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. DIFFERENTIATING APPLE SPORTS BY POLLEN ULTRASTRUCTURE A thesis presented in partial fulfilment of the requirements for the degree of Master of Horticultural Science at Massey University Alastair John Currie 1995 11 ABSTRACT Cultivars are plants that form distinct, uniform and stable phenotypes. New cultivars can be protected by Plant Variety Rights (PVR) which allow the owner exclusive rights to the propagation and sale of the plant material. Current PVR identification methods for apple cultivars require detailed records of tree, flower and fruit characteristics to differentiate the new cultivars from known cultivars. This method is c;low, expensive and unable to cope with the increasing numbers of sports. Biochemical identification methods such as isozymes, restriction fragment length polymerisation (RFLP), random amplified polymorphism DNAs (RAPD), and minisatellite probes, can quickly and objectively differentiate cultivars, but cannot differentiate apple sports. Previous research suggested that pollen ultrastructure could be an alternative method for plant identification. This thesis is concerned with the development of a technique to differentiate apple sports using pollen exine patterns. Scanning electron microscopy was used to capture images of the apple pollen grain and the exine surface. A digital image analysis algorithm was developed to extract quantitative data from the pollen grain dimensions and pore characteristics, and a Fast Fourier transform extracted quantitative data from the ridge patterns. Statistical methods were applied to the data to differentiate the sports. Pollen harvested from apple flowers in the spring were wider than pollen harvested from flowers forced out of season under artificial conditions. Significant differences between trees were found for pollen grain length:width ratio, percent pore coverage, pore area and pore length but further research is required. However, apple cultivars types 'Red Delicious' and 'Gala' were successfully differentiated by pore and pollen grain variables, and 'Aversang' and 'Ultrared' sports of Red Delicious', and 'Splenola' and 'Galalea' sports of 'Gala' were successfully differentiated by exine ridge patterns and pollen grain measurements. Differentiation of apple sports by pollen requires further development but may be one of the only quick, objective identification methods that can differemiate sports. Sport lll differentiation would greatly aid PVR establishment and enforcement. IV ACKNOWLEDGEMENTS My sincere thanks to my supervisors. Dr. G. S. Lawes, senior lecturer in the Plant Science Department of Massey University, supplied guidance and encouragement throughout the term of my project. Dr. D. Bailey, director of the Image Analysis Unit at Massey University, explained the concepts of image analysis and provided expert help in developing the image analysis system. Dr. D. Noiton, science group manager for the Plant Improvement Division at the Horticulture and Food Research Institute of New Zealand (HortResearch), Havelock North, initiated the project, supplied encouragement and greatly helped with the writing of the thesis. I would like to thank ENZA for their generous contribution in the funding of this project, and the Leonard Condell trust scholarship for financial assistance with living expenses. I wish to thank Doug Hopcroft and Raymond ('Crunch') Bennett of the Electron Microscope Unit, HortResearch, Palmerston North. Doug Hopcroft for his advice, explanations and work with the scanning electron microscopy, and Crunch for his work in developing the micrographs. I am gratefully indebted to Dr. S. Ganeshanandam of the Statistics Department, Massey University, for his expert advice on the multivariate statistical analysis. Peter Alspach of HortResearch, Riwaka, provided invaluable assistance with the experimental design. I would also like to thank Cath Snelling and Sue Knowles, technical staff of HortResearch, for collecting budwood and pollen at Havelock North and Clyde Research centres (respectively); and to the field staff of HortResearch for their part in maintaining the trials. My appreciation is expressed to Kathie Brown for the help with proof reading the manuscript. This degree could not have been undertaken without the financial ._ support and v encouragement of my parents, Ted and Barbara, and sister, Margaret. Finally, a special thanks to Digna Sandoval for her friendship, generosity and support. Vl TABLE OF CONTENTS ABSTRACT ............................ ... .......... . ............. 11 ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv TABLE OF CONTENTS ..... ..... . .. .... . . ... . .. . ................. v1 LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix LIST OF TABLES ........................ . . ............. .......... Xl 1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2. LITERATURE REVIEW ......................................... 4 2.1. New cultivars and plant variety rights . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.2. Plant identification techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.3. Pollen analysis and plant identification .................. .. . ... . 7 2.4. Image analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.5. Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.6. Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3. MATERIALS AND METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3 .1. Pollen source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.2. Pollen collection .................................. ....... . 16 3.3. Pollen preparation for light and scanning electron microscopy . . . . . . 17 3.3.1. Air drying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 3.3.2. Acetolysis ................................... .... 18 3.3.3. Hot Potassium Hydroxide wash .. .. .. .. . ...... ....... 18 3.3.4. Rehydration ...................................... 18 Vll 3.3.5. Scanning electron microscopy ....................... 18 3.4. Scanning electron microscopy .............................. . 19 3.5. Pollen characteristics and analysis .................... . ...... 20 3.6. Image analysis ........................................... 21 3.6.1. Software and Hardware Systems ..................... 21 3.6.2. Image capture .................................... 22 3.6.3. VIPS Commands and Sub-routines ................... 23 3. 7. Comparison between spring flowers and forced flowers . . . . . . . . . . . 23 3.8. Statistical methodology ........................... . ........ 24 3.8.1. Experimental design ............................... 24 3.8.2. Data collected .................................... 24 3.8.3. Analysis ......................................... 25 4. RES UL TS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.1. Pollen collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.2. Pollen preparation for scanning electron microscopy ...... . ... . . . 27 4.2.1. Air dried pollen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.2.2. Acetolysed pollen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.2.3. Hot potassium hydroxide solution treated pollen ......... 28 4.2.4. Rehydrated pollen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4.3. Scanning electron microscopy ............................... 29 4.4. Image analysis algorithm ............................. . ..... 35 4.4.1. Whole pollen grain measurements .......... .. .... . ... 36 4.4.2. Fourier analysis of the ridge patterns ...... . .. . ........ 39 4.4.3. Analysis of pores in the exine ........................ 46 4.5. Comparison between spring flowers and forced flowers .. . ........ 51 4.6. Statistical analysis ........................................ 52 4.6.1. Nested factorial multivariate analysis of variance (MANOV A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.6.2. Univariate analysis of variance on non-ridge data . . . . . . . . 52 4.6.3. Canonical Variate Analysis (CVA) .................... 56 4.6.4. Discriminant Analysis for group classification . . . . . . . . . . . 64 viii 5. DISCUSSION .................................................. 59 6. CONCLUSION ................................................. 68 7. REFERENCES ................................................. 70 APPENDIX A: Summary of VIPS5 commands, functionals and utilities . . . . . . . 92 APPENDIX B: Plant Variety Rights evaluation form ..................... 102 ix LIST OF FIGURES Fig. 1. Fig. 2. Fig. 3. Fig. 4. Fig. 5. Fig. 6. Fig. 7. Fig. 8. Fig. 9. Fig. 10. Fig. 11. Fig. 12. Fig. 13. Fig. 14. Fig. 15. A diagram demonstrating the distortion of measurements such as pollen grain length if the object is rotated in three dimensions and then viewed in two dimensions (as in a micrograph). . . . . . . . . 19 A cross section of the pollen surface structure. . . . . . . . . . . . . . . . . 20 A schematic diagram of the hardware set-up for image analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Scanning electron micrograph of air-dried 'Red Delicious' apple pollen which was stored for 12 months in a freezer. Most ....... 29 Light micrograph of freshly prepared air-dried 'Red Delicious' apple pollen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Light micrograph of acetolysed 'A versang' apple pollen suspended in water. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Light micrograph of acetolysed 'A versang' apple pollen dried and - sprinkled on double-sided tape. . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Light micrograph of 'A versang' apple pollen prepared with a hot potassium hydroxide ( 10%) wash and suspended in water. . . . . . . . 31 Light micrograph of rehydrated and fixed 'A versang' apple pollen suspended in water. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Light micrograph of rehydrated and fixed 'A versang' apple pollen, dried and mounted on double-sided tape. . . . . . . . . . . . . . . 32 Scanning electron micrograph of air-dried 'A versang' apple pollen ................................................. 33 Scanning electron micrograph of rehydrated and fixed 'A versang' apple pollen ............................................ 33 Flow diagram outlining the three -sections and general structure of the image analysis algorithm. . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 The processing sequence of a 512 x 512 pixel image from a scanning electron micrograph of air dried 'A versang' pollen. . . . . . 37 Images of the central region of an 'A versang' pollen grain between the germinal furrows prepared for the Fourier analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Fig. 16. Fig. 17. Fig. 18. Fig. 19. Fig. 20. Fig. 21. Fig. 22. Fig. 23. Fig. 24. x The Fast Fourier transform of the windowed exine pattern image. . . .. .. . . .................. .. ......................... 42 The Fourier image after mapping from a radial plot to a ยท rectangular plot. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 The Fast Fourier transform image of 'A versang' exine ridge patterns processed and Gaussian smoothed, ready to extract ridge data ...... . ...... . .. . ..... . ............. . .......... . .. 44 The processed Fourier image after applying a simple threshold to remove low frequency and harmonic frequency information. This isolated information on the strongest ridge patterns of the pollen exine ...... . . . . . .. . . . .. . ........... . ........... . . . . . .. 44 The apple pollen exine surface in the central area of the pollen grain between the germinal furrows. The shaded area was selected for enlargement and pore dimensions analysis. . . . . . . . . . 48 The central area of the pollen exine surface enlarged for pore analysis ....... . ...... . .. .... . ......................... 48 Steps in the processing and measurement of pores in the exine of an apple pollen grain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Canonical scores for the first two canonical variables calculated from individual pollen grain data and plotted to show the separation achieved between the apple sports of 'Red Delicious' and 'Gala'. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 The canonical scores for the first two canonical variables calculated from averaged and weighted data and plotted to show the separation between the sports of 'Red Delicious' and 'Gala'. . . . 63 Xl LIST OF TABLES Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Comparison between spring pollen and forced pollen yield components for apple cultivar 'Harrold Red'. . . . . . . . . . . . . . . . . . . 51 Mean pollen grain dimensions from spring and forced flowers of the apple cultivars 'Red Delicious' and 'Gala' .................. 54 Mean pollen grain and pore dimensions for bulked pollen from 'Red Delicious' and 'Gala' cultivar types. . . . . . . . . . . . . . . . . . . . . . 55 P values of Mahalanobis Distance for Squared Distance to apple sports of 'Red Delicious' and 'Gala' .......................... 56 Top ten DRCs for the first canonical variable (CANl) used to select the most important variables for the separation of the apple sports groups in CAN 1 using pollen grain characteristics. . . . . . . . 57 Top ten DRCs for the second canonical variable (CAN2) to select the most important variables for the separation of the apple sports groups in CAN2 using pollen grain characteristics. . . . . . . . 59 P values of Mahalanobis distance for squared distance to apple sport ................................................. 60 Mean canonical scores and significant groupings for sports of Red Delicious and Gala. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 1 1. INTRODUCTION Cultivars are cultivated varieties of a crop and are the product of change to the genome, for example sexual reproduction or mutation. The cultivars that are the products of mutations are called sports and can either occur naturally or be induced with a mutagenic agent (for example radiation). In Brooks and Olmo's Fruit and Nut register there are approximately 2500 cultivars listed for the period between 1962 and 1991. Over 12% of these cultivars are sports. Sports form an even more significant proportion of apple cultivars; nearly 31 % of the commercial cultivars in the Brooks and Olmo Register 1962-1991 are sports (Brooks and Olmo, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969, 1970, 1971, 1972, 1973, 1974, 1975, 1978, 1982, 1983, 1984, 1991). If it can be demonstrated that a cultivar is new, distinct, homogenous, and genetically stable then the developer may apply for a plant patent. In New Zealand new varieties have only been covered by plant patent laws since 1975 (Whitmore, 1992). The current laws dealing with plant patents are in the Plant Variety Rights (PVR) Act of 1987 which is based on the International Union for the Protection of New Varieties of Plants (UPOV) convention. The PVR office has to esta~lish that each applicant for PVR status meets the criteria of distinctness, homogeneity and genetic stability. Applicant plants are planted in the field beside other known, similar cultivars and are observed over several seasons to remove the influence of environment. This can be quite a large undertaking. The Brooks and Olmo Register recorded 57 sports of Delicious between 1962 and 1991 (Brooks and Olmo, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969, 1970, 1971, 1972, 1973, 1974, 1975, 1978, 1982, 1983, 1984, 1991). Once the PVR is established the developer usually licences nurseries to propagate the plant material and charges a royalty on each tree sold. The produce of the new cultivar can also be protected by marketing the fruit under a trade mark which can be licensed out as well (Selby, 1995). Over the years there has been a move away from government backed research for the public good and a move towards research for commercial returns which has meant an increase in interest in developing and patenting cultivars for financial returns. In New Zealand, apples are an important crop. The annual return from fresh apple Chapter]. Introduction 2 exports was about $346.3 million in 1992/3, which represented more than 38% of the total export fruit sales (Fruit Research Council, 1994). Many cultivars of apple are used for export and each cultivar has minimum thresholds for colour, size and other quality factors. Mutations in the genome can enhance the quality of the fruit, and increase the value of the crop. The New Zealand Apple and Pear Marketing Board (NZAP1\1B) encourages the discovery of naturally occurring, new, improved sports. Recently there has been an increase in the numbers of sports as applicants for PVR status (Whitmore, 1992) and this may either be a reflection of the NZAPMB encouragement, the opportunity for financial gains from PVR, or the apple crop may be prone to mutation. The numbers of cultivars submitted with little morphological or agronomic difference from established cultivars is a major concern for the PVR office (Whitmore, 1992). Plants can be identified or differentiated by differences in agronomy, morphology, or biochemistry. Field observation of morphology has been adequate for most apple cultivars but requires the use of a large area to grow the standards as well as the applicant plant, and an expert to record the large amounts of data (PVR evaluation form, appendix B). Although most of the measurements are objective, some like colour require judgement. Unfortunately biochemical markers do not, so far, offer appropriate solutions to differentiate sports. DNA techniques like polymerase chain reaction (PCR), restriction fragment length polymerisation (RFLP), and minisatellite probes do not show sufficient polymorphism to differentiate apple sports. Isozyme techniques can differentiate some sports, but not all, and isozymes are sensitive to the environment. Other simple biochemical tests can be applied in specific cases, for example phenol test for wheat, but cannot be applied to all crops. There is a need for a plant identification method for PVR that will easily and objectively differentiate sports. The aim of this project was to develop a system that would differentiate apple sports using pollen grain ultrastructure. Scanning electron microscopy was used to capture the image and a new digital image analysis algorithm was developed to extract data from the image. Statistical methods were then applied to the data to differentiate the sports. This method could further be used to assist with the establishment of PVR status as well Chapter 1. Introduction 3 as identifying cultivars and sports. Although the technique was developed on apples, it could also be applied to other crops.