remains unknown. The course of xylem dysfunction could be of special importance to the mineral composition o f the fruit. A steady decline in xylem conductance under constant fruit transpiration must force a shift in the balance of supply to the fruit of xylem-borne nutrients, notably calcium. Indeed, xylem function decays from relatively early in the season, which mirrors the decline in calcium import (Wilkinson, 1 968; Jones et al., 1 983). This naturally-occurring mechanism could be principally responsible for the increased phloem inflow to the fruit that maintains growth (Lang, 1 990). Hence, programmed xylem breakdown is likely to reduce apoplastic backflow of solutes (from fruit to the tree), which promotes the partitioning of assimilates to reproductive sinks (Lang and Thorpe, 1 989). A rather variable timing of xylem dysfunction can, therefore, create high variability in fruit mineral composition. This could explain the high variability observed in the incidence of calcium-related disorders (Drazeta et al., 200 1 ). - - � --����----------- Chapter 4. Causes and effects of changes in xylem fonctionality in apple fruit 92 Hence, the earlier start of xylem dysfunction in ' Braeburn' fits with the observation that, of the two examined cultivars, ' Braeburn' is the more susceptible to bitter pit. 4.4.3. Spatial changes in xylem functionality Significant changes in xylem function were established using contiguous sectioning to evaluate D in vascular bundles. The most conspicuous gradient was obtained in the dorsal bundles that on average scored higher D values than the other bundle classes and differed between the two cultivars (Fig. 4.7). However, dye concentration declined in nearly all bundles with increasing distance from the stalk end of the fruit suggesting gradual disruption of the flow along the bundles. If the xylem vessels are only partially broken, some dye can still move through the bundle, whereas a complete gap will stop all flow at this point. The extent of xylem dysfunction signifies the direction and relative intensity of fruit expansion. By cutting ' Gala' through the centre of the fruit and measuring the increase in the ' gape ' openings, Opara ( 1 993) showed that the first noticeable growth stress occurred around 60 DAFB and reached maximum just before harvest. Moreover, the extent of gaping was greater in a vertical than a horizontal cut. Thus it may be inferred that internal fruit growth is not uniform in the two directions, leading to an asymmetry of the resulting growth-induced stress. Two possible explanations emerge from this : • The extent of xylem dysfunction could be greater in the faster-growing regions of the fruit due to the larger strains generated, or • Tensile stresses created by an expanding fleshy matrix develops evenly along the bundle path so that xylem breakage appears at random as fruit growth continues. The appearance of multiple breaks along each bundle suggests that bundles are well attached to the expanding flesh tissues. The trend in longitudinal dye accumulation indicates irregular multiple breakages. For this reason, spatial assessment o f D supplements the data on seasonal xylem functionality although fruit were examined only once, close to harvest maturity, and contains no information on seasonal breakdown in regard to the position within the fruit. - --- - --------------- - - - - - Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 93 The spatial patterns in xylem function are important to postharvest physiology. The reduction in dye accumulation towards the calyx end of the fruit appears to fit with the reported longitudinal gradient in calcium concentration found in apple fruit (Lewis and Martin, 1 97 3 ; Lewis, 1 980). As the fruit expands, progressive xylem dysfunction is likely to result in a reduction of calcium uptake and displacement within the fruit, leading to the uneven spatial distribution and imbalance with phloem-borne minerals. Consequentially, the distal regions of the fruit are depleted in calcium and are at higher risk of developing calcium-related disorders than the proximal zones. Indeed, bitter pit lesions are commonly observed near the calyx-end of apple fruit (Ferguson and Watkins, 1 989). This suggests the localised nature of the disorder and fits with the spatial profile of dye accumulation in the fruit. 4.4.4. Anatomy of xylem breakage The forms and arrangements of wall thickenings in dysfunctional primary bundles of ' Braebum ' (Fig. 4.9A-D) conform to the pattern of vascular ontogeny in the developing plant body (Esau, 1 9 77). Scattered elements with annular thickening represent the remnants of pro to xylem structures that probably cease to have functional importance as soon as organ expansion obliterates their structure. In contrast, spiral depositions dominate the wall thickenings of later formed vessels and appear in two distinctive forms. Deformed and fragmented spiral elements are possibly the earliest metaxylem structure, initiated before protoxylem growth was complete. The common ' uncoiled' appearance of the thickening helices of the broken vessels offers strong evidence of excessive structural damage due to growth-induced strain. The remaining vessels in the xylem strand have less steep helices and appear functional, which suggests that later vessel differentiation has occurred to replace the older, overly stretched and dysfunctional elements. The mechanical properties of vascular elements make it possible to assign likely changes in the xylem of developing fruit. When a growth-induced strain is applied, cell walls will stretch resulting in the fractional change in length (strain), which is specific for a given structure (Nob el, 1 999). The magnitude of breaking strain, therefore, relies on the mechanical properties of the xylem tissue. Lignified tube-like conducting cells of the xylem seem to provide good mechanical support but exhibit rather low elasticity. When Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 94 expressed as a percentage of the broken segment length for the peripheral vascular bundles in the grape berry, first breakage occurs close to veraison, with fractional change of c. 3% (DUring et al., 1987). This bears close resemblance with calculated breaking strain of c. 2-3% for xylem fibers in a variety of woody species (Kollmann and Cote, 1 968). Furthermore, the magnitude of cell wall deformation is proportional to the time of applied stress (Nobel, 1 999), indicating that the growth dynamics could principally determine cell functionality in stressed tissue. Indeed, high fruit growth rates decrease calcium supply (Marschner, 1986), and larger fruit are usually more prone to the incidence of bitter pit (Ferguson and Watkins, 1 989). Vigorous fruit growth, particularly from the onset of cell expansion is likely to initiate earlier xylem dysfunction, as a result of which, the relationship between xylem conductance and potential fruit size must not be taken at the time of maturity but expressed periodically. In addition, the persistence of xylem conductance depends not only on the active tissue growth but also on the amplitude of the vessel differentiation by the vascular cambium. Only knowledge o f the periodicity and intensity of these processes could elucidate actual dynamics o f xylem dysfunction. 4.4.5. Growth of the different fruit tissues Dimensional measurements of ATG and RTG gave a more comprehensive picture of the growth pattern in the two cultivars. It should be noted, however, that the growth of fruit structures was obtained by measuring linear dimensions in a growing spherical organ, which does not capture the exact picture of its volume increments (Bollard, 1 970). The exponential phase in ATG was followed by diminishing growth later in the season, which implies a sigmoidal growth pattern as commonly reported in apple (Opara, 2000). In contrast, RTG accounts for growth efficiency of a particular part relative to the growth of the whole, which allows for more equitable inter-tissue comparisons. From the beginning of the growth measurements, the core and cortex showed different relative contributions such that cortex development dominated the change in fruit size (Fig. 4.1 1 ), which is in the line with the relative contribution of different tissues to the overall fruit size (Smith, 1 950). Chapter 4. Causes and effects of changes in xylem functionality in apple jiiiit 95 The patterns in fruit tissue growth refined the idea of specific structural features leading to analogous physiological responses. Because xylem collapse in the bundles is related to the growth of the fruit, the extent of damage depends upon the rate of tissue enlargement. Although fractional increase in fruit size is mainly driven by cortical enlargement, it is the rate of core growth that governs the intensity of xylem dysfunction in primary bundles. Since fruit development initiates a drift o f tissues outwards, growth-induced stretching of primary bundles is exclusively determined by core expansion. The dissimilar spatial layout ofthe primary bundles observed between ' Granny Smith' and 'Braeburn' conforms with the fact that the cross-sectional outline of apples is one of the structural features used for pomological description (Bultitude, 1 983). In mature ' Granny Smith ' , primary bundles form a somewhat angular pattern (with petal bundles outlying the arrangement of sepal bundles), whereas in ' Braebum' the layout of primary bundles is ring-shaped. This is in line with conspicuous cultivar differences in the mean ratio of the radial distance to primary bundles (Fig. 4. 1 2), and coincides with the prolonged functionality of sepal bundles in ' Granny Smith ' . This observation implicates a 'protective' role o f the locule. Slower growth o f a locule is likely to reduce early season motion of sepal bundles that are positioned closer to the locule. This can also restrict the influence of core enlargement by producing fewer cell layers of the pith. Hence, the magnitude of the drift of the sepal bundles is not fully attained at the time that locule growth ceases, since the core continues to grow through the increase of the pith (Fig. 4. 1 0). As a result, the difference in drift of sepal bundles appears by the latter half of the season and increases towards fruit maturity (Fig. 4. 1 3). However, no degree of protection applies to the petal bundles that are fully exposed to the rapid flesh growth. These morphological and physiological aspects of tissue growth emphasise the significance of variations in the internal structure of apple fruit. Since the pattern of fruit growth is a peculiar cultivar feature (Pratt, 1 988), particular structural attributes could determine functional response of the whole fruit. Moreover, the extent of xylem breakdown as the season progresses is of considerable relevance to the nutrition of the expanding fruit. It was observed that some earlier-maturing cultivars grew more rapidly Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 96 (Tukey and Young, 1 943), which could lead to premature xylem dysfunction and change the balance of minerals in the developing fruit. The importance of growth dynamics is further strengthened by the fact that calcium concentration peaked early in the season (lones et al. , 1 98 3), when the rate of growth was notably slower (Percy et al., 1 996). Consequentially, the outcome of increased growth rates over the brief period of time (i.e. early in the season) could result in greater calcium deficiency. 4.5. CONCLUSIONS The phenomenon of progressive xylem dysfunction in apple fruit was studied by assessing the intensity of dye infusion in the bundles. The structural nature of the changes in function of the xylem was assessed using a microscopic procedure. Fruit growth dynamics were also analysed and critical morphological features were identified. The literature clearly supports the pattern of xylem dysfunction, the observation of physical damage to the xylem vasculature and features of the gross development of fruit structure. This study has shown that the xylem vasculature of apple fruit suffers progressive dysfunction during the growing season. Visual signs of physical disruption of the bundles seem to explain this dysfunction. Dysfunction was greatest in the primary bundles suggesting a likely impact on fruit mineral uptake. Interestingly, in the two cultivars studied, dysfunction occurred earlier and was more pronounced in the one with the heightened susceptibility to bitter pit. Spatial patterns of dye infusion indicate a marked reduction in function of the xylem towards the calyx-end o f the fruit. This fits with the observation that calcium (mobile only in the xylem) is least concentrated at this end of the fruit. The observation supports the view that progressive xylem breakdown underlies and is the principal detenninant of fruit mineral status and distribution. Frequent sampling of fruit during the season has been used to evaluate internal tissue growth and the corresponding changes in the spatial arrangement of the primary bundles within the flesh. It was found that fruit grew mainly through an increase in the volume of Chapter 4. Causes and effects of changes in xylem functionality in apple fruit 97 the cortical tissue while the extent of difference in the layout o f the primary bundles may account for the different degrees of dysfunction between cultivars. In conclusion, the entire work contained in this chapter suggests that growth-induced strains damage the vessels and impact negatively on xylem inflow resulting in changes in mineral composition as the fruit matures. There is a need to further elucidate these processes in relation to fruit growth dynamics during the season. It is hoped that the results and discussion presented will contribute to a more comprehensive understanding of developmental changes in xylem function of the fruit. In this way, the apple industry might be able to minimise the impact of early vigorous growth in apples (i.e. through better cultivar selection and through appropriately adjusted crop load) so as to reduce the occurrence of calcium-related disorders. Chapter 5. Air and cell mapping of apple fruit 98 Chapter 5. Air and cell mapping of apple fruit 5.1 . INTRODUCTION An apple fruit is composed of many different tissues but the greatest volume comprises the flesh (Smith, 1 950). This is made of parenchyma permeated with vasculature and intercellular spaces (Esau, 1 977). The fractional air volume increases during fruit growth so as to occupy a considerable proportion of the flesh volume by harvest (Bain and Robertson, 1 95 1 ; Skene, 1 966; Westwood et al., 1 967; Soudain and Phan Phuc, 1 979; Goffinet and Maloney, 1 987; Harker and Ferguson, 1 98 8 ; Yamaki and Ino, 1 992; Ruess and StOsser, 1 993). The increase in air space during development is accompanied by a proportional decline in fruit density; the density of the fruit cells themselves remains roughly constant (Bain and Robertson, 1 95 1 ; Skene, 1 966; Westwood et al., 1 967). Furthermore, the volume of the intercellular air spaces continues to expand during storage (Soudain and Phan Phuc, 1 979; Hatfield and Knee, 1 98 8 ; Harker and Hallett, 1 992; Ruess and Stosser, 1 993). The fraction of intercellular air in the flesh differs between the cultivars (Reeve, 1 95 3 ; Baumann and Henze, 1 98 3 ; Vincent, 1 989; Ruess and StOsser, 1 993), while larger fruit of the same cultivar have a higher proportion of intercellular air than smaller fruit (Reeve, 1 95 3 ;_Ruess and StOsser, 1 993; Volz et al., 2002). However, the significance of air volume on fruit size is a long-standing matter of controversy of whether the late fruit growth was due mainly to an increase in the size of intercellular spaces (Tedey, 1 930; MacArthur and Wetmore, 1 94 1 ; Reeve, 1 953) or not (Skene, 1 966; Goffinet et al., 1 995). Intercellular air volume is also an important factor in understanding the postharvest quality of apple fruit. Fruit with greater fractional air volumes have been shown to be softer (Hatfield and Knee, 1 988; Yearsley et al. , 1 997a, 1 997b; Volz et al., 2002), or more mealy (Harker and Hallett, 1 992) and to have greater internal gas diffusion rates (Rajapakse et al., 1 990). Chapter 5. A ir and cell mapping of apple fruit 99 A range of techniques has been developed to measure the intercellular air in apple fruit. One is based on weighing the amount of water that replaces an equal amount of air removed by a vacuum from the submerged tissue. A number of studies have used this principle with a range of sampling methods, vacuum pressures, treatment durations and infiltration media (Reeve, 1 95 3 ; Harker and Ferguson, 1 988; Vincent, 1 989; Raj apakse et al., 1 990; Yamaki and Ino, 1 992; Yearsley et ai. , 1 996). In contrast, Calbo and Sornrner ( 1 987) estimated intercellular air volume using a gasometric technique that measured the amount of air extracted from samples when subj ected to vacuum. An alternative way to calculate air fraction has been to assess the weight and volume of a whole fruit when submerged in water (Bain and Robertson, 1 95 1 ; Skene, 1 966; Westwood et al., 1 967; Baumann and Henze, 1 98 3 ; Hatfield and Knee, 1 98 8 ; Harker and Hallett, 1 992). Other authors have taken a stereo logical approach to estimate air volume in the tissue (Soudain and Phan Phuc, 1 979; Ruess and StOsser, 1 993; Goffinet et al., 1 995). The intent o f this study was to develop methods through which to measure internal air in fruit flesh and to elucidate tissue structure at the cellular level. ' Braeburn' was used for this study as it is known to have a relatively dense and firm flesh (Dadzie, 1 992; Yearsley et al., 1 997a; 1 997b), poor flesh gas-diffusivity (Rajapakse et al., 1 990) and low skin gas-permeance (Rajapakse et al. , 1 990; Dadzie, 1 992). All these features could lead to increased susceptibility to tissue browning and damage. 5.2. MATERIALS AND METHODS 5.2.1 . Plant material The study used mature trees of 'Braeburn' apples grown at the Massey University Fruit Crops Unit, Palmerston North, New Zealand. Trees were planted on MM 1 06 rootstock and managed according to standard commercial practice. Fruit were picked at late harvest and selected to be without obvious blemishes. To minimise variation, fruit of similar size were chosen (c. 200-250 g). They were stored in a cool-store (O.soC, c. 95% RH) for up to S months pending analysis. Day-to-day storage was in a cold room (2°C, no humidity control). Chapter 5. Air and cell mapping of applefruU 1 00 5.2.2. Air volume measurement The intercellular air volume of freshly excised blocks of apple flesh was measured using methods based upon Archimedes' principle. The technique employs two experimental arrangements involving I mg sensitivity electronic balances. The first arrangement measures sample volume by recording the buoyant upthrust experienced by a sample when it is fully immersed in water (Fig. 5 . 1 A). To do this, a small block of flesh was impaled on the tip of a fine (0.24 mm diameter) steel entomological pin and this attached to a magnet mounted on a light stand that rested on the bench. The use of a pair of plastic (non magnetic) forceps simplified sample handling. The components were so positioned that when attached to the magnet the sample was held just beneath the surface of water that was contained in a small (50 cm3) plastic container that rested on the pan of a pre-tared electronic balance. The buoyant upthrust experienced by the sample appears to the balance as an equal and opposite downthrust and is expressed as an apparent increase in the weight of the water in the container. Because the density of water at 20°C is 0.99823 gcm-3, the balance reading (g) can be directly interpreted as a sample volume ( V) (cm\ With a 1 mg sensitivity electronic balance the system yields sample-volume estimates to the nearest 1 J.!l. A wetting agent (0. 1 % Silwet L-77) was added to the water to help eliminate entrainment of air bubbles as the sample was immersed. The force of the meniscus acting upon the pin as it broke the surface (about 1 mg) was minimised both by the use of a fine pin and by the wetting agent. This surface tension induced error was not only small but it was also systematic and appeared equally in all measurements so tended to cancel. A second balance arrangement was used to measure sample air volume. This involves a subtle variation of the first (Fig. 5 . 1 B). Here, the small container instead rested on a light 'bridge' mounted on the body of a similar pre-tared balance in such a way that the container was supported just above the pan. A similar magnet was mounted on a wire support and a light stand and this assembly rested on the balance pan. B y attaching the pin holding the flesh sample to the magnet, the sample was again held j ust beneath the water. Using this arrangement, the balance recorded the weight of the sample while under water. Prior to air extraction the recorded weight was usually negative because apple flesh floats. By weighing the sanlple just before (W1) and just after (W2) the removal of Chapter 5. Air and cell mapping of apple fruit 1 0 1 tissue air under vacuum i t is possible to determine the volume of air extracted as an apparent increase in sample weight (WA = W2 - Wl). Again, because the density of water is very close to unity, the difference in balance readings (g) can be interpreted directly as the volume of air removed by vacuum extraction (cm\ A 1 mg sensitivity balance measures air volumes to the nearest 1 Ill. Air extraction (after the first weighing and before the second) was accomplished by transferring the sample still with its pin to a 50 cm3 plastic vial of water containing 0. 1 % Silwet L-77. A small sinker weight made of a spiral of wire solder was threaded onto the pin to hold the sample beneath the water. While under water in this vial, the sample was exposed for 60 s to reduced atmospheric pressure causing the intercellular air to expand and to bubble out through the cut tissue surfaces. The vacuum generated by a vacuum pump in the presence of water at 20 QC is approximately 1 8 mm Hg (the vapour pressure of water at this temperature). This causes the intercellular air, initially at the atmospheric pressure (760 mm Hg), to expand its initial volume so that most of it bubbles out. At this low pressure a small residue of air remains within the tissue (about 1 8 x 1 001760 or 2.3% of the initial amount) but the vast maj ority (i.e. 97.6%) is removed. This small error is systematic and can either be ignored or allowed for in calculation i f the water temperature is approximately known. Following release of the vacuum, the sample was held under the water for a further 60 s. This caused the residue of internal air to contract and the internal spaces (previously air-filled) to became occupied with water (now at atmospheric pressure). The fractional air content (A) of the sample (%) is, therefore, given as the difference between the two under-water weights expressed as a percentage of the sample volume A = (W", I V) 1 00 ( 1 ) Chapter 5. Air and cell mapping of apple fruit A. Measuring sample volume by di splacement B. Measuring sample wei ght under water 1 02 Fig. 5 . 1 . Experimental arrangements for measuring fractional air volume. (A) Sample volume was first recorded by measuring the upthrust on the sample when fully immersed (equal to the weight of displaced water and expressed as an equal and opposite downthrust on the balance). (B) Sample air content was measured by weighing the same sample under water before and after vacuum air extraction. 5.2.3. Timings A preliminary study showed that 60 s of air extraction (under vacuum) followed by 60 s of water replacement (at atmospheric pressure) was sufficient for most of the air in a 3 x 1 2 x 1 2 mm tissue block to be extracted and replaced by water. It was desirable, however, to determine optimal timings so some further study was required. For this, equatorial discs ( 1 5 -20 mm thick) were taken from several fruit. For excision of flesh samples, we used Leica Model 8 1 8 disposable microtome blades to minimise tissue damage close to the cut surfaces. The discs were cut so as to produce pairs of blocks 1 2x 12xr mm with their long axes (r) aligned with the fruit radius. One of each block pair was used as control (with fixed air-extraction and water-replacement timings) while the other was exposed to a range of timings. This plan allowed data to be normalised (i.e. to remove errors due to fruit:fruit and radial variability) by taking the ratio of air volume estimates for a pair of matched blocks (treatment/control). To determine the optimum time for water replacement, the control blocks were exposed to 120 s of vacuum extraction and to 480 s of water replacement. Both these timings Chapter 5. A ir and cell mapping of apple fruit 1 03 were judged to be longer than the times required for completion. To determine the optimal water-replacement timing, extraction times for the treatment blocks were fixed at 1 20 s (like the controls) while their changing weights (as they took up water) were recorded every 30 s until a steady state was achieved « 480 s). Five replicate pairs of blocks of three sizes (r :::::: 3 , 6 and 1 2 mm) were used to assess the influence of sample size on water-replacement times. Once the optimum water-replacement time had been found (300 s for a 1 2x 1 2x 1 2 mm block), five replicate pairs of blocks were used to determine the optimal vacuum extraction time. Tested times were 5 , 1 5 , 30, 60, 1 20, 240 and 480 s. The control blocks were all extracted for 480 s. 5.2.4. Spatial distribution of intercellular spaces Fruit were cut at the equator to excise transverse discs about 1 5 - 20 mm thick. Radial transects ( 1 2x 1 2xr mm) were cut from these discs. Transects extended for the full distance (c. 30 mm) from the skin to the fruit axis and included the petal bundle located between a pair of locules (Fig. 5 .2A). To facilitate vacuum infiltration, the skin and the cartilaginous walls of the locules were shaved off very thinly. This radial transect was then sliced (blade aligned parallel to the skin) so as to create from it ten == 3 mm thick blocks. Even spacing of cuts was facilitated by making these cuts while the block rested atop a 1 mm grid. Each block was measured carefully and each series of 1 0 blocks provided the samples for measuring air volume change along a fruit radius. Tangential variation in air distribution around the fruit was examined using a single fruit with five symmetrically excised transects. This provided 5 0 blocks (five replicates for each tissue depth). To investigate the radial disposition of tissue air we tested ten fruit. One transect was taken from each fruit giving 1 00 blocks (ten replicates for each tissue depth). The volumes of intercellular air were combined across replicates for each tissue depth so as to yield a mean fractional air content (%) along the radii . Each transect contained eight rectangular 3 x 1 2x 1 2 mm blocks (BJ to Bs) plus two trapezoidal blocks close to the fruit axis (B9 and BJO)' These blocks were also 3 mm thick but their shape was dictated by the proximity of the locules (Fig. S.2A). Chapter 5. A ir and cell mapping of apple fruit A. Sample for ai r vol ume measurement 12mm 12mm B. Sample for cel l measurement 1 04 Fig. 5 .2 . Tissue sampling for air volume and cell measurements. A flesh disc about 1 5 - 20 mm thick was cut from the fruit equator. (A) Each of ten transects ( 1 2x 1 2xr mm) was subsequently cut into ten blocks for air volume measurement (note the trapezoidal shape of the two inner blocks, B9 and B I O). (B) Five pairs of thin transects were also taken from one fruit and stained to reveal the walls of cells just beneath the surface (entire cells were visible in 2-dimensional view). The two orthogonal diameters of 400 well-defined cells were randomly measured from two fields in the middle of the cortex for each transect. 5.2.5. Cell measurements 5.2.5.1. Staining. Fruit were cut transversely to form five pairs of flat equatorial blocks (Fig. 5 .2B) and these rinsed briefly with water to remove cell-content residues from the cut surfaces. For ferric tannate staining (Goffinet et al., 1 995), the transects were infused for three minutes in 5% aqueous tannic acid (this infiltrates the cell walls), then briefly rinsed with water to prevent extra-cellular depositions of precipitate. Next, the tissue was flooded with 1 % aqueous iron alum (ferric ammonium sulphate) for 60 s. This reacts with tannic acid to form a black tannic-iron precipitate within the cell wall matrix (Austin, pers. comm.). Lastly, blocks were rinsed to remove surplus precipitate and analysed immediately to minimise possible dimensional change due to osmotic water uptake. Chapter 5. Air and cell mapping of apple fruit 1 05 5.2.5.2. Cell dimeltsiolts. Cell dimensions were measured using a video-microscope system (a Wild M3Z dissecting microscope coupled to a JVC TK- 1 280E video camera and this connected to a PC running Video Master™ software). Transects were mounted in a petri dish under water to create an optically flat meniscus that minimised light scattering from the irregular block surface. To stop transects from moving they were lightly impaled on a thin spike attached to the bottom of the dish. This arrangement presented a sharp image of whole cortical cells from just beneath the cut surface onto a 1 7" flat VDU screen at x 1 95 magnification. The software allowed measurement, using the mouse, of a live (unfrozen) image and direct data acquisition to disc. This greatly facil itated decision-making because it was possible to adjust the focus slightly as each measurement was being made. For each block, cell dimensions (2 orthogonal diameters) were obtained for forty cells from two adj acent fields in the middle of the cortex approximately lO mm in from the skin (Fig. 5 .2B). The first diameter measurement (dl) was always made in line with the long axis of the cell with the second measurement (d2) at right angles to this. For later calculations, the geometrical mean o f these diameters (D) was calculated as D == (d] d2 )0.5 (2) 5.2.6. Adjustment for air loss When a block of apple flesh is immersed in water there is a tendency for water to invade the air spaces near the cut surface. Air displaced in this manner at the initial weighing ( WJ ) is not recorded by this method. This air loss causes, therefore, underestimation of air content. The magnitude of the error will depend on the surface area to volume ratio of the sample (and thus on its size and shape). This loss can be quantified, and corrected for, by considering it as equal to a loss of air in a superficial layer of thickness (1) all around the sample. The value taken by T will likely be similar in the tangential directions (t) relating to the 2 transverse and 2 radial-longitudinal faces of a tissue block but may be greater in the radial direction (tx) relating to the 2 tangential block faces. This is because apple parenchyma is highly anisotropic (Vincent, 1 989), with a network of elongated air spaces radiating out from the centre (Reeve, 1 95 3 ; Khan and Vincent, 1 990). - - ---------------- Chapter 5. A ir and cell mapping of apple fruit 1 06 It is unlikely that on sample immersion, water will invade an air space where the cut aperture to this is smaller than its maximum diameter. For this to happen would require an increase in the area of invading meniscus and thus the input of energy (e.g. from a vacuum pump). It is more likely that only those cut air spaces whose diameters decrease with depth will be invaded. This insight leads to the hypothesis that in the two tangential directions, t will take a value equal to about Y4 of the average cell diameter. Recall that, and with random sectioning, the air spaces between close-packed spheroidal cells will be so cut that either: the aperture falls before the space ' s mid-point (in this case no air is lost), or it falls after the mid-point (in this case water enters freely and all air is displaced). Air spaces, having similar maximum dimensions to the spheriodal cells that form their boundaries, and these two scenarios occurring with the similar likelihood, it is likely that air loss will occur to a depth of a Y4 cell diameter. The conspicuous radial air channels o f some cultivars may offer freer access to water invasion in the radial direction. In this case the effective thickness from which air is lost (Ix) may be greater in this direction. To check these predictions, pairs of 1 2x 1 2x 1 2 mm blocks of cortical tissue (cut from 4 1 6 mm below the skin), were taken from a number o f apples. The air volume of one block was measured intact while the air volume of the other was measured (by summing) after it had been cut up into a number of slices. Sixteen such pairs of blocks were measured. In six cases the sliced block was cut into two 6 mm slices, in four into three 4 mm slices, in three into four 3 mm slices, in two into six 2 mm slices, and in one into twelve 1 mm slices. For each pair, the sum of the air volumes in the slices (.EV) was compared with the value for the paired intact block ( V) where, because of the block ' edge' effect, V >IV. A model was then used to evaluate a correction factor (F) that takes as input the shape, size and orientation of the block where Vcorrected = V measured F (3) The error is expected to be inversely related to the size of the sample. The model was fitted by least squares using PROC NUN (SAS, 2000). Chapter 5. A ir and cell mapping of apple fruit 1 07 5.3. RESULTS 5.3.1. Timings The Figure 5 . 3 shows a plot of the relative air volume evaluated for blocks of different sizes and times. Data were normalised by dividing the air content value obtained for the matching block pair that had been exposed for times judged to be excessive (viz 1 20 s for extraction and 480 s for water replacement). This creates the property that results in an asymptote to unity. The largest blocks ( 1 2x 1 2x 1 2 mm) required 300 s for water replacement and this was longer than for the smaller ones. Thus 6x 1 2x 1 2 mm blocks required 1 20 s and 3 x 1 2x 1 2 mm blocks required just 3 0 s (see Fig. 5 . 3A). Air extraction, however, even from the largest ( 1 2x 1 2x 1 2 mm) blocks was completed within about 60 s (see Fig. 5 .3B). 1 .2 1 1 1 1 1 1 1 1 .2 1 I I 1 I I I I A. B . Steady state Steady state 1 . 1 " �\ 1 1 / u 1 .0 " 4�- 1 .0 " I'-"�- I rr-... ·· � . .... et! 0.85 '.;::l ctI -� 0.80 0.75 I I I I I I I I I I / / / / / 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 B lock thickness (mm) Fig. 5 . 5 . Average scores for relative air volume in blocks of different radial dimensions. Each mean value (e) represents nonnalised measurements for each block size. The lines were fitted using T value fixed (dashed line) and T value varying with block thickness (solid line). Chapter 5. A ir and cell mapping of apple fruit 1 1 1 Table 5 . 1 . Air loss in the radial direction (Ix) due to water invasion calculated using Eq. 4. An effective surface layer thickness from which air is lost due to water invasion is dependent on the dimension of the block (in the radial direction). Block thickness (x) Effective surface layer (mm) thickness (Ix) (mm) 1 0. 1 23 2 0. 1 94 3 0.235 4 0.259 6 0.2 8 1 The blocks bordering the locules (B9 and B I O) are more difficult as they are 'trapezoidal' in shape. For simplicity, a similar approach can be applied, taking off a thickness tx in the radial direction and t in the tangential direction (Fig. 5 .6). Given the minimum (a) and maximum (b) widths of the trapezoid and the height (x), then assuming symmetry the correction factor becomes F _ [ t(a + b) 1 2 x ] trapezoid - Ha + b) - 2t[sin(B)r1 1 2 - 2t x - 2tx 2x where tan(B) = -- . b - a (8) The adjustment factors for the rectangular block samples can be calculated as 1 .2 1 (B I - Bg), and for two trapezoids as 1 .22 (B9) and 1 .23 (B I O) . Although these factors are not very different, it is clear that air loss will be proportionately higher in the smaller blocks. Chapter 5. A ir and cell mapping of apple fruit a I 2 (a + h) - 2 . e sm h I - x 2 1 1 2 Fig. 5 .6. The calculated adjustment for a 'trapezoidal ' sample (B9 and B l O). The shaded area represents the effective surface layer thickness from which air is lost due to water invasion. The adjustment for air loss is calculated from the difference in the volumes of the inner and outer trapezoids based on average dimensions of a, h and x measured in a number of samples. 5.3.4. Spatial distribution of intercellular spaces When air distribution was measured tangentially around the fruit, no significant asymmetry was found (p = 0. 545). This removed the requirement to adopt any special criterion when deciding along which radius to sample. The results show clearly that fractional air content reduces strongly as one moves from the skin towards the fruit axis (p ::;; 0.00 1 ). With the data now adjusted for the small air losses suffered on water immersion, a more accurate picture of radial air distribution in the fruit emerges (Fig. 5 .7). The air content change follows a distinct pattern in the cortex and this pattern is recapitulated in the pith - these two concentric fleshy tissues originate from different parts of the flower (Esau, 1 977). In the cortex, the air fraction first increases from a subepidermal level of about 1 8%, to reach a peak of about 23 % in the layer just beneath. It then falls steadily to about 1 2% at the core line. Here, at the boundary between cortex and pith, a sharp discontinuity occurs with the air fraction rising suddenly to about 1 5 % before falling steadily again to about 7% at the centre of the fruit. Chapter 5. Air and cell mapping of apple fruit 1 1 3 30 Skin Core line Centre 25 J • I J J ,--, � '-' I�' 0 I § 20 • • "0 .�--� ;;.- • I-; 1 5 . .... <.:I C; § 1 0 ..... ..... () � � 5 • 0 0-3 3-6 6-9 9-1 2 1 2-1 5 1 5- 1 8 1 8-21 2 1 -24 24-27 27-30 Radial tissue depth (mm) Fig. 5 .7 . Changes in fractional air content along a radius in 'Braebum' fruit. Ten tissue blocks (B J B I O) were measured sequentially between the fruit surface (0-3 mm) and the centre (27-30 mm). Mean values (0) are calculated from 1 0 measured values (e). 5.4. DISCUSSION 5.4.1 . Timings The magnitude of convective fluid flow through the air spaces of fruit flesh tissue will depend on the viscosity of the fluid, the cross-sectional area of the channels and the length and tortuousity of the path (the last three properties being dependant on cell and tissue structure and on the size of the sample). The air measurement method described here involves the removal of air from the flesh sample and its replacement by water. It is known that the flow of a fluid is inversely proportional to its viscosity - the viscosity of air at 20°C being 1 . 8 1 3 x 1 0-5 Pa s and this is much lower than the value for water of 1 .002x 1 0-3 Pa s (Nobel, 1 999). Hence, air will experience much less resistance to its outflow than will water to its inflow, which is Chapter 5. Air and cell mapping of apple fruit reflected in the distinct differences in time required for air extraction and for water replacement (compare Figs. 5 .3A and 5.3B). The same effect is also evident from Vincent ( 1 989) who observed a markedly slower fluid infiltration. This can be also attributed to the fact that he used 5% aqueous mannitol, which is more viscous than water. 1 1 4 Moreover, an increase in the size of the sample will have an influence on the time taken for each of these flows to run to completion. For example, every increase of block size will increase the volume of air that must flow out (or of water that must flow in) while it will reduce the steepness of the pressure gradients driving these flows. Again this effect is reflected in the equilibration times of Fig. 5 .3A with larger blocks requiring longer timings. 5.4.2. Cell measurements 5.4.2.1. Staining. The staining method proved simple and well suited for the staining of mature apple parenchyma containing large cells. Entry into the cell walls was fast and enabled a relatively good definition of individual cells. The tannic-iron also penetrated deeper into the tissue and stained some cell walls beneath the surface layer (Fig. 5 .4). The contrast improved if stained samples were left in a refrigerator for 1 - 2 days. 5.4.2.2. Cell dimensions. This study has found that the mean cell diameter of ' Braeburn' (c. 250 J!m in the mid-cortex) is in the close agreement with values reported by Tukey and Young ( 1 943), Bain and Robertson ( 1 95 1 ), Khan and Vincent ( 1 990) and Harker et al. ( 1 997), but slightly higher than those of Reeve ( 1 953). From this information, it is possible to arrive at a reasonably good estimate of the number of cortical cells in an apple fruit. Cell number can be calculated using the mean cell diameter value to arrive at a value for mean cortical cell volume (using 4/3n:r\ Cortical cell number is then calculated as the ratio of this volume to the fruit's cell-fraction volume (i.e. the total volume of the fruit minus the volume of the contained air). Cell number estimates arrived at in this manner are especially useful where it is desired to record changes in cortical cell number in response to some treatment (e.g. to a range of imposed crop loads). Hence, the various Chapter 5. A ir and cell mapping of apple fruit 1 1 5 systematic errors contained in the underlying assumptions (e.g. the contribution to fruit weight made by the skin or core tissues has not been allowed for) will tend to cancel. 5.4.2.3. Cell shape and cell packing. Information on cell shape and cell packing provide an important tool for understanding the textural and physiological properties of fruit flesh. Cells taken from the mid-cortex of 'Braeburn ' are prolate spheroids having an aspect ratio close to 1 .5 (their long diameter is about half as great as their short diameter). Furthermore, the flesh cells gradually increase in length and also become more aligned towards the interior of the fruit (e.g. compare Figs. 5 .4A and 5 .4B) . This conforms to the pattern seen in a number of other apple cultivars (Bain and Robertson, 1 95 1 ; Reeve, 1 95 3 ; Khan and Vincent, 1 990; Ruess and StOsser, 1 993). It would seem that changes in cell shape and packing cannot be explained simply on the basis of mechanical factors (e.g. forces) acting on the cells - the physical environment in which the cells develop is unlikely to be all that different in the different regions within the fruit. Instead, other factors (e.g. gradients in hormone concentration along which the cells align their long axes) could be involved. These changes result in the numerous air spaces between the cells that became increasingly organised deeper into the flesh so as to form distinct radial air channels closer to the centre of the fruit (Reeve, 1 95 3 ; Vincent, 1 989; Khan and Vincent, 1 990). The channels have little lateral contact with one another and communicate effectively only in the inner parenchyma (Vincent, 1 989). Their presence strongly contributes to the anisotropy and heterogeneity of apple flesh tissue (Vincent, 1 989; Khan and Vincent, 1 990). For example, when compressed radially, the tissue tends to fail in sympathy with the air channels (Khan and Vincent, 1 993). Since flesh textural attributes are very important determinants of fruit quality and the different apple cultivars vary widely in texture (Janick et al., 1 996), it is likely that differences in the size, orientation and distribution of the air channels will make an important contribution to the perceived textural properties of the fruit. It is also likely that the flesh of an apple is under pressure while the skin is in tension (Vincent, 1 992) and that these opposing forces operate to form the structural unit of the whole. The relatively large proportion of air and the relatively rounded shape of the flesh Chapter 5. A ir and cell mapping of apple fruit 1 1 6 cells suggest, however, that this 'tissue pressure' (the force imposed by the tension in the skin that tends to press the cells against one another) is quite gentle. The degree of faceting in the fully turgid tissue will, therefore, depend upon the balance between cell turgor and tissue pressure. Given a certain tissue pressure, a higher cell turgor will tend to increase tissue air volume, and create more rounded cells with smaller facet areas. Higher internal air volumes will also lead to better internal gas exchange. 5.4.3. Adjustment for air loss An air:water meniscus is in tension and thus will always tend to minimise its area in order to minimise surface energy. This means that when the cut surface of a block of apple flesh tissue is immersed in water, water will tend to invade any air-filled cavities in the irregular surface but only where to do so involves a reduction in the meniscus area. Correspondingly, water will tend not to displace air from cavities whose shape is such that where to do so the meniscus area is required to increase. In practice, only air spaces lying close to the surface are invaded whereas the maj ority of the air is contained deep within the tissue and thus is measured. For water to invade the deeper compartments, energy must be input (in this study it was contributed by a vacuum pump). It can be predicted from a simple geometrical consideration that, before the vacuum is applied, air spaces lying between regularly packed spheroidal cells will, in general, not be invaded beyond a depth of about a Y.J cell diameter. However, the radial air channels should allow water to penetrate deeper in this direction. If this increased depth of penetration approaches half the sample's dimension (in the radial direction) there is the possibility that some of the internal air will be displaced even from the deepest regions of the block. The model proposed, therefore, accounts not only for all the air displaced around the edges of the block but also from air spaces beneath the cut surfaces that are open to water invasion in the radial direction. Thus for the range of samples ( 1 to 6 mm in the radial direction), tx is observed to change gradually with slice thickness (Table 5 . 1 ). In a 1 mm slice, air loss occurs from around 25% of the radial thickness, whereas for a 6 mm slice, loss occurs from less than 1 0% of the radial thickness. Not surprisingly, therefore, measurement error is inversely related to block size such that 1 mm thick slices account Chapter 5. Air and cell mapping of apple fruit 1 1 7 for only 0.79 of the total air volume, while for 6 mm slices the figure is 0.95 (Fig. 5 . 5 .). Consistent with the degree of spatial resolution required, when using this technique for air content measurement, block sizes should be maximised in order to minimise the impact of air loss on the result. 5.4.4. Spatial distribution of intercellular spaces In 'Braeburn' , a progressive decrease in the volume of intercellular air occurs along a radius (Fig. 5 .7). The decrease is not uniform though, and the transect contains two peaks - one in the cortex and another in the pith with the highest fractional air volumes being recorded j ust inside the bounding tissues (viz the skin and the core line). Other authors have investigated air space distribution in apple. Reeve ( 1 95 3 ) and Vincent ( 1 989) reported a gradient in the volume of the intercellular spaces along a radius. Their results were obtained by vacuum infiltration of rather large samples that gave a result that referred to the whole of the outer or inner flesh. Soudain and Phan Phuc ( 1 979) and Ruess and Stasser ( 1 993) also observed a general decline in the intercellular air content in a radial direction although the latter obtained a distinct bell-shaped air distribution. Their results were based on stereological estimates from 2-dimensional sections. In this study, a more direct measurement method has been employed, using a contiguous sampling of still-turgid tissues with adjustment for air loss around the sample edges. The pattern of change in air content, with its sharp discontinuity at the core line (Fig. 5 . 7), is mirrored by changes in both tissue organisation and cell packing. This leads to the idea that the fruit can usefully be considered as comprising two fleshy parts in line with the traditional botanical perspective (Esau, 1 977). The cells at the core line are distinctively smaller, are more elongated and have thicker walls than the other parenchyma cells (MacDaniels, 1 940). This phenomenon is expressed to such a degree that the core-line cells form a visible ' sheet' between the outer cortex and the inner pith (Fig. 1 .2). Furthermore, some apples possess a ' cleavage plane' at the core line associated with a sharp increase in intercellular spaces (MacDaniels, 1 940). Various degrees of demarcation between the two fleshy tissues may have functional implications in terms of the physiology of gas transfer. Chapter 5. Air and cell mapping of apple fruit 1 1 8 A question here is whether these factors could impact on the storage quality in 'Braeburn' . Several lines of evidence support this idea. Steep gradients in the concentration of gases were found between the centre and the skin in ' Braebum' (Rajapakse et al., 1 990; Dadzie, 1 992) suggesting a heterogeneity of internal atmospheres and thus a degree of gaseous compartmentation. Gas movement will experience a greater resistance to diffusion in a tight-packed tissue, with low air content such as the fruit of 'Braeburn' (Rajapakse et al., 1 990; Yearsley et al., 1 997a; 1 997b). Whether gas diffusion occurs mainly through the air channels (Solomos, 1 987) and/or in a combination with fluid/solid matrix of the flesh (Rajapakse et al., 1 990), a region of reduced intercellular air content will act to impede gaseous diffusion. A partial diffusion barrier at the core line could further restrict gas exchange and lead to increased C02 and lower O2 in the inner parenchyma. The internal concentration of respiratory gases is also dependent on the permeance of the skin. The skin acts as the primary barrier to gas diffusion in apple fruit having a 1 0 to 20- fold higher resistance than the flesh (Solomos, 1 987). ' Braeburn' is known to exhibit high skin resistance (Rajapakse et al., 1 990; Dadzie, 1 992), high internal partial pressure of C02 (Dadzie, 1 992; Yearsley et al., 1 997b) and low respiration rates (Rajapakse et ai., 1 990; Yearsley et al., 1 997a) relative to other cultivars. Hence, when the skin is intact and the tissue is not water soaked, diffusional gas flow seems to be dominated by the resistance of the skin. However, if the tissues suffer a partial loss of membrane functionality such that the air spaces become flooded with water, capillarity dictates that this extracellular fluid will migrate towards the tissue regions containing the smallest pores. These lie around the core line and near the centre of the fruit (see Fig. 5 . 7). Flooding of air spaces will effectively seal off the inner tissues to respiratory gas exchange (diffusion in a liquid phase is 1 04 times slower than that in a gas phase, Nobel, 1 999). In ' Braebum' this would cause a sharp rise in the internal partial pressure of CO2 and a lowering in that of O2, which could lead to tissue breakdown. Chapter 5. A ir and cell mapping of apple fruit 1 1 9 A number of studies refer to CO2-related disorders in 'Braeburn' although the mechanism by which CO2 causes damage remains unclear (Watkins et al. , 1 997; Elgar et al., 1 998). Internal injury in ' Braeburn' is seen as randomly distributed discrete lesions that enlarge and merge in storage (Clark and Burmeister, 1 999). It was also shown that severe watercore may lead to cavity development and mealiness after visible symptoms have disappeared (Clark and Richardson, 1 999). These observations are consistent with a model in which the tight-packed internal structure of ' Braeburn' (particularly at the core line) reduces 02 and C02 diffusion resulting in a special susceptibility to internal injury. It is reported that a number of pre­ harvest factors contribute to physiological changes in the fruit that predispose it to injury (Ferguson et al., 1 999). It is suggested that the link may in fact be mediated through patterns of flesh structure (viz. air-space distribution) that are modified by factors in the pre-harvest environment. 5.5. CONCLUSIONS The study was undertaken to develop methods through which to determine patterns of intercellular air content and of cell shape, size, number, and packing within the flesh tissue as a whole. A gravimetric technique based on Archimedes ' principle was developed and refined through which accurate assessments of air content can be made in plant tissues. In ' Braeburn' , the intercellular air content of the flesh shows a systematic decline between the fruit surface and the centre with a marked discontinuity at the core line. This discontinuity effectively separates the fruit into two concentric parts. The radial decline in air content is accompanied by an increase in the degree of cellular organisation towards the centre. Quite steep gradients in air content underline the importance o f location of flesh samples for air content assessment. They also render relatively meaningless the assumption that an air content measurement made at one point can be taken to refer to the whole fruit. Chapter 6. General discussion and conclusions 1 20 Chapter 6. General discussion and conclusions The preceding parts of this thesis contain a number of experimental studies that explore various aspects of developmental fruit physiology. Research carried out included studies on the dynamics of initiation and functionality of fruit and stalk xylem, seed effects on fruit shape and fruit minerals and air and cell alignments in mature fruit. The work describes and illustrates the relationships between structure and function of the tissues with special emphasis on the mechanisms underlying physiological responses of the fruit. The obj ective of this chapter is to provide a final discussion from the results of the experimental studies and reviewed literature conceptualised in the model that describes the interactions between the elements that contribute to the development of fruit quality (Fig. 1 .3) and thus the management policies for the crop. 6.1 . GENERAL DISCUSSION 6.1 . 1 . Seed as a source of developmental stimuli The causative relationship between seed set and shape defects is well established (Heinicke, 1 9 1 7 ; Latimer, 1 93 1 , 1 937; Roberts, 1 946; Way, 1 978; Brault and de Oliveira, 1 995; Brookfield et al., 1 996) but lacks an intensive evaluation in regard to the morphology of the apple fruit. By sectioning the whole fruit on the basis of extrapolated carpellary partitioning (Fig. 2.2), fruit asymmetry, size and seed distribution were assessed with a three-dimensional perspective. The fruit tends to grow lopsided when seeds confined to a particular part fail to develop (Fig. 2 .4). The growth behaviour, therefore, appears to be modular where each carpel exerts a somewhat autonomous effect on fruit growth and, at least partially, develops as a unit. The dynamics of fusion between the carpels comprising an inferior ovary conform to this outcome. Delayed fusion along the carpellary margins (Esau, 1 977) indicate that, at least in the early stages of growth, carpels are physically separated and thus are likely to exhibit a physiological autonomy within the assemblage (fruit). ------ --------------- - Chapter 6. General discussion and conclusions 1 2 1 Several important implications emerge from these premises. Seed and fruit development are related, such that seed aberrations parallel that of fruit shape. Seed development appears to be a prerequisite for fruit development and a likely source of developmental stimuli that radiate towards the skin. Fruit development, therefore, occurs through the combined growth of the carpels that strongly influence the growth of the extracarpellary tissue in a centrifugal direction (first-order effect) and to a lesser extent laterally (second­ order effect), (Fig. 2 .6). Hence, the overall growth of the apple fruit appears to be determined by the dynamics of ovary growth, which in a biological sense represents the 'true' fruit o f a pome (Abercrombie et al., 1 98 1 ). The localised trend in fruit asymmetry was also related to the spatially confined mineral composition. The larger sectors of lopsided fruit tended to have lower mineral concentrations but only that of the calcium was significantly affected (Figs. 2 . 7 and 2.8). This is in line with the expected influence of an increased growth rate that could lead to a dilution o f the mineral concentration within the rapidly expanding tissues of the fruit (Ferguson and Watkins, 1 989). The ratio of calcium to magnesium and potassium in the larger sectors thus changed unfavourably with respect to calcium although no symptoms of bitter pit were noticed in the stored apples assessed in this study. This implies that better seed set will on average lower the calcium concentration by promoting tissue growth, which contradicts the claim that an increase in seed number will increase fruit calcium in the population as a whole (Tomala and Dilley, 1 989; 1 990; Bramlage et al., 1 990; Broom et al. , 1 998). If seeds can directly influence fruit calcium this effect should be independent of environmental conditions. In addition, more seeds should stimulate auxin export from the fruit (Bangerth et al., 1 989; Callej as and Bangerth, 1 997; Bangerth, 2000), which Banuelos et al. ( 1 987) considered promoted calcium import by the fruit. However, where seed set has been affected by pollination, the relationship with fruit calcium tended to be masked by tree effects (Brookfield et al. , 1 996; Volz et al., 1 996b). The relationship between seed set and the calcium concentration of the fruit is clearly complex and requires a better understanding of the impacts of a range of factors. It is common for an over-simplistic description to lead to a contradiction. However, it is Chapter 6. General discussion and conclusions possible to create a more sophisticated explanation of the ' seed-set: fruit calcium­ concentration' relationship (Fig. 6 . 1 ). I Seed set I � I Auxin I I I • • • � • Flesh Flesh Xylem Xylem cell No -0- cell size conduits No H ... conduits size Calcium ... mass - Fruit calcium - .. concentration .. Flesh ... volume 1 22 Fig. 6. 1 . Causative links between seed set and fruit calcium concentration (the mass of calcium in a particular volume of flesh). The sequence in Fig. 6 . 1 describes the dependence of fruit calcium concentration on the parallel developments of fruit vasculature and of flesh volume growth. Both pathways appear to be regulated by seed-derived auxin. However, both pathways will also likely be influenced by a host of ' external ' factors such as genotype, crop load, temperature, light, and water availability. Where the rate of a process depends on many factors, its sensitivity to change in any one factor will depend upon the extent to which this factor is in deficit. This concept of ' limiting factors' was first enunciated by Liebig ( 1 84 1 ) and Blackman ( 1 905) and is discussed in Salisbury and Ross ( 1 992). The concept of ' limiting factors' suggests that the relative strengths of the two auxin effects will depend upon the extent to which auxin is limiting for each pathway compared with the other potentially limiting factors. For example, auxin will have its dominant effect on fruit calcium concentration via calcium mass (the numerator, and governed by the right-hand pathway in Fig. 6 . 1 ) when the other factors affecting calcium mass are in Chapter 6. General discussion and conclusions 1 23 relative excess and auxin is limiting. Here more seeds (auxin) will increase calcium mass. If, instead, one of the other factors is in deficit (i.e. is more limiting than auxin), increasing seed number will not have much of an effect on calcium mass. The same arguments can be erected for flesh-volume (the denominator and left-hand pathway). Because the physiology of the two pathways are quite different, and are affected differently by the external factors, this simple model can explain many paradoxical results. Changing seed set could increase, decrease, or have no significant effect upon, fruit calcium concentration. 6.1 .2. Nature of the developmental stimulus The presence of seeds provides the necessary stimulus for fruit growth including mineral accumulation (Chapter 2). However, determining whether seeds act as source or sink of such stimuli will establish a key link in the model of fruit quality development (Fig. 1 .3). Exogenous applications of the auxin transport inhibitor (NPA) altered vessel differentiation in the stalk while controlling the extent of premature abscission, seed and fruit development (Chapter 3). This supports the hypothesis that auxin emanating from the fruit is the principal developmental stimulus. The timing and dynamics of vessel differentiation suggest a pattern of auxin efflux that is closely related to the stage of flower/fruit growth (Fig. 3 .5). In addition, the changes in vessel formation along the stalk (Fig. 3 . 7) and the adverse response to auxin inhibition in terms of seed and fruit size (Tables 3 .3 and 3 .4) indicate a possible negative feedback regulation of auxin biosynthesis in the fruit. Clearly, auxin efflux stimulates vascular differentiation in the stalk, which supports the increasing nutritional demands of the rapidly growing fruit tissues. Considering that the relative strength of the auxin efflux governs the dominance of the fruit (Gruber and Bangerth, 1 990), it seems likely that the rate of descending vessel differentiation will physically limit the rate of ascending sap flow. As a consequence, the relative strength of auxin efflux may not just determine the sink capacity of the fruit according to its dominance hierarchy but also promote the flow capacity of its xylem. Furthermore, the region of lowest conductance across the abscission zone (Fig. 3 .2) dominates convective Chapter 6. General discussion and conclusions flow in the vessels and creates the weakest point of fruit attachment to the tree that enables efficient fruit shedding. 1 24 6.1 .3. Character of fruit tissue growth as a determinant of fruit quality development Spatial variations in the mineral status of lopsided fruit (Figs. 2.7 and 2.8) highlight the link between mineral uptake and fruit growth. Dye infusion of the fruit revealed a progressive breakdown in xylem flow, more pronounced in the primary bundles, the principal suppliers of the fruit flesh (Figs. 4.5 and 4.6). A microscopic assessment showed that the mode of dysfunction is due to the physical damage of the xylem structure in the bundle (Fig. 4.9), similar to that observed in some other fruit crop species (Findlay et aI., 1987; DUring et al., 1987; Wolswinkel et al., 1 999). The suggestion that expansion of the fruit induces breakage of the vessels implies that fruit growth rates govern the patterns of change in the hydraulic properties of the fruit xylem and thus can affect fruit mineral composition. Several lines of evidence support this suggestion. Because the xylem tissue is the principal pathway for calcium transport (Marschner, 1 986) any changes in the functionality of the xylem will affect calcium accumulation in the fruit. Indeed, the xylem dysfunction in primary bundles (Fig. 4.5) can be positively correlated with the decline in fruit calcium import shown by Wilkinson ( 1 968) and Jones et al. ( 1 983). Furthermore, the sudden drop in fruit calcium concentration in apples coincides with the onset of rapid fruit growth (Jones et al., 1 983). Because the pattern of fruit growth is specific for a given cultivar (Pratt, 1 988), the dynamics of fruit xylem dysfunction appear cultivar-dependent. Significant differences in the pattern of xylem flow were observed between cultivars that vary in their susceptibility to the calcium-related disorder, bitter-pit (Fig. 4 . 5 ; Lang, 1 990; Lang and Ryan, 1 994). In addition, the decline in dye import towards the calyx end of the fruit (Fig. 4.7) again positively correlates with the more pronounced calcium deficiency and bitter pit frequency in the distal part of the fruit (Ferguson and Watkins, 1 989). Chapter 6. General discussion and conclusions 1 25 However, the ring of ten primary bundles appears to lie along the dorsal margins of the core (Fig. 1 .2). While fruit growth will cause the bundles to drift outwards, only the growth of the core can affect the primary bundles. Hence, cultivar differences in the dynamics of xylem dysfunction depend upon the growth of the carpellary tissues. Both examined cultivars ('Braeburn' and ' Granny Smith') showed similar tissue growth patterns (Figs. 4. 1 0 and 4. 1 1 ) but the relative increase in the distance between the locule tips and the aligned sepal bundles differed significantly (Fig. 4. 13). The proximity of the sepal bundles to the locules coincides with a prolonged functionality of these bundles during the later stages of fruit growth (Fig. 4. 14). This suggests a 'protective' role of the locules whose slower growth will slow the drift of a bundle in close proximity and thus prolong the functionality of its xylem. It has been shown in apple fruit that the direction of xylem movement reverses diurnally such that sap flow is from fruit to tree during the day and from tree to fruit at night (Lang, 1 990; Lang and Volz, 1 998). Recently, it has also been found that the concentration of osmotic materials (sugars, minerals etc) in the cell-wall free space rises as apple mature (Lang and Diack, unpublished). This rise is suspected to follow some level of dysfunction of the cell membranes (Lang and DUring, 1 99 1 ). It follows that the outflowing xylem sap will be rich in these osmotically active materials - the xylem sap being continuous with the cell-wall free space (apoplast). Hence, xylem dysfunction can be seen as minimising the loss of these materials from the fruit but at the expense of reduced import of xylem borne minerals to the fruit. Two consequences arise from these changes. First, phloem sap inflow to the fruit will be stimulated. The increased osmotic concentration of the fruit apoplast (Lang and Diack, unpublished) will reduce the turgor in the terminal sieve tubes strongly promoting phloem sap inflow to the fruit (Lang et al., 1 986, Lang and Thorpe, 1 989, Lang and During, 1 99 1 ) and will explain the observed increase in phloem import (Lang, 1 990). Second, the likelihood of calcium deficiency disorders will be increased. Phloem immobile minerals like calcium (Marschner, 1 986) will tend to be excluded from the fruit giving rise to relative deficiency and thus increased occurrence of physiological disorders like bitter pit. The view is, therefore, taken that dysfunction of the xylem and cell membrane discussed above is an integral part of normal fruit development. Chapter 6. General discussion and conclusions 1 26 6.1 .4. Tissue air and cell alignments and fruit storage quality It is likely that the textural properties of an aggregated tissue will arise from the character of its components such that specific changes in texture will be founded in the structure and individual properties of the elements of which it is composed. The dynamics of apple parenchyma development can create distinctly different textural properties in the fruit, some of which are peculiar to particular cultivars (Janick et al., 1 996). Anisotropy can also increase complexity by creating textural properties that vary with direction (Vincent, 1 989; Khan and Vincent, 1 993). Changes in cell shape and packing along the radius (Fig. 5 .4) conform to a general pattern of there being increased tissue organisation with increasing depth beneath the skin (Bain and Robertson, 1 95 1 ; Reeve, 1 95 3 ; Khan and Vincent, 1 990; Ruess and Stasser, 1 993). Also, the radial gradient in air content (with a sharp discontinuity at the core line) was measured with high accuracy and by contiguous sampling (Fig. 5 . 7) . These various patterns evidence the distinctive disposition of the intercellular air network in ' Braebum' . The nature o f tissue air distribution in 'Braebum' might b e related to its susceptibility to internal storage disorders. The tight-packed internal structure (Dadzie, 1 992; Yearsley et al., 1 997a; 1 997b) and potential diffusion barrier at the core line is likely to impact on gas transfer within the fruit. Furthermore, the appearance o f free water in the intercellular spaces (e.g. due to partial loss of membrane functionality) will further impede gas diffusion. These features are likely to reduce the ability of respiratory gases to move freely across the flesh. Together with a low skin permeance to gases (Rajapakse et al., 1 990; Dadzie, 1 992) these flesh properties will determine the storage behaviour of 'Braeburn' . Chapter 6. General discussion and conclusions 1 27 6.2. GENERAL CONCLUSIONS In conclusion, some aspects that ultimately affect quality development in apple fruit have been studied and the causative relationships between structure and function for some of its components assessed. The literature clearly points out the reasons why some fruit fail to meet particular quality criteria but often provides no more than an overly simplified relationship between a likely cause and its effect on quality. The conceptual model (Fig. 1 .3 ) is based on the premise that the seed represents the primary reproductive organ and is the principal source of developmental stimuli for the surrounding fruit tissues. It is apparent from the model that fruit quality development is controlled by a complex of synchronised physiological responses interacting within and across the paths of the major quality attributes. Hence, the assessment of physiological properties can frequently include properties, processes or factors from other developmental paths. This makes the model of quality development in apple fruit generally applicable to other fruit crop species. In summary, the studies reported in this thesis provide an assessment of the mechanisms that control some aspects of quality development in the apple fruit. These are: • That the modular growth of the syncarpous ovary affects shape and spatial mineral distribution in the fruit; • That auxin emanating from the fruit controls vessel differentiation, abscission, seed and fruit development; • That it is the structure and character of tissue growth that controls the progressive xylem dysfunction in the fruit; and • That it is the structure of the flesh that determines the nature of air distribution and thus the storage behaviour of the fruit. Hence, identifying the mechanisms underlying apple fruit development has been valuable in conceptualising the mechanistic links between structure, function and quality. The implications of these findings in a wider developmental context are considerable. Chapter 6. General discussion and conclusions 1 28 From a biological standpoint, the fruit is the structure produced to protect and disperse seeds and thus ensure the propagation of the species. Hence, the role that fruit play in plant ontogeny determines the nature of its development. Some defects of the fruit that we define as 'quality defects' are, in fact, an integral part of the developmental process. The dependence of fruit development on seed development conforms to the principal role of seeds in plant organ hierarchy. The peculiar character of apple fruit asymmetry is the consequence of modular growth of ' lopseeded ' syncarpous ovary that presumably provides developmental stimuli for all fruit tissues (including fruit stalk). Textural and physiological properties of fruit flesh will, therefore, serve to determine an organoleptic perception of the fruit. In addition, the progressive xylem dysfunction is programmed to minimise daily losses of solutes from the growing flesh (through diurnal cycling), and so to increase assimilate partitioning in favour of the developing fruit. The increased assimilate content of the fruit enhances its attractiveness to dispersal agents. However, inter-linked processes that serve to adjust for high sink demands are not without cost. The bias of assimilate partitioning at the expense of xylem sap flow will presumably affect mineral composition (notably calcium) and storage potential of the fruit. This naturally­ occurring mechanism is, therefore, the putative link to the occurrence of calcium-related disorders that also affect other fruit crop species, such as pear, avocado, mango, tomato, watermelons, cabbage, carrots, celery and capsicum. Hence, it is hoped that this thesis will improve the understanding of the physiological processes investigated here and will help initiate new approaches to optimise crop production. The techniques described here can be used to further understanding of developmental processes in other fruit crop species. However, due to the morphological diversity of fruit, the interactions between the processes of fruit development should be assessed within the limits of fruit structure. In general, apple can be used as a model fruit to illustrate patterns in quality development but more specifically, development of a particular quality attribute should be aligned with the relationships between structure and function o f its fruit tissues. For this reason, an integrative structure: function approach to the processes of fruit tissue development will lead to better strategies to remedy the negative effects of some developmental patterns in fruit. References 129 References Abbott, D.L. 1 959. The effects of seed removal on the growth of apple fruitlets. Annual Reports of the Long Ashton Research Station for 1 95 8 : 52-56. Abercrombie, M., Hickman, C.J. and Johnson, M.L. 1 98 1 . The Penguin Dictionary of Biology, ih edition. Penguin Books, England. Abruzzese, A., Mignani, 1. And Cocucci, S.M. 1 995. Nutritional status in apples and june drop. Journal of the American Society for Horticultural Science, 1 20: 7 1 -74. Alderman, W.H. 1 9 1 8 . Experimental work on self-sterility of the apple. Proceedings of the American Society for Horticultural Science, 1 4: 94- 1 0 1 . Allan, A.C. and Rubery, P.H. 1 99 1 . Calcium deficiency and auxin transport in Cucurbita pepo L. seedlings. Planta, 1 83 : 604-6 1 2 . Aloni, R . 1 980. Role of auxin and sucrose in the differentiation of sieve and tracheary elements in plant tissue cultures. Planta, 1 50: 255-263. Aloni, R. 1 987. Differentiation of vascular tissues. Annual Review of Plant Physiology, 3 8 : 1 79-204. Aloni, R. 1 99 5 . The induction of vascular tissues by auxin and cytokinin. In: Plant Hormones. Physiology, Biochemistry and Molecular Biology, 2nd edition. PJ. Davies (ed). Kluwer Academic Publishers, Dordrecht, Boston, London. Pp 5 3 1 -546. Aloni, R. and Bamett, J.R. 1 996. The development of phloem anastomoses between vascular bundles and their role in xylem regeneration after wounding in Cucurbita and Dahlia. Planta, 1 98 : 595 -603 . References --------------- -_ .. - 1 30 Aloni, R and Zimmermann, M.H. 1 983 . The control of vessel size and density along the plant axis. A new hypothesis. Differentiation, 24: 203-208. Aloni, R., Feigenbaum, P. , Kalev, N. and Rozovsky, S . 2000. Hormonal control of vascular differentiation in plants: the physiological basis of cambium ontogeny and xylem evolution. In: Cell & Molecular Biology of Wood Formation. R Savidge, J. Barnett and R. Napier (eds). BIOS Scientific Publishers Ltd, Oxford. Pp 223-236. Andrich, G., Fiorentini, R., Tuci, A. and Galoppini, C. 1 989. Skin permeability to oxygen in apples stored in controlled atmosphere. Journal of the American Society for Horticultural Science, 1 1 4: 770-775. Anon., 200 1 . Horticulture Facts and Figures. An annual publication from: The Horticulture & Food Research Institute of New Zealand Ltd. Avanzi, S . , Vernieri, P. , Ceccarelli, N., Tagliasacchi, A.M., Forino, L.M.C., Lorenzi, R and Bonomo, R. 1 988. Developmental stages and gibberellin activity in ovules of abscission-affected fruits of Malus domestica Borkh. Journal of Plant Physiology, 1 32 : 1 0- 1 5 . Bain, J.M. and Robertson, RN. 1 95 1 . The physiology o f growth i n apple fruits. L Cell size, cell number and fruit development. Australian Journal of Scientific Research, 4: 75-9 1 . Bangerth, F . 1 973 . Investigations upon Ca related physiological disorders. Phytopathologische Zeitschrift, 77: 20-37. Bangerth, F. 1 976. A role for auxin and auxin transport inhibitors on the Ca content of artificially induced parthenocarpic fruits. Physiologia Plantarum, 37: 1 9 1 - 1 94. Bangerth, F. 1 989. Dominance among fruits/sinks and the search for a correlative signal. Physiologia Plantarum, 76: 608-6 1 4. References Bangerth, F. 1 993. Polar auxin transport as a signal in the regulation o f tree and fruit development. Acta Horticulturae, 329: 70-76. Bangerth, F.K. 1 997. Can regulatory mechanism in fruit growth and development be elucidated through the study of endogenous hormone concentrations. Acta Horticulturae, 463 : 77-87. 1 3 1 Bangerth, F. 2000. Abscission and thinning of young fruit and their regulation by plant hormones and bioregulators. Plant Growth Regulation, 3 1 : 43-59. Bangerth, F. and Firuzeh, P. 1 97 1 . Der einfluB von 2,3,5 -trij odbenzoesaure (Tiba) auf den mineralstoffgehalt und die stippigkeit von "Boskoop"-ftiichten. Zeitschrift fUr Pflanzenkrankheiten und Pflanzenschutz, 78: 93-97. Bangerth, F.K., Gruber, J.D. and Shehata, S. 1 989. Auxin transport in relation to dominance and development of reproductive structures. R.K. Atkin and D.R. Clifford (eds). British Plant Growth Regulator Group, Monograph No 1 8 : 5 5-69. Banuelos, G.S . , Bangerth, F. and Marschner, H. 1 987. Relationship between polar basipetal auxin transport and acropetal Ca2+ transport into tomato fruits. Physiologia Plantarum, 7 1 : 3 2 1 -327. Barden, J.A. and Thompson, A.H. 1 963. Developmental anatomy of vascular tissues in York Imperial apple, with special emphasis on the pedicel. University of Maryland Agricultural Experiment Station Bulletin, A- 1 3 1 : 1 -5 3 . Barritt, B.H. 1 970. Ovule development i n seeded and seedless grapes. Vitis, 9 : 7- 1 4 . Batlle, 1 . , Alston, F.H. and Evans, K.M. 1 995. The use of the isoenzymic marker gene Got-l in the recognition of incompatibility S alleles in apple. Theoretical and Applied Genetics, 90: 303-306. References 1 32 Baumann, H. and Henze, J. 1 98 3 . Intercellular space volume o f fruit. Acta Horticulturae, 1 3 8 : 1 07- 1 1 1 . Bell, H.P. 1 93 7 . The protective layers of the apple. Canadian Journal of Research, 1 5 C: 3 9 1 -402. Ben-Arie, R., Kislev, N. and Frenkel, C. 1 979. Ultrastructural changes in the cell walls of ripening apple and pear fruit. Plant Physiology, 64: 1 97-202. Bergh, O. 1 98 5 . Effect of the previous crop on cortical cell number o f Malus domestica cv. Starking Delicious apple flower primordia, flowers and fruit. South African Journal of Plant and Soil, 2 : 1 9 1 - 1 96. Bergh, O. 1 990. Effect of time of hand-thinning on apple fruit size. South African Journal of Plant and Soil, 7 : 1 - 1 0. Blackman, F.F. 1 905. Optima and limiting factors. Annals o f Botany, 1 9: 2 8 1 -295. Blanpied, G.D. and Wilde, M.H. 1 968. A study of the cells in the outer flesh of developing McIntosh apple fruits. Botanical Gazette, 1 29: 1 73 - 1 8 3 . Bollard, E.G. 1 95 3 . The use of tracheal sap i n the study of apple-tree nutrition. Journal o f Experimental Botany, 4: 363-368. Bollard, E.G. 1 970. The physiology and nutrition of developing fruits. In: The Biochemistry of Fruits and their Products, Vol. 1 . A.c. Hulme (ed). Academic Press, London, New York. Pp 3 87-425 . Boskovic, R . and Tobutt, K.R. 1 999. Correlation o f stylar ribonuclease isoenzymes with incompatibility alleles in apple. Euphytica, 1 07 : 29-43. Bradfield, E.G. 1 976. Calcium complexes in the xylem sap of apple shoots. Plant and Soil, 44: 495-499. References ----------------- ---- - 1 3 3 Bramlage, WJ., Weis, S .A. and Greene, D.W. 1 990. Observations on the relationships among seed number, fruit calcium, and senescent breakdown in apples. HortScience, 25 : 3 5 1 -3 5 3 . Brault, A-M. and d e Oliveira, D. 1 995. Seed number and an asymmetry index of ' McIntosh' apples. HortScience, 30: 44-46. Brookfield, P.L., Ferguson, LB., Watkins, C.B. and Bowen, J.H. 1 996. Seed number and calcium concentrations of 'Braebum' apple fruit. Journal of Horticultural Science, 7 1 : 265-27 1 . Broom, F.D., Smith, G.S. , Miles, D.B. and Green, T.G.A. 1 998. Within and between tree variability in fruit characteristics associated with bitter pit incidence of 'Braebum ' apple. Journal of Horticultural Science & Biotechnology, 73 : 5 5 5-56 1 . Broothaerts, W., Janssens, G.A., Proost, P . and Broekaert, W.F. 1 99 5 . cDNA cloning and molecular analysis of two self-incompatibility alleles from apple. Plant Molecular Biology, 27: 499-5 1 1 . Broothaerts, W., Verdoodt, L., Keulemans, J. , Janssens, G.A. and Broekaert, W.F. 1 996. The self-incompatibility gene in apple and determination of the S-genotype of apple cultivars by PCR. Acta Horticulturae, 423 : 1 03- 1 09. Brown, K.M. 1 997. Ethylene and abscission. Physiologia Plantarum, 1 00: 567-576. Bultitude, J. 1 98 3 . Apples. A guide to the Identification of Intemational Varieties. Macmillan Press, London. Burton, W.G. 1 982. Post-harvest physiology of food crops. Longman, London, New York. Bustan, A., Emer, Y. and Goldschmidt, E.E. 1 995. Interactions between developing Citrus fruits and their supportive vascular system. Annals of Botany, 76: 65 7-666. References 1 34 Byers, R.E., Carbaugh, D.H., Presley, C.N. and Wolf, T.K. 1 99 1 . The influence of low light on apple fruit abscission. Journal of Horticultural Science, 66: 7 - 1 7. Calbo, A.G. and Sommer, N.F. 1 987. Intercellular volume and resistance to air flow of fruits and vegetables. Journal of the American Society for Horticultural Science, 1 1 2 : 1 3 1 - 1 34. Calkin, H.W., Gibson, A.C. and Nobel, P.S. 1 986. Biophysical model of xylem conductance in tracheids of the fern Pteris vittata. Journal of Experimental Botany, 3 7 : 1 054- 1 064. Callej as, R. and Bangerth, F. 1 997. Is auxin export of apple fruit an alternative signal for inhibition of flower bud induction. Acta Horticulturae, 463 : 27 1 -277. Child, R.D. 1 966. Pollen tube growth in relation to temperature and ovule longevity in the cider apple michelin. Annual Reports of the Long Ashton Research Station for 1 965: 1 1 5- 1 20. Clark, C.J. and Burmeister, D.M. 1 999. Magnetic resonance imaging of browning development in 'Braeburn' apple during controlled-atmosphere storage under high C02. HortScience, 34: 9 1 5-9 1 9. Clark, C.J. and Richardson, c.A. 1 999. Observation of water core dissipation in ' Braeburn' apple by magnetic resonance imaging. New Zealand Journal of Crop and Horticultural Science, 27: 47-52. Costa Tura, J. and MacKenzie, K.A.D. 1 990. Ovule and embryo sac development in Malus pumila L. cv. Cox' s Orange Pippin, from dormancy to blossom. Annals of Botany, 66: 443-450. Creasy, G.L., Price, S.P. and Lombard, P.B. 1 993. Evidence for xylem discontinuity in Pinot Noir and Merlot grapes: dye uptake and mineral composition during berry maturation. American Journal of Enology and Viticulture, 44: 1 87- 1 92. References 1 3 5 Currie, A.l., Ganeshanandam, S. , Noiton, D.A., Garrick, D. , Shelbourne, C.J.A. and Oraguzie, N. 2000. Quantitative evaluation of apple (Malus x domestica Borkh.) fruit shape by principal component analysis of Fourier descriptors. Euphytica, 1 1 1 : 2 1 9-227. Dadzie, B.K. 1 992. Gas exchange characteristics and quality o f apples. PhD Dissertation, Massey University, Palmerston North, New Zealand. Davies, P.l. 1 995. The plant hormones: their nature, occurrence, and functions. In: Plant Hormones. Physiology, Biochemistry and Molecular Biology, 2nd edition. P.l . Davies (ed). Kluwer Academic Publishers, Dordrecht, Boston, London. Pp 1 - 1 2 . Deckers, T. and Porreye, W. 1 984. Influence o f the temperature o n pollen germination of different cultivars of apple and pear: trials in vitro. Acta Horticulturae, 1 49 : 1 23 - 1 29. DeGrandi-Hoffrnan, G., Hoopingarner, R. and Baker, K.K. 1 985. The influence of honey bee "sideworking" behavior on cross-pollination and fruit set in apples. HortScience, 20: 397-399. DeGrandi-Hoffrnan, G., Terry, 1., Morales, F. and Weinhold, C . 1 990. Influence of blossom density, bloom phenology, honey bee foraging activity and blossom quality on apple fruit set. Southwestern Entomologist, 1 5 : 2 8 1 -289. dela Fuente, R.K. and Leopold, A.C. 1 973. A role for calcium in auxin transport. Plant Physiology, 5 1 : 845-847. del a Fuente, R.K. 1 984. Role of calcium in the polar secretion of indoleacetic acid. Plant Physiology, 76 : 342-346. de Guzman, C.c. and dela Fuente, R.K. 1 984. Polar calcium flux in sunflower hypocotyl segments. 1. The effect of auxin. Plant Physiology, 76: 347-3 5 2 . References 1 36 de Guzman, C.C. and dela Fuente, R.K. 1 986. Polar calcium flux in sunflower hypocotyl segments. n. The effect of segment orientation, growth, and respiration. Plant Physiology, 8 1 : 408-4 1 2 . Delbarre, A., Muller, P . , Imhoff, V. and Guern, J . 1 996. Comparison of mechanisms controlling uptake and accumulation of 2,4-dichlorophenoxy acetic acid, naphthalene- I -acetic acid, and indole-3-acetic acid in suspension-cultured tobacco cells. Planta, 1 98 : 532-54 1 . del Campillo, E. and Bennett, A.B. 1 996. Pedicel breakstrength and cellulase gene expression during tomato flower abscission. Plant Physiology, 1 1 1 : 8 1 3-820. Demarty, M., Morvan, C. and Thellier, M. 1 984. Calcium and the cell wall. Plant, Cell and Environment, 7: 44 1 -448 . Denne, M.P. 1 960. The growth of apple fruitlets, and the effect of early thinning on fruit development. Annals of Botany, 24: 397-406. Denne, M.P. 1 963. Fruit development and some tree factors affecting it. New Zealand Journal of Botany, 1 : 265-294. Drazeta, L., Lang, S. , Hall, A., Volz, R. and Jameson, P.E. 2000. Seed set and the development of fruit shape in apple. Proceedings of a Seed Symposium. Massey University, Palmerston North, New Zealand, 1 2 February 1 999. M.T. McManus, H.A. Outred and K.M. Pollock (eds). Pp 99- 1 0 1 . Drazeta, L., Lang, A., Morgan, L., Volz, R. and Jameson, P.E. 200 1 . Bitter pit and vascular function in apples. Acta Horticulturae, 5 64: 3 87-392. During, H., Lang, A. and Oggionni, F. 1 987. Patterns of water flow in Riesling berries in relation to developmental changes in their xylem morphology. Vitis, 26: 1 23 - 1 3 1 . References 1 3 7 During, H. and Lang, A. 1 993 . Xylem development and function in the grape peduncle: Relations to bunch stem necrosis. Vitis, 32: 1 5-22. Ebadi, A., Sedgley, M., May, P. and Coombe, B.G. 1 996. Seed development and abortion in Vitis vinifera L., cv. Chardonnay. International Journal of Plant Sciences, 1 5 7 : 703-7 1 2 . Einset, O. 1 93 9. Correlation of number o f seed to weight of fruit i n the apple. Gartenbauwissenschaft, 1 3 : 3 5 1 -3 5 3 . Elgar, H J . , Bunneister, D.M. and Watkins, C . B . 1 998. Storage and handling effects o n a CO2-related internal browning disorder of 'Braeburn' apples. HortScience, 3 3 : 7 1 9-722. ENZA. 1 995. Specifications Manual. ENZA New Zealand (International) Operations, Hastings. Esau, K. 1 977. Anatomy of Seed Plants, 2nd edition. John Wiley & Sons, New York, Chichester, Brisbane, Toronto, Singapore. Ferguson, LB. 1 984. Calcium in plant senescence and fruit ripening. Plant, Cell and Environment, 7 : 477-489. Ferguson, LB. and Bollard, E.G. 1 976. The movement of calcium in woody stems. Annals of Botany, 40: 1 057- 1 06 5 . Ferguson, LB. and Watkins, C.B. 1 989. Bitter pit i n apple fruit. Horticultural Reviews, 1 1 : 289-3 5 5 . Ferguson, 1 . , Volz, R . and Woolf, A . 1 999. Preharvest factors affecting physiological disorders of fruit. Postharvest Biology and Technology, 1 5 : 255-262. References 1 3 8 Ferree, D.C. and Palmer, J.W. 1 982. Effect of spur defoliation and ringing during bloom on fruiting, fruit mineral level, and net photosynthesis of ' Golden Delicious' apple. Journal of the American Society for Horticultural Science, 1 07 : 1 1 82- 1 1 86. Findlay, N., Oliver, KJ., Nii, N. and Coombe, B.G. 1 987. Solute accumulation by grape pericarp cells. IV . Perfusion of pericarp apoplast via the pedicel and evidence for xylem malfunction in ripening berries. Journal of Experimental Botany, 3 8 : 668-679. Forino, L.M.C., Tagliasacchi, A.M. and A vanzi, S. 1 987. Embryo-sacs frequency in ovules of abscission affected fruits in Malus domestica Borkh. Advances in Horticultural Science 1 : 65-67. Free, J .B. 1 962. The effect of distance from pollinizer varieties on the fruit set on trees in plum and apple orchards. Journal of Horticultural Science, 37: 262-27 1 . Free, J.B. and Spencer-Booth, Y. 1 964. The effect of distance from pollinizer varieties on the fruit set of apple, pear and sweet-cherry trees. Journal of Horticultural Science, 39: 54-60. Fuller, M.M. 1 976. The ultrastructure of the outer tissues of cold-stored apple fruits of high and low calcium content in relation to cell breakdown. Annals of Applied Biology, 8 3 : 299-304. Glenn, G.M. and Poovaiah, B.W. 1 990. Calcium-mediated postharvest changes in texture and cell wall structure and composition in 'Golden Delicious' apples. Journal of the American Society for Horticultural Science, 1 1 5 : 962-968. Goffinet, M.C. and Maloney, K. 1 987. Histological studies of the developing fruit of apple (Malus domestica). American Journal o f Botany, 74: 629. References 1 39 Goffinet, M.C., Robinson, T.L. and Lakso, A.N. 1 995. A comparison of 'Empire' apple fruit size and anatomy in unthinned and hand-thinned trees. Journal of Horticultural Science, 70: 375-387. Goldway, M., Shai, 0., Yehuda, H., Matityahu, A. and Stem, R.A. 1 999. 'Jonathan' apple is a lower potency pollenizer of 'Topred' than ' Golden Delicious' due to partial S-allele incompatibility. Journal of Horticultural Science & Biotechnology, 74: 3 8 1 -3 8 5 . Goldway, M., Schneider, D . , Yehuda, H . , Matityahu, A . , Eisikowitch, D. and Stem, R.A. 200 1 . The effect of apple S-allele compatibility on fruit set levels in non-optimal fertilization conditions. Acta Horticulturae, 5 6 1 : 23 1 -234. Greene, D.W., Lord, WJ. and Bramlage, WJ. 1 982. Effects of gibber ell ins �+7 and 6- benzylamino purine on fruit set, fruit characteristics, seed content, and storage quality of 'McIntosh' apples. HortScience, 1 7 : 653-654. Gross, K.C. and Sams, c.E. 1 984. Changes in cell wall neutral sugar composition during fruit ripening: a species survey. Phytochemistry, 2 3 : 2457-246 1 . Gruber, J. and Bangerth, F. 1 990. Diffusible lAA and dominance phen<:>mena in fruits of apple and tomato. Physiologia Plantarum, 79: 3 54-35 8 . Gustafson, F.G. 1 93 9 . Auxin distribution i n fruits and its significance i n fruit development. American Journal of Botany, 26: 1 89- 1 94. Haasbroek, F.J., Strydom, D.K., Thiele, 1. and Hurter, N. 1 967. Pollen tube growth in apple flowers as detennined with 32P-Iabelled pollen. South African Journal of Agricultural Science, 1 0 : 1 0 1 5 - 1 02 1 . Habdas, H., Jankiewicz, L.S. and Borkowska, B . 1 982. Changes in anatomical structure of apple fruitlet pedicels preceding June drop. Acta Agrobotanica, 3 5 : 1 1 -2 3 . References Haga, K. and lino, M. 1 998. Auxin-growth relationships in maize coleoptiles and pea internodes and control by auxin of the tissue sensitivity to auxin. Plant Physiology, 1 1 7 : 1 473- 1 486. 1 40 Hamamoto, H. , Shishido, Y., Furuya, S. and Yasuba, K. 1 99 8 . Growth and development of tomato fruit as affected by 2,3,5-Triiodobenzoic acid (TIBA) applied to the peduncle. Journal of the Japanese Society for Horticultural Science, 67: 2 1 0-2 1 2. Harker, F.R. and Ferguson, LB. 1 988a. Transport of calcium across cuticles isolated from apple fruit. Scientia Horticulturae, 36: 205-2 1 7. Harker, F.R. and Ferguson, LB. 1 988b. Calcium ion transport across discs of the cortical flesh of apple fruit in relation to fruit development. Physiologia Plantarum, 74: 695-700. Harker, F.R. and Hallett, LC. 1 992. Physiological changes associated with development of mealiness of apple fruit during cool storage. HortScience, 2 7 : 1 29 1 - 1 294. Harker, F.R., Redgwell, RJ., Hallett, LC., Murray, S.H. and Carter, G. 1 997. Texture of fresh fruit. Horticultural Reviews, 20: 1 2 1 -224. Harley, c.P. 1 942. Seasonal Growth and dry matter accumulation in Winesap apples. Proceedings of the American Society for Horticultural Science, 40: 1 65 - 1 68. Hartman, F.O. and Howlett, F.S. 1 954. Fruit setting of the Delicious apple. Ohio Agricultural Experiment Station Research Bulletin, 745 . Hatfield, S.G.S. and Knee, M. 1 988. Effects of water loss on apples in storage. International Journal of Food Science and Technology, 23 : 5 75-583 . Heinicke, AJ. 1 9 1 7. Factors influencing the abscission of flowers and partially developed fruits of the apple (Pyrus malus L.). Cornell University Agricultural Experiment Station Bulletin, 393 : 43- 1 1 4. ---- --- ��- - -----� References 1 4 1 Heinicke, A.J. 1 920. The seed content and the position o f the fruit as factors influencing stippen in apples. Proceedings of the American Society for Horticultural Science, 1 7 : 225-232. Higgs, K.H. and Jones, H.G. 1 984. A microcomputer-based system for continuous measurement and recording fruit diameter in relation to environmental factors. Journal of Experimental Botany, 3 5 : 1 646- 1 65 5 . Himelrick, D.G. and Ingle, M. 1 98 1 . Calcium levels of apple leaves an d fruit following tree sprays with EDT A, oxalic acid, TIBA, and calcium chloride. HortScience, 1 6 : 1 67- 1 68. Howlett, F.S. 1 928. Fruit setting in the Delicious apple. Proceedings of the American Society for Horticultural Science, 25 : 1 43 - 1 48 . Howlett, F .S. 1 93 8 . Factors affecting the rate and course of development of the female gametophyte in apple varieties. Proceedings of the American Society for Horticultural Science, 3 5 : 1 05- 1 1 0. Imhoff, V. , Muller, P., Guem, J. and Delbarre, A. 2000. Inhibitors of the carrier-mediated influx of auxin in suspension-cultured tobacco cells. Planta, 2 1 0: 5 80-5 8 8 . Jacobs, W.P. 1 952. The role of auxin i n differentiation of xylem around a wound. American Journal of Botany, 39: 3 0 1 -309. Janick, J., Cummins, J.N., Brown, S.K. and Hemmat, M. 1 996. Apples. In: Fruit Breeding, Vol. I, Tree and Tropical fruits. J. Janick and J.N. Moore (eds). John Wiley & Sons, Inc., New York, Chichester, Brisbane, Toronto, Singapore. Pp 1 -77. Janssens, G.A., Goderis, l.J., Broekaert, W.F. and Broothaerts, W. 1 99 5 . A molecular method for S-allele identification in apple based on allele-specific PCR. Theoretical and Applied Genetics, 9 1 : 691 -698. References 1 42 Johnson, D.S. 1 976. Influence of water loss on the storage quality of apples. Chemistry and Industry, 1 8 : 1 044- 1 046. Jona, R. and Foa, E. 1 979. Histochemical survey of cell-wall polysaccharides of selected fruits. Scientia Horticulturae, 1 0: 1 4 1 - 1 47. Jones, H.G. and Higgs, K.H. 1 982. Surface conductance and water balance of developing apple (Malus pumila Mill.) fruits. Journal of Experimental Botany, 3 3 : 67-77. Jones, H.G. and Higgs, K.H. 1 98 5 . Water movement into and out of apple fruits. Acta Horticulturae, 1 7 1 : 3 5 3 -359. Jones, H.G. and Samuelson, TJ. 1 98 3 . Calcium uptake by developing apple fruits. 1I. The role of spur leaves. Journal of Horticultural Science, 5 8 : 1 83 - 1 90. Jones, H.G., Higgs, K.H. and Samuelson, TJ. 1 98 3 . Calcium uptake by developing apple fruits. 1. Seasonal changes in calcium content of fruits. Journal of Horticultural Science, 5 8 : 1 73 - 1 82. Khan, A.A. and Vincent, J.F.V. 1 990. Anisotropy of apple parenchyma. Journal of the Science of Food and Agriculture, 5 2 : 455-466. Khan, A.A. and Vincent, J.F.V. 1 993 . Compressive stiffness and fracture properties of apple and potato parenchyma. Journal of Texture Studies, 24: 423-43 5 . Klee, H J . , Horsch, R.B., Hinchee, M.A., Hein, M.B. and Hoffmann, N.L. 1 987. The effects of overproduction of two Agrobacterium tumefaciens T -DNA auxin biosynthetic gene products in transgenic petunia plants. Genes & Development, 1 : 86-96. References Knee, M . and Bartley, LM. 1 98 1 . Composition and metabolism of cell wall polysaccharides in ripening fruits. In: Recent Advances in the Biochemistry of Fruits and Vegetables. J. Friend and M.J.C. Rhodes (eds). Academic Press, London, New York. Pp 1 33 - 1 48. 1 43 Knoche, M., Schroder, M. and Hinz, M. 2000. Control of the development of mummified fruit of ' Elstar' apple. Journal of Horticultural Science & Biotechnology, 75 : 328-3 3 5 . Kobel, F . , Steinegger, P . and Anliker, J. 1 939. Weitere untersuchungen iiber die befruchtungsverhaltnisse der apfel- und birnsorten. Landwirtschaftliches Jahrbuch der Schweiz, 5 3 : 1 60- 1 9 1 . Kollmann, F.F.P. and Cote, W.A. 1 968. Principles o f Wood Science and Technology. 1. Solid Wood. George Allen & Unwin Ltd, London. Springer-Verlag, Berlin, Heidelberg, New York. Korn, R.W. 1 980. The changing shape of plant cells: transformations during cell proliferation. Annals of Botany, 46: 649-666. Korn, RW. 1 984. Cell shapes and tissue geometries. In: Positional Controls in Plant Development. P.W. Barlow and DJ. Carr (eds). Cambridge University Press, Cambridge, New York. Pp 33-52. Kraus, E.J . and Ralston, G.S. 1 9 1 6. The pollination of the pomaceous fruits. IU. Gross vascular anatomy of the apple. Oregon Agricultural College Experiment Station Bulletin, 1 3 8 : 4- 1 2 . Kron, P . , Husband, B .C., Kevan, P.G. and Belaoussoff, S . 200 l a. Factors affecting pollen dispersal in high-density apple orchards. HortScience, 3 6 : 1 03 9- 1 046. References 1 44 Kron, P., Husband, B.c. and Kevan, P.G. 200 1 b. Across- and along-row pollen dispersal in high-density apple orchards: insights from allozyme markers. Journal of Horticultural Science & B iotechnology, 76: 2 86-294. Kuhn, E.D. and Ambrose, J.T. 1 982. Foraging behavior of honey bees on ' Golden Delicious' and ' Delicious' apple. Journal of the American Society for Horticultural Science, 1 07 : 39 1 -395 . Lang, A. 1 979. A relay mechanism for phloem translocation. Annals of Botany, 44 : 1 4 1 - 1 45 . Lang, A . 1 990. Xylem, phloem and transpiration flows i n developing apple fruits. Journal of Experimental Botany, 4 1 : 645-65 1 . Lang, A. and Thorpe, M.R. 1 989. Xylem, phloem and transpiration flows in a grape: application of a technique for measuring the volume of attached fruits to high resolution using Archimedes' principle. Journal of Experimental Botany, 40: 1 069- 1 078. Lang, A. and During, H. 1 99 1 . Partitioning control by water potential gradient: evidence for compartmentation breakdown in grape berries. Journal of Experimental Botany, 42 : 1 1 1 7- 1 1 22. Lang, A. and Ryan, K.G. 1 994. Vascular development and sap flow in apple pedicels. Annals of Botany, 74 : 3 8 1 -388. Lang, A. and Volz, R.K. 1 998. Spur leaves increase calcium in young apples by promoting xylem inflow and outflow. Journal of the American Society for Horticultural Science, 1 23 : 956-960. References 1 45 Lang, A., Thorpe, M.R and Edwards, W.RN. 1 986. Plant water potential and translocation. In: Phloem transport. Proceedings of an International Conference on Phloem Transport, Asilomar, California, USA, 1 8-23 August 1 98 5 . J. Cronshaw, WJ. Lucas and Giaquinta, R.T. (eds). Alan R Liss, Inc. , New York. Pp 1 93 - 1 94. Latimer, L.P. 1 93 1 . Further observations on factors affecting fruit setting of the McIntosh apple in New Hampshire. Proceedings o f the American Society for Horticultural Science, 2 8 : 87-92. Latimer, L.P. 1 93 7. Self- and cross-pollination in the McIntosh apple and some of its hybrids. Proceedings of the American Society for Horticultural Science, 34: 1 9-2 1 . Lee, n.R 1 98 1 . Elasticity of phloem tissues. Journal o f Experimental Botany, 3 2 : 25 1 -260. Lee, n.R. 1 989. Vasculature of the abscission zone of tomato fruit: implications for transport. Canadian Journal of Botany, 67: 1 898- 1 902. Lewis, F.T. 1 923. The typical shape of polyhedral cells in vegetable parenchyma and the restoration o f that shape following cell division. Proceedings o f the American Academy of Arts and Sciences, 5 8 : 5 3 7-552. Lewis, T.L. 1 980. The rate of uptake and longitudinal distribution of potassium, calcium and magnesium in the flesh of developing apple fruits of nine cultivars. Journal of Horticultural Science, 5 5 : 5 7-63 . Lewis, T.L. and Martin, D. 1 973 . Longitudinal distribution o f applied calcium, and of naturally occurring calcium, magnesium, and potassium, in Merton apple fuits. Australian Journal of Agricultural Research, 24: 363-3 7 1 . References Liebig, J . 1 84 1 . Organic Chemistry in Its Applications to Agriculture and Physiology, 1 st American edition. John Owen, Cambridge. 1 46 Lomax, T.L., Muday, G.K. and Rubery, P.H. 1 995. Auxin transport. In : Plant Hormones. Physiology, Biochemistry and Molecular Biology, 2nd edition. P.J. Davies (ed). Kluwer Academic Publishers, Dordrecht, Boston, London. Pp 5 09-530. Luckwill, L.C. 1 948. The hormone content of the seed in relation to endosperm development and fruit drop in the apple. Journal of Horticultural Science, 24: 3 2-44. Luckwill, L.C . 1 95 3 . Studies of fruit development in relation to plant hormones. I . Hormone production by the developing apple seed in relation to fruit drop. Journal of Horticultural Science, 2 8 : 1 4-24. MacArthur, M. and Wetmore, R.H. 1 939. Developmental studies in the apple fruit in the varieties McIntosh Red and Wagener. I. Vascular anatomy. Journal of Pomology and Horticultural Science, 1 7 : 2 1 8-232. MacArthur, M. and Wetmore, R.H. 1 94 1 . Developmental studies of the apple fruit in the varieties McIntosh Red and Wagener. H. An analysis of development. Canadian Journal of Research, 1 9C : 3 7 1 -3 82 . MacDaniels, L.H. 1 937. Some anatomical aspects of apple flower and fruit abscission. Proceedings of the American Society for Horticultural Science, 34: 1 22 - 1 29. MacDaniels, L.H. 1 940. The morphology of the apple and other pome fruits. Cornell University Agricultural Experiment Station Memoir, 230. MacKenzie, K.A.D. 1 98 8 . The anatomy of fruit abscission in loganberries. Annals of Botany, 62: 249-263 . References - --- ---- ------------ 1 47 MacKenzie, K.A.D and. Costa Tura, J. 1 99 1 . Xylem development in the gynoecium of the apple (Malus pumila L.) Cv. Cox 's Orange Pippin. Annals of Botany, 67: 3 83-389. Maggs, D.H., Martin, G.J. and Needs, RA. 1 97 1 . The spread of cross-pollination in a solid block of Granny Smith apples. Australian Journal of Experimental Agriculture and Animal Husbandry, 1 1 : 1 1 3 - 1 1 7. Maguire, K.M., Banks, N.H., Lang, A. and Gordon, LL. 2000. Harvest date, cultivar, orchard, and tree effects on water vapor penneance in apples. Journal of the American Society for Horticultural Science, 1 2 5 : 1 00- 1 04. Manganaris, A.G. and Alston, F.H. 1 987. Inheritance and linkage relationships of glutamate oxaloacetate transaminase isoenzymes in apple. Theoretical and Applied Genetics, 74: 1 54- 1 6 1 . Mapfumo, E., Aspinall, D., Hancock, T. and Sedgley, M. 1 993. Xylem development in relation to water uptake by roots of grapevine ( Vitis vinifera L.). New Phytologist, 1 2 5 : 93-99. Marini, R and Byers, R. 1 988. Methods for evaluating chemical inhibitors of apple abscission. HortScience, 2 3 : 849-85 1 . Martin, D. and Lewis, T.L. 1 952. The physiology of growth in apple fruits. Ill. Cell characteristics and respiratory activity of light and heavy crop fruits. Australian Journal of Scientific Research, 5 : 3 1 5-327. Martin, D., Lewis, T.L. and Cerny, J. 1 967. Nitrogen metabolism during storage in relation to the breakdown of apples. I. Changes in protein nitrogen level in relation to incidence. Australian Journal of Agricultural Research, 1 8 : 27 1 -278. Marschner, H. 1 986. Mineral Nutrition of Higher Plants. Academic Press, London, San Diego, New York, Berkeley, Boston, Sidney, Tokyo, Toronto. References 1 48 Marvin, J.W. 1 939. The shape of compressed lead shot and its relation to cell shape. American Journal of Botany, 26: 280-288. Matsumoto, S., Kitahara, K., Komori, S. and Soejima, J. 1 999. A new S-allele in apple, 'Sg', and its similarity to the 'Sf allele from 'Fuji ' . HortScience, 34: 708-7 1 0. Matzke, E.B. 1 939. Volume-shape relationships in lead shot and their bearing on cell shapes. American Journal of Botany, 26: 288-295 . McAlpine, D. 1 9 1 2 . The fibro-vascular system of the apple (pome) and its functions. Proceedings of the Linnean Society of New South Wales, 3 6 : 6 1 3 -62 5 . Michalke, W . , Katekar, G.E. and Geissler, A.E. 1 992. Phytotropin-binding sites and auxin transport in Cucurbita pepo: evidence for two recognition sites. Planta, 1 8 7: 254-260. Miller, R.H. 1 982. Apple fruit cutic1es and the occurrence of pores and transcuticular canals. Annals of Botany, 50: 3 5 5 -37 1 . Milutinovic, M., Surlan-Momirovic, G. and Nikolic, D. 1 996. Relationship between pollinizer distance and fruit set in apple. Acta Horticulturae, 423 : 9 1 -94. Modlibowska, I. 1 945. Pollen tube growth and embryo-sac development in apples and pears. Journal of Pomology and Horticultural Science, 2 1 : 5 7-89. Murneek, A.E. and Schowengerdt, G.C. 1 93 5 . A study o f the relation o f size o f apples to number of seeds and weight of spur leaves. Proceedings o f the American Society for Horticultural Science, 3 3 : 4-6. Nii, N. 1 980. Seasonal changes in growth and enlargement of the Japanese pear fruit, Pyrus serotina cv Shinsheiki, in relation to vascular bundle development in the pedicel and flesh. Journal of Horticultural Science, 5 5 : 3 85-396. References 1 49 Nitsch, J.P. 1 950. Growth and morphogenesis of the strawberry as related to auxin. American Journal of Botany, 3 7 : 2 1 1 -2 1 5 . Nitsch, J.P. 1 970. Hormonal factors in growth and development. In: The Biochemistry of Fruits and their Products, Vol. 1 . AC. Hulme (ed). Academic Press, London, New York. Pp 427-47 1 . Nobel, P.S. 1 999. Physicochemical & Environmental Plant Physiology, 2nd edition. Academic Press, San Diego, London, Boston, New York, Sidney, Tokyo, Toronto. Olmo, H.P. 1 934. Empty-seededness in varieties of Vitis vinifera. Proceedings of the American Society for Horticultural Science, 3 2 : 3 76-380. Opara, L.v. 1 993. A study of stem-end splitting in apples. PhD Dissertation, Massey University, Palmerston North, New Zealand. Opara, L.U. 2000. Fruit growth measurement and analysis. Horticultural Reviews, 24: 373-43 1 . Osborne, D.J. 1 989. Abscission. CRC Critical Reviews in Plant Sciences, 8 : 1 03 - 1 29. Pandita, V.K. and Jindal, K.K. 1 99 1 . Enzymatic and anatomical changes in abscission zone cells of apple fruits induced by ethephon. Biologia Plantarurn (Praha), 3 3 : 20-2 5 . Pearson, J . A and Robertson, R.N. 1 95 3 . The physiology of growth i n apple fruits. IV. Seasonal variation in cell size, nitrogen metabolism, and respiration in developing Granny Smith apple fruits. Australian Journal of Biological Sciences, 6: 1 -20. Percy, AE., O 'Brien, LE.W., Jameson, P.E., Melton, L.D., MacRae, E.A. and Redgwell, RJ. 1 996. Xyloglucan endotransglycosylase activity during fruit development and ripening of apple and kiwifruit. Physiologia Plantarum, 96: 43-50. References Pitt, R.E. 1 982. Models for the rheology and statistical strength of unifonnly stressed vegetative tissue. Transactions of the ASAE, 25 : 1 776- 1 784. 1 50 Pratt, C. 1 98 8 . Apple flower and fruit: morphology and anatomy. Horticultural Reviews, 1 0: 273-308. Prive, J-P. , Elfving, D.C. and Proctor, J.T.A. 1 988. Pedicel characteristics of four apple cultivars. Fruit Varieties Journal, 42 : 1 22- 1 25 . Prive, J-P. , Elfving, D.C. and Proctor, J.T.A. 1 989. Paclobutrazol, gibberellin and cytokinin effects on growth, development, and histology o f apple pedicels and fruits. Journal of the American Society for Horticultural Science, 1 1 4 : 2 73-278 . Proctor, J.T.A. and Palmer, J .W. 1 99 1 . The role of spur and bourse leaves of three apple cultivars on fruit set and growth and calcium content. Journal of Horticultural Science, 66: 275-282. Proctor, J.T.A. and Schechter, L 1 992. Effect of ovule damage on fruit development in three apple cultivars. HortScience, 27: 1 8- 1 9. Quinlan, J.D. and Preston, A.P. 1 97 1 . The influence of shoot competition on fruit retention and cropping of apple trees. Journal of Horticultural Science, 46: 525-534. Rajapakse, N.C., Banks, N.R., Hewett, E.W. and Cleland, DJ. 1 990. Development of oxygen concentration gradients in flesh tissues o f bulky plant organs. Journal of the American Society for Horticultural Science, 1 1 5 : 793-797. Read, S.M. and Bacic, A. 1 996. Cell wall porosity and its detennination. In: Modem Methods of Plant Analysis, Vol. 1 7, Plant Cell Wall Analysis. H.F. Linskens and J.F Jackson (eds). Springer-Verlag, Berlin, Heidelberg. Pp 63-80. References ---� � --� Reeve, R.M. 1 95 3 . Histological investigations of texture in apples. n. Structure and intercellular spaces. Food Research, 1 8 : 604-6 1 7. 1 5 1 Riederer, M. and Schreiber, L. 1 995. Waxes - the transport barriers of plant cuticles. In: Waxes: Chemistry, Molecular Biology and Functions. RJ. Hamilton (ed). The Oily Press, West Ferry. Pp 1 3 1 - 1 56. Roberts, RH. 1 945. Blossom structure and setting of Delicious and other apple varieties. Proceedings of the American Society for Horticultural Science, 46: 87-90. Roberts, RH. 1 946. Notes on apple set and growth, 1 945. Proceedings of the American Society for Horticultural Science, 48 : 59-62. Robinson, W.S. 1 979. Effect of apple cultivar on foraging behavior and pollen transfer by honey bees. Journal of the American Society for Horticultural Science, 1 04: 5 96-598. Robinson, W.S. and Fell, RD. 1 98 1 . Effect of honey bee foraging behaviors on 'Delicious' apple set. HortScience, 1 6: 3 26-328. Rohitha, B.H. and Klinac, DJ. 1 990. Relationships between seed set and fruit weight and shape of nashi (Pyrus serotina Rehder var culta Rehder). New Zealand Journal of Crop and Horticultural Science, 1 8 : 1 33 - 1 36. Romano, c.P., Hein, M.B. and Klee, H.J. 1 99 1 . Inactivation of auxin in tobacco transformed with the indoleacetic acid-lysine synthetase gene of Pseudomonas savastanoi. Genes & Development, 5 : 438-446. Ruess, F. and SWsser, R. 1 993. Untersuchungen iiber das interzellularsystem bei apfelfiiichten mit methoden der digitalen bildverarbeitung. Gartenbauwissenschaft, 5 8 : 1 97-205 . References 1 52 Sachs, T. 1 98 1 . The control of the patterned differentiation of vascular tissues. In: Advances in Botanical Research, Vol. 9. H. W. Woolhouse (ed). Academic Press, London, New York, Toronto, Sydney, San Francisco. Pp 1 5 1 -262. Safiner, R.A., Conway, W.S. and Sarns, C.E. 1 998. Effect of post harvest calcium chloride treatments on tissue water relations, cell wall calcium levels and postharvest life of ' Golden Delicious' apples. Journal of the American Society for Horticultural Science, 1 2 3 : 893-897. Sakurai, K., Brown, S .K. and Weeden, N.F. 1 997. Determining the self-incompatibility alle1es of Japanese apple cuItivars. HortScience, 3 2 : 1 25 8- 1 25 9 . Sakurai, K . , Brown, S.K. and Weeden, N. 2000. Self-incompatibility alIe1es of apple cultivars and advanced selections. HortScience, 3 5 : 1 1 6- 1 1 9. Salisbury, F.B. and Ross, C.W. 1 992. Plant Physiology, 4th edition. Wadsworth Publishing Company, Belmont, California. SAS Institute Inc. 1 996. SAS/ST A T Software: Changes and enhancements through Release 6. 1 1 , Cary, North Carolina. SAS Institute Inc. 2000. SAS OnlineDoc(rtm), Version 8. SAS Institute, Cary, North Carolina. Sassa, H., Mase, N., Hirano, H. and Ikehashi, H. 1 994. Identification of self­ incompatibility-related glycoproteins in styles of apple (Malus x domestica) . Theoretical and Applied Genetics, 8 9 : 2 0 1 -205 . Sato, M. and Kanbe, K. 1 984. Studies on the ovule abortion of the triploid apple cultivar 'Mutsu ' . Bulletin of the Akita Prefecture College of Agriculture, 1 0 : 33-42 . References 1 5 3 Schneider, D., Stem, RA., Eisikowitch, D. and Goldway, M. 200 1 . Detennination of the self-fertilization potency of ' Golden Delicious' apple. Journal of Horticultural Science & Biotechnology, 76: 259-263 . Schulte, P.J., Gibson, A.C. and Nobel, P.S. 1 989. Water flow in vessels with simple or compound perforation plates. Annals of Botany, 64: 1 7 1 - 1 78 . Schultz, H.R. and Matthews, M.A. 1 993 . Xylem development and hydraulic conductance in sun and shade shoots of grapevine ( Vitis vinifera L.): evidence that low light uncouples water transport capacity from leaf area. Planta, 1 90: 393-406. Scott, KoJ. and Roberts, E.A. 1 967. Breakdown in Jonathan and Delicious apples in relation to weight lost during cool storage. Australian Journal of Experimental Agriculture and Animal Husbandry, 7: 87-90. Scott, KoJ. and Roberts, E.A. 1 968. The importance of weight loss in reducing breakdown of Jonathan apples. Australian Journal of Experimental Agriculture and Animal Husbandry, 8 : 3 77-3 80. Sexton, R and Roberts, J.A. 1 982. Cell biology of abscission. Annual Review of Plant Physiology, 3 3 : 1 33 - 1 62. Sharples, RO. 1 964. The effects of fruit thinning on the development of Cox ' s Orange Pippin apples in relation to the incidence of storage disorders. Journal o f Horticultural Science, 39: 224-23 5 . Sharples, RO. 1 968. The structure and composition of apples i n relation to storage quality. Reports of the East Malling Research Station for 1 967. 1 85 - 1 89. Siddiqui, S. and Bangerth, F. 1 996. The effect of calcium infiltration on structural changes in cell walls o f stored apples. Journal of Horticultural Science, 7 1 : 703-708. References 1 54 Simon, E.W. 1 977. Leakage from fruit cells in water. Journal of Experimental Botany, 2 8 : 1 1 47- 1 1 52. Simons, R.K. 1 965. Tissue development in the apple associated with embryo abortion. Proceedings of the American Society for Horticultural Science, 87: 5 5-65 . Simons, R.K. 1 974. Placental tissue and ovule development in ' Lodi ' apple. Journal of the American Society for Horticultural Science, 99: 69-73. Simons, R.K. and Chu, M.C. 1 968. Ovule development in the apple as reJated to morphological and anatomical variation in supporting tissues. Proceedings of the American Society for Horticultural Science, 92: 3 7-49. Skene, D.S. 1 966. The distribution of growth and cell division in the fruit of Cox 's Orange Pippin. Annals of Botany, 30: 493-5 1 2 . Smith, W.H. 1 940. The histological structure o f the flesh o f the apple i n relation to growth and senescence. Journal of Pomology and Horticultural Science, 1 8 : 249-260. Smith, W.H. 1 950. Cell-multiplication and cell-enlargement in the development of the flesh of the apple fruit. Annals of Botany, 1 4 : 23-38. Solomos, T. 1 987. Principles of gas exchange in bulky plant tissues. HortScience, 22: 766-77 1 . Soudain, P. and Phan Phuc, A. 1 979. La difussion des gaz dans les tissus vegetaux en rapport avec la structure des organes massifs. In : Perspectives Nouvelles dans la Conservation des Fruits et Legumes Frais. Seminaire International, Centre de Recherches en Sciences Appliquees it } ' Alimentation, L'Universite du Quebec cl Montreal, Canada. Pp 67-86. References Stahly, E.A. 1 986. Time of application of calcium sprays to increase fruit calcium and reduce fruit pitting of apples sprayed with TIBA. HortScience, 2 1 : 95-96. 1 5 5 Stahly, E.A. and Benson, N .R. 1 982. Seasonal accumulation of calcium and potassium in the cortex of ' Golden Delicious' apple fruit sprayed with 2,3,5-Triiodobenzoic acid. HortScience, 1 7 : 78 1 -783. Stott, K.G. 1 972. Pollen germination and pollen-tube characteristics in a range of apple cultivars. Journal of Horticultural Science, 47 : 1 9 1 - 1 98. Stout, A.B. 1 936. Seedlessness in grapes. New York State Agricultural Experiment Station (Geneva) Technical Bulletin, 238. Stow, J. 1 989. The involvement of calcium ions in maintenance of apple fruit tissue structure. Journal of Experimental Botany, 40: 1 05 3 - 1 05 7 . Swamp, R . , Marchant, A . and Bennett, M.J. 2000. Auxin transport: providing a sense o f direction during plant development. Biochemical Society Transactions, 2 8 : 48 1 -485. Tamas, LA. 1 995. Hormonal regulation of apical dominance. In: Plant Hormones. Physiology, Biochemistry and Molecular Biology, 2nd edition. P J . Davies (ed). Kluwer Academic Publishers, Dordrecht, Boston, London. Pp 5 72-597. Tetley, U. 1 930. A study of the anatomical development of the apple and some observations on the "pectic constituents" of the cell-walls. Journal of the Pomology and Horticultural Science, 8 : 1 5 3 - 1 72. Theiler, R. and Coombe, B.G. 1 985. Influence of berry growth and growth regulators on the development of grape peduncles in Vitis vinifera L. Vitis, 24: 1 - 1 1 . Thompson, D.W. 1 966. On Growth and Form. J.T. Bonner (ed). Cambridge University Press. References 1 56 Thomson, K-S ., Hertel, R., Miiller, S. and Tavares, J. 1 973 . 1 -N-Naphthylphthalamic acid and 2,3,5-Triiodobenzoic acid. In-vitro binding to particulate cell fractions and action on auxin transport in corn coleoptiles. Planta, 1 09 : 3 3 7-352. Tomala, K. and Dilley, D.R 1 989. Calcium content of Mclntosh and Spartan is influenced by the number of seeds per fruit. Proceedings of the Fifth International Controlled Atmosphere Research Conference. Wenatchee, Washington, USA, 1 4- 1 6 June, 1 989. Pp 75-8 1 . Tomala, K. and Dilley, D.R 1 990. Some factors influencing the calcium level in apple fruits. Acta Horticulturae, 274: 48 1 -487. Tromp, J. 1 979. Seasonal variations in the composition of xylem sap of apple with respect to K, Ca, Mg, and N. Zeitschrift fuer Pflanzenphysiologie, 94: 1 89- 1 94. Tromp. J. 1 984. Diurnal fruit shrinkage in apple as affected by leaf water potential and vapour pressure deficit of the air. Scientia Horticulturae, 22: 8 1 -87. Tucker, M.L., Sexton, R , del Campillo, E. and Lewis, L.N. 1 988. Bean abscission cellulase. Plant Physiology, 88: 1 25 7- 1 262. Tukey, L.D. 1 964. A linear electronic device for continuous measurement and recording of fruit enlargement and contraction. Proceedings of the American Society for Horticultural Science, 84: 653-660. Tukey, H.B. and Young, J.O. 1 943 . Gross morphology and histology of developing fruit of the apple. Botanical Gazette, 1 04: 3-25. Tyree, M.T. and Zimmermann, M.H. 1 97 1 . The theory and practice of measuring transport coefficients and sap flow in the xylem of red maple stems (Acer rubrum). Journal of Experimental Botany, 22: 1 - 1 8. References 1 5 7 Tyree, M.T. and Ewers, F.W. 1 99 1 . The hydraulic architecture of trees and other woody plants. New Phytologist, 1 1 9: 345-360. van de Geijn, S.C., Petit, C.M. and Roelofsen, H. 1 979. Measurement o f the cation exchange capacity of the transport system in intact plant stems. Methodology and preliminary results. Communications in Soil Science and Plant Analysis, 1 0: 225-236. Varga, A. and Bruinsma, J . 1 976. Roles of seeds and auxins in tomato fruit growth. Zeitschrift fuer Pflanzenphysiologie, 80: 95- 1 04. Vincent, J.F.V. 1 989. Relationship between density and stiffness of apple flesh. Journal of the Science of Food and Agriculture, 47: 443-462 . Vincent, J.F.V. 1 992. Plants. In: Biomechanics - Materials. A Practical Approach. J.F.V. Vincent (ed). Oxford University Press, Oxford, New York, Tokyo. Pp 1 65- 1 9 1 . Visser, T . and Verhaegh, J.J. 1 980. Pollen and pollination experiments. II. The influence of the first pollination on the effectiveness of the second one in apple. Euphytica, 29: 385-390. Visser, T. and Marcucci, M.C. 1 98 3 . Pollen and pollination experiments. IX. The pioneer pollen effect in apple and pear related to the interval between pollinations and the temperature. Euphytica, 3 2 : 703-709. Visser, T., Verhaegh, J.J., Marcucci, M.C. and Uijtewaal, B .A. 1 98 3 . Pollen and pollination experiments. VIII. The effect of successive pollinations with compatible and self-incompatible pollen in apple and pear. Euphytica, 3 2 : 5 7-64. Volz, R.K., Ferguson, LB., Hewett, E.W. and Woolley, D.J. 1 994. Wood age and leaf area influence fruit size and mineral composition of apple fruit. Journal of Horticultural Science, 69: 3 85-395. References 1 5 8 Volz, RK., Tustin, S.D. and Ferguson, LB. 1 996a. Mineral accumulation in apple fruit as affected by spur leaves. Scientia Horticulturae, 6 5 : 1 5 1 - 1 6 1 . Volz, RK., Tustin, S.D. and Ferguson, LB. 1 996b. Pollination effects on fruit mineral composition, seeds and cropping characteristics of 'Braeburn' apple trees. Scientia Horticuiturae, 66: 1 69- 1 80. Volz, R.K., Harker, F.R, Hallett, I.C. and Lang, A. 2002. Development of texture in apple fruit - a biophysical perspective. Acta Horticulturae (in press). Ward, D. and Marini, RP. 1 999. Growth and development of young apple fruits following applications of ethephon plus carbaryl for thinning. HortScience, 34: 1 05 7 - 1 059. Ward, D.L., Beers, E.P., Byers, RE. and Marini, RP. 1 999. Cutting apple fruits induces cellulase activity in the abscission zone. HortScience, 34: 601 -603 . Watkins, C.B., Burrneister, D.M., EIgar, H J . and Liu, F.W. 1 997. A comparison of two carbon dioxide-related injuries of apple fruit. Proceedings of the 7th International ConrolIed Atmosphere Research Conference. University of California, Davis, California, USA, 1 3- 1 8 July 1 997, Vol. 2. E J . Mitcharn (ed). Pp 1 1 9- 1 24. Way, RD. 1 978. Pollination and fruit set of fruit crops. New York's Food and Life Sciences Bulletin, 76: 1 -9 . Weinbaum, S.A. , DeJong, T.M. and Maki, J. 200 1 . Reassessment of seed influence on return bloom and fruit growth in 'Bartlett' pear. HortScience, 3 6 : 295-297. Wertheim, S.J. 1 97 1 . The drop of flowers and fruits in apple, with special reference to the j une drop of Cox ' s Orange Pippin and its control with growth regulators. Mededelingen Landbouwhogeschool Wageningen, 7 1 - 1 7 : 1 -73. References 1 59 Wertheim, S.J. 1 99 1 . Malus cv. Baskatong as an indicator of pollen spread in intensive apple orchards. Journal of Horticultural Science, 66: 635-642 . Westwood, M.N., Batj er, L.P. and Billingsley, H.D. 1 967. Cell size, cell number and fruit density of apples as related to fruit size, position in cluster and thinning method. Proceedings of the American Society for Horticultural Science, 9 1 : 5 1 -62. Wilkinson, B.G. 1 965. Some effects of storage under different conditions of humidity on the physical properties of apples. Journal of Horticultural Science, 40: 58-65. Wilkinson, B.G. 1 968. Mineral composition of apples. IX. Uptake of calcium by the fruit. Journal of the Science of Food and Agriculture, 1 9 : 646-647. Wilkinson, B . G. 1 970. The effect of evaporation on storage disorders of apples. Reports of the East Malling Research Station for 1 969. 1 2 5- 1 27. Williams, RR 1 970. Factors affecting pollination in fruit trees. In: Physiology of Tree Crops. L.C. Luckwill and C.V. Cutting (eds). Academic Press, London, New York. Pp 1 93-207. Williams, R.R and Smith, B.D. 1 967. Observations on factors influencing the effective distance of pollinator trees in 1 966. Annual Report of the Long Ashton Agricultural and Horticultural Research Station for 1 966: 1 26- 1 34. Williams, RR, Brain, P., Church, RM. and Flook, V.A. 1 984. Flower receptivity, pollen transfer and fruit set variations during a single flowering period of Cox's Orange Pippin apple. Journal of Horticultural Science, 59: 337-347. Wills, R.B.H. and Scott, KJ. 1 972. Methods of increasing water loss from apples to reduce low temperature breakdown. Journal of Horticultural Science, 47: 349-3 5 5 . References 1 60 Wolswinkel, P., Ammerlaan, A. and Koerselman-Kooij, J. 1 999. Changes in the function of the xylem pathway during the development of fleshy fruits: implications for phloem transport, water movement and fruit growth. International Conference on Assimilate Transport and Partitioning, Newcastle, NSW, Australia, 1 5 -20 August 1 999. Scientific Programme and Abstracts Book, ISSN 0 7259 1 0690, Abstract No 1 0-34. Pp 1 64. Yamaki, S. and lno, M. 1 992. Alteration of cellular compartmentation and membrane permeability to sugars in immature and mature apple fruit. Journal of the American Society for Horticultural Science, 1 1 7 : 95 1 -954. Yearsley, C.W., B anks, N.H., Ganesh, S. and Cleland, DJ. 1 996. Determination of lower oxygen limits for apple fruit. Postharvest Biology and Technology, 8 : 95- 1 09. Yearsley, C.W., Banks, N.H. and Ganesh, S. 1 997a. Temperature effects on the internal lower oxygen limits of apple fruit. Postharvest Biology and Technology, 1 1 : 73-83 . Yearsley, C.W., Banks, N.H. and Ganesh, S . 1 997b. Effect of carbon dioxide on the internal lower oxygen limits of apple fruit. Postharvest Biology and Technology, 1 2 : 1 - l 3 . Zimmermann, M.H. 1 98 3 . Xylem Structure and the Ascent o f Sap. Springer-Verlag, Berlin, Heidelberg, New York, Tokyo. Zimmermann, M.H. and Milburn, l .A. 1 982. Transport and storage of water. In: Physiological Plant Ecology 11. Encyclopaedia of Plant Physiology, New Series Vol. 1 2B . O.L. Lange, P.S. Nobel, C.B. Osmond and H.Ziegler (eds). Springer-Verlag, Berlin, Heidelberg, New York. Pp 1 3 5 - 1 5 1 .