HONEY BEE POLLINATION REQUIREMENTS OF HYBRID CUCUMBERS CUCUMIS SATIVUS L. Thesis for the Degree of M. S. MICHIGAN STATE UNIVERSITY LAWRENCE JOHN CONNOR 1969 ABSTRACT HONEY BEE POLLINATION REQUIREMENTS OF HYBRID CUCUMBERS CUCUMIS SATIVUS L. By Lawrence John Connor Pickling cucumbers grown in Michigan relied almost entirely upon honey bees for pollination, with very few bumble bees and other native bees participating. Pistillate and staminate cucumber flowers produced a quality of nectar that was relatively attractive to bees. Honey bees did not gather pollen in the field but did gather small amounts in the greenhouse. The number of ovules in an ovary increased with size of the ovary. Pistillate flowers produced near the crown (root end) of the plant were shorter and developed into shorter fruit with lower seed numbers. The removal of one lobe of the stigma of a pistillate flower at the time of pollination did not prevent seed formation in the three carpels of the fruit. This fact and morphological evidence indicated that pollen tubes could penetrate from any lobe of the stigma to any of the 3 carpels. Parthenocarpic fruit formation of standard pickling varieties was observed in the field and in the greenhouse during the fall and Spring months. Shape of parthenocarpic fruit ranged from the hollow crooked types to those indistinguishable from fruit fully pollinated. The num- ber of seeds found in perfectly shaped fruit varied from one to 520 seeds, Lawrence John Connor with 300 to #00 being most frequent. The first one to 3 fruit forming anywhere on a vine usually completely inhibited the development of sub- sequent fruit. Plants grown in soil of low fertility reached bloom more slowly, developed fewer flowers, set a higher percentage of pistil- late flowers and had more short fruit than plants in high fertility plots. Honey bee visits to cucumbers began around 8:30 a.m. EDT at temperatures around 62°F, although flights were not abundant until the temperature reached 70°F. The first visit to a cucumber flower frequently lasted 36 to 39 seconds, but decreased to 8 to II seconds for the hth to l2th visits. Length of the first visit varied with the time of day as follows: 8 to 9 a.m. visits typically lasted 9-ll seconds; ID to II a.m. typically about 30 seconds; ll a.m. to 6 p.m. 38 to #5 se- conds. Bees were most active from l0 a.m. to 3 p.m. with peak flights around noon. Fruit set was most successful during this period. The number of visits per flower per hour could be increased by adding extra numbers of colonies per acre. Cold temperature resulted in few flights in the morning while warm temperature resulted in numerous morning flights. Single visits of honey bees to pistillate flowers frequently produced well-shaped cucumbers. However, the number of seeds increased with additional visits. Data indicated that at least IO bee visits were needed to insure pollination under the environmental variables. The number of seeds in fruit increased with total length of accumulated visits, A staminatezpistillate flower ratio of l:2 provided better Lawrence John Connor fruit yield than either 2.7:], l.2:l, or 0.03:l. An S:P ratio of 1:2 produced earlier fruit with more short, non-curved, non-constricted cu- cumbers than the other blends. The S:P ratio was found to be very sen- sitive to environmental factors, especially light, an increase of which resulted in increased staminate flower production. Cucumber production on gynoecious plants decreased on a gradient with distance from the pol- len source, with yield being satisfactory up to 25 feet. Colony rotation from different fieldswas found to have little effect on bee visits to cucumbers. Delaying the introduction of bees to test plots resulted in more fruit with larger length/width ratios and more seeds. As an overall recommendation, it is suggested that one colony of bees containing a minimum of h0,000 workers be supplied for every 50,000 cucumber plants in a field. HONEY BEE POLLINATION REQUIREMENTS OF HYBRID CUCUMBERS CUCUMIS SATIVUS L. By Lawrence John Connor A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Entomology I969 ACKNOWLEDGMENTS My sincere thanks and deep appreciation are extended to Dr. E. C. Martin, who deserves total credit for this pollination project. Through long hours in the field, the office, the laboratory and even while on sabbatical, he has directed this project with constant encouragment and tactful criticism of the author's work. Using patience and well timed suggestions, he has supervised this project through two years of labor. I am grateful to USDA Agricultural Research Service, Entomology Research Division, Apicultural Research Branch for financial and advisory support under a Cooperative Agreement arrangement. In particular, I wish .to thank Mr. S. E. McGregor and Dr. Floyd Moeller of the Apiculture Research Branch, USDA for suggestions and ideas which were generously offered. I wish to thank those individuals who have served on my guidance committee, Dr. Roger Hoopingarner, Dr. Roland Fisher, Dr. Larry Baker and Dr. Clinton E. Peterson. Dr. Peterson is thanked for the original suggestion and for presenting ideas and materials that have proven very useful in the project. Special appreciation is given for his generosity in supplying seed. Thanks are extended to Dr. Gordon Guyer, Chairman of the Department of Entomology, who provided field, greenhouse and laboratory facilities for the project and has shown great enthusiasm for the significance of pollination in crop production. To Dr. Clark Nicklow of the MSU Horticulture Department, I extend appreciation for help in precision planting of the I968 research plots. To Art Wells I extend appreciation for help in preparing field plots and furnishing fertilizer and other necessities, To Clarence and Sally Collison, who collaborated in this adventure, I extend thanks for covering for me when the schedule became tight, when I needed extra help, and for listening to me dream, complain and speculate. To my summer help, who have seen more flowers, cucumbers, cucumber seeds, and honey bees than any one of them is willing to admit, I extend my thanks. Special thanks go to Curtis Jensen, who helped keep the field plots in order during I967, and Calvin Erickson, for his dependability with the stopwatch, the hand counter, the key punch, the desk computer, and the art materials. My greatest thanks to to my part-time cucumber seed counter, full-time consultant, usual lunch partner, part-time psychoanalysist, and full'time wife, Catherine, for all that she has had to put up with during our marriage and the time previous, regarding this thesis. Appreciation is extended to Thanin Taychachaiwongse for typing the final draft. Finally, I thank my family and friends, who have offered their support during this research program, even if it meant graciously accepting gift cucumbers when they already had enough pickled to last ten years. It seemed a shame to bury all those well-pollinated cucumbers. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION FLOWER ANATOMY AND PLANT PHYSIOLOGY OF THE CUCUMBER AS RELATED TO FLOWER FERTILIZATION AND FRUIT DEVELOPMENT Flower anatomy . . Staminate flower structure . Pistillate flower structure Number of ovules . . Materials and methods Results and Discussion . Pollination process. Fertilization process . . Path of the pollen tube - PolIen placement on stigma Materials and Methods Results . Time from pollination to fertilization . Literature . Materials and Methods Results Influence of the plant on fruit formation Plant age . Materials and Methods Results Node position . Materials and Methods Results Discussion . . Parthenocarpy in pickling cucumbers Literature . . Observations and Discussion Presence of inhibitive fruit . I Literature . Materials and Methods Results Discussion . . . Effect of fertilizer on set and shape of fruit . Literature . Materials and Methods Results. Page coax-t4:- 4-” II ll 13 l4 Ih IS IS l7 I7 18 18 l9 I9 20 2] 26 26 26 BI 32 32 33 3h 35 35 36 ho no 40 hi FLIGHT ACTIVITY OF HONEY BEES 0N PICKLING CUCUMBERS Literature . . Factors influencing bee flight activity . Honey bee behavior on cucurbits . Materials and Methods . Flight activity . . . . Bee visits and colonies per acre Length of individual bee visits Results . . Flight activity . . . . . Bee visits and colonies per acre Length of individual bee visits Discussion . Flight activity . . . . Bee visits and colonies per acre Length of individual bee visits . . Attractiveness of cucumber to honey bees FLOWER SET AND DEVELOPMENT AND SEED NUMBERS AS INFLUENCED BY THE NUMBER, LENGTH, AND TIME OF DAY OF BEE VISITS . . . . Literature. . Number of bee visits and length of visit Number of seeds and shape of cucumbers . Time of day and length of time for optimum pollin- ation of cucumbers Materials and Methods . . . Observations of individual flowers Variable eXposure of honey bees to cucumber flowers Results and Discussion . Single flower observations . . . . . . . . Number of bee visits and percentage fruit set .............. Number of bee visits and fruit shape Number of bee visits and seed counts Length of bee visits and resulting seed counts . Time of day of pollination . Variable exposure of honey bees to flowers Minimum number of hours of exposure to pollination for maximum fruit set ... Time of peak pollination activity . ESSCENTAGE. STAMINATE FLOWERS OPTIMUM FOR FRUIT YIELD AND SEED Literature . . Materials and Methods . I967 Cage study . Page I968 Isolation plot study Staminate: pistillate flower ratios for different varieties Percentage fruit development per day of flowering. Results I967 Cage Study I968 Isolation plot study Staminate: pistillate flower ratios for different varieties Percentage fruit deveIOpment per day of flowering. Discussion . CUCUMBER POLLEN MOVEMENT BY HONEY BEES Literature . . Materials and Methods . . . I967 Gradient Study I968 Gradient Study Results . . . . . . . . . . I967 Gradient Study I968 Gradient Study Discussion . . . . . . . DELAYED POLLINATION AND EFFECT ON YIELD . Literature . . Materials and Methods . Preliminary investigation - August I967 Flower removal on different varieties Delayed pollination cage plots - I968. Results . Preliminary investigation - August I967 Flower removal on different varieties Delayed pollination cage plots - I968 CONCLUSIONS . . . . . . . . . . . . . ..... LITERATURE CITED APPENDIX vi Page 97 98 98 99 I02 I08 llh II6 II8 II8 II9 II9 I20 I2I I2I 122 I28 129 l29 I30 I30 I30 I3] I32 l30 I32 l32 I38 Ihh ISO Table LIST OF TABLES Correlation of ovary length at anthesis with ovules per longitudinal section . Effect on number of seeds formed per carpel after removing one stigmatic lobe from a pollinated flower . Time from pollination to fertilization . Effects of age of the cucumber plants on seeds per fruit Effect of age of the cucumber plant on length/width ratio Of mature fruit . . . . . ... ... . . . Effect of node position on cucumber shape. (Average length/width ratio of main vinerfruit)'. . Effect of node position on cucumber shape of fruity; from the main vine. (Grading percentages). . . Effect of node position on cucumber shape. (Average length/width ratio of lateral vine fruit) Effect of node position on shape of fruit from lateral vine. (Grading percentages) . . Influence of position on the vine on fruit set and shape . Influence of number of other fruit on fruit set and fruit shape Fertilizer level effect on fruit shape, constriction, curvature and length/width ratio . Effect of high and low fertilizer on flower set of cucumbers. Percentage of fruit deveIOpment from 25 pollinated flowers per day . Number of bee visits to IO flowers in thirty minutes, East Lansing, I967 . vii Page 19 22 2h 27 29 3O 3O 37 38 Al A3 #8 Table 20. 2|. 22. 23. 2h. 25. 26. 27. 28. 29. 30. 3I. 32. 33. Air temperature at the time of the first flights of the day on cucumbers . Weather observations for central Michigan during comparative counts of bee populations on three dates . Bee flight counts for Springport and East Lansing, Mich- igan, August 9, I967 . . . . . . . . . . . . Plant yield data from three fields for the first harvest duplicating mechanical harvesting Bee flight counts for Cedar Springs and East Lansing, August 23, I967. . Bee flights for Cedar Springs and East Lansing, August 25, I967 . . . . . . . . . . . . . . . . . . . Number of bee visit and average bee visit length . Effect of time of day upon length of initial bee visit Weather data for Lansing, Michigan and the average length of the first bee visit to cucumber flowers . Number of bee visits and percentage of fruit set Percentage fruit set and average seed counts from 2, IO, and 20 visits Number of bee visits and shape of fruit, I967 and I968 . Effect of the number of bee visits on the number of seeds per fruit. Average number of seeds obtained from three shapes of fruit. Fifteen fruit randomly selected in each class Total length of bee visits and resultant fruit development Total length of bee visit and average number of seeds. Effect of time of day of pollination on fruit development and the number of seeds per fruit. Controlled exposure of flowers to bees. Effect of time of day and length of exposure to pollination on fruit formation. Inclusion method . Controlled exposure of flowers to bees. 'Effect of time Of day and length of time of exposure to pollination on fruit formation. Exclusion method . viii Page 54 SS 56 57 58 59 6O 62 65 76 79 80 8I 82 83 8h 86 88 89 Table 3h. 35. 36. 37. 38. 39- #0. RI. #2. A3. #5. #6. #7. #8. A9. Time of day of pollination - Timed open study . Weight yield per acre and percentage pollinator I967. Based upon 50,000 plants per acre . . . . Average number of cucumbers produced per plant and percentage pollinator - I967 Percentage perfect, neck, nub, and crook shaped fruit from plots of different percentage pollinator I967. Staminate and pistillate flower production of different varieties and variety mixtures during the first l5 days of bloom . Staminate and pistillate flower production of different varieties and variety mixtures each day during the first l5 days of bloom Daily StaminatezPistillate flower ratio in different varieties and variety mixtures Dollar value of percentage pollinator tests in isolation plots -l968 . Number of fruit per plant of percentage pollinator tests in isolation plots - l968.. Shape of fruit from four different pollinator blend/variety treatments - I968 . Staminate and pistillate flowers produced during the first l5 days of bloom in different varieties grown in the greenhouse at different times of the year . Percentage of pistillate flowers which developed into fruit each day in percentage pollinator tests - I968 Pollen movement study- I967. Yield Data Pollen movement study - I968. Dollar value and average number of fruit per plant of 225 plants at increased distances from pollinator source Average seed counts from mature fruits grown at increasing distances from pollinator plot Percentage fruit set for 25 flowers tagged daily in one group planted with pollen source and second group planted without pollen source . ix Page 9] 99 IOO IOI IOZ l09 IIO III III II2 Il3 IIS I2I 12h l25 l27 Table 50. SI. 52. 53- SA. Delayed pollination - preliminary investigation August I967 . Effect of delayed pollination on yield of different varieties of cucumbers - July I968. Number of fruit per plant Effect on cucumber yield of delaying pollination by ex- cluding bees from the flowers. Number of cucumbers per plant . Effect of cucumber yield of delayed pollination by ex- cluding bees from the flowers. Dollar value per acre estimate based upon 50,000 plants per acre Effect of delayed pollination on cucumber shape. Number of fruit per shape per IOOO plants Page I33 I3h I35 I36 I37 Figure Ui-t’w I2. I3. lh. I5. I7. LIST OF FIGURES Overall view of staminate flower . Bithecal stamens . Monothecal stamen Pistillodia Cucumber pollen mounted in glycerine gel Fresh cucumber pollen as removed from anthers Cross section through ovary at anthesis. The infolded edges form the placentae from which the ovules arise . Longitudinal section of ovary at anthesis. Iodine shows the ovules and the conductive tissue forming the path of the pollen tube Overall view of petals, sepals, and ovary of pistillate flower . View of the three stigmatic lobes and five petals Pistillate flower with corolla and calyx removed, showing 3 stigmatic lobes and nectary Cross-section through pistillate flower showing pollen tube path, nectary tissue,rstigma and style . Section slightly above base of perianth tube showing trilobate stigma, conductive material in style, and hairs on inner wall Close-up of iodine-stained conductive material in general tri-radiate pattern. Section through base of perianth tube showing stained conductive material. Section through apex of ovary showing conductive material Redationship of plant age and the average number of seeds produced . xi Page I0 l0 l0 l6 l6 I6 23 Figure l8. I9. 20. 2|. 22. 23. 29. 25. 26. 27. 28. 29. 30. 3|. 32. 33. 3h. 35. 36. Relationship of plant age and average length/width ratio . . . . . . . . . . . . . . . . . . . . Relationship of node position and length/width ratio Relationship of position of a flower on the vine and the number of other fruit on the vine upon fruit deveIOpment Relationship» of fertilizer level and shape, constriction and curvature of cucunbers Relationship of time of day and flight pattern of honey bees on cucumber . . . . . . . . . . . . Bee VIsits per flower per hour - August 9, I967 . Bee visits per flower per hour - August 25, I967 Bee visits per flower per hour - August 23, I967 Relationship of the number of bee visits to a flower and the length of the visit . - . . Relationship of the time of day and the length of the first visit to a flower . . . . . . . Pistillate flowers bagged to prevent bee visits Nucleus hive (6-frame) in cage for cucumber pollination . I967 field research plots, including percentage pollinator plots and pollen movement studies . Greenhouse plants used for several parts of the research program . Cucumber with full seed development which received a full day of pollination Cucumber with only a few seeds resulting from one bee visit . Cucumber with a full seed development resulting from one bee visit . Cucumber with full seed development resulting from l0 bee visits Relationship of the time of day of pollination and the percent development . . . . . . xii Page 25 28 39 42 #9 SI 52 53 6| 63 77 77 77 77 78 78 78 78 87 Figure 37- 38. 39- #0. #I. #2. #3. ##. #5. #6. #7. #8. #9. 50. Relationship of the length of exposure to bee pollination and the degree of fruit deveIOpment . Relationship of one hour periods of exposure to bee pollination at different times of the day and the degree of fruit deveIOpment . . . . . . . Bee arriving on staminate flower Bee position for effective pollination visit Bee in greenhouse with a small amount of cucumber pollen on corbiculae. This was not observed in the field Bee removing nectar from staminate flower with part of corolla and calyx removed . Gynoecious F] hybrid plant Spartan Dawn grown without bees in field cage. Note the dominate pistillate flower production... Typical monoecious plant in greenhouse. Staminate flowers are produced at nearly every node . Influence of seed blend on staminate flower production. Influence of seed blend on pistillate flower production . Influence of seed blend on staminate:pistillate flower ratio Effect of staminate:pistillate flower ratio on the number of fruit per acre . Relationship of the closeness to the pollen source and the number of fruit per plant . The number of seeds produced in fruit at increasing distances from the pollen source xiii Page 90 92 93 93 93 93 95 95 IO# 105 106 107 123 126 INTRODUCTION During the past decade, the cucumber industry has experienced a series of rapid changes in production technology. The termination of Public Law #78 on December 31, I96# prevented Mexican laborers from entering the United States to harvest many crops, including cucumbers. Michigan growers were faced with the prospect of reduced production or finding an alternative to hand harvesting. Michigan has been able to meet the challenge by developing mechanical harvesters and gynoecious hybrid cucumber varieties. The development of mechanical cucumber harvesters introduced new problems for the cucumber grower. Plants were pulled completely from the ground and the fruit removed by a series of rollers. Plant populations were increased to allow for a single, profitable harvest (Stout.g£.§l., l96#; Morrison and Ries, 1967). During 1968, the majority of cucumber fields in Michigan were planted at p0pulations of 50 to 100 thousand plants per acre. (Schwarz, 1968). Gynoecious F] hybrid varieties are providing higher yields and more uniform growth. Hayes and Jones (1916), first reported hybrid vigor in cucumbers. Weigle (1956) indicated that considerable genetic diversity was needed in the parent varieties in order to obtain increased yields from hybrid pickling cucumbers. In spite of superior yields of hybrid cucumbers, little attention was paid to them because hybrid seed produced by hand Pollination was too expensive for commercial use. Peterson and WeigIe 2 (1958), described work in which they crossed a gynomonoecious variety from Korea named Shogoin (Pl 220860) with a pickling cucUmber selection (Wisconsin SMR 18), and then backcrossed gynoecious segregates which occurred in the F] to pickling type parents. Continued backcrossing and selection led to the conclusion that the gynoecious flowering habit could be incorporated into cucumber lines and such lines could be used to cross with selected monoecious lines to produce hybrid seed. The best inbred lines segregated up to 80 percent gynoecious plants and a high degree of heterosis appeared in the hybrid fruits. Peterson and Anhder (1960), showed that gibberellin A3 could be used to produce staminate flowers on predominately gynoecious lines and these authors produced homozygous gynoecious lines by successive self-pollination with pollen from gibberellin-induced staminate flowers. Self-pollination was accomplished by honey bees in a screen isolation cage. Peterson (1960) described a new gynoecious inbred line MSU 713-5 which produced no staminate flowers when tested in Michigan and Mexico. Peterson and DeZeeuw (1963) reported the development and release of the hybrid pickling cucumber, Spartan Dawn, which was resistant to cucumber mosaic and scab. Yields were higher than standard monoecious varieties with earlier fruiting. Peterson recommended that 10 percent of a monoecious pollinator be planted with the hybrid. He based this recommendation on a 20:1 ratio of staminate:pistillate flowers common in monoecious varieties and the fact that a 10 percent blend provided about 2 staminate flowers for each pistillate flower in a commercial planting. He pointed out that data were needed on pollen supply and insect pollinators. The rapid transition from low plant populations of # to 20 3 thousand plants per acre to high plant populations of #0,000 to over 200,000 plants per acre and the use of hybrids with more pistillate flowers vastly changed the pollination requirements of cucumbers. Honey bee pollination has been recognized as necessary for normal cucurbit production (Robinson, 1952; Nevkryta, I953; Tuljyenkova, 1955; Taylor, I955; Alex, I957; Markov and Romanchuk, I959; Verdieva and Ismailova, I960; Warren, 196l; Shahin, 1967; Williams and Kauffeld, I967). In practice, bee pollination was a minor problem for the grower due to low plant populations and multiple harvests. Now, with an average of one pistillate flower per plant per day in gynoecious hybrids, high plant populations may produce from #O to 200 thousand pistillate flowers per acre per day. As a comparison, one acre of 12 thousand monoecious plants might produce approximately 3,000 pistillate flowers per day. It may readily be seen that rapid and adequate pollination has now become an important, critical factor in the production of pickling cucumbers. The purpose of this research was to investigate major factors associated with the pollination of gynoecious hybrid cucumbers. Six areas for investigation were selected. Each of these areas is discussed in a separate chapter of the thesis. The six areas of interest include: 1. Flower anatomy and plant physiology of the cucumbers as related to flower fertilization and fruit development. 2. Flight activity of honey bees on cucumbers. 3. Fruit set and development and seed numbers as influenced by number, length and time of day of bee visits. #. Percentage staminate flowers optimum for fruit yield and seed counts. 5. Cucumber pollen movement by honey bees. 6. Delayed pollination and effect on yield. FLOWER ANATOMY AND PLANT PHYSIOLOGY OF THE CUCUMBER AS RELATED TO FLOWER FERTILIZATION AND FRUIT DEVELOPMENT FLOWER ANATOMY Staminate flower structure Cucumber flowers are basically 5-merous with staminate flowers having 5 petals, 5 sepals and 5 anther lobes arranged symetrically. However there are not five stamens, but three. Chakravarty (1958), concluded that five was the basic number of stamens and that two pair of stamens fused together to form two bithecal stamens leaving one mono- thecal stamen. Each theca contains two microsporangia, therefore each bithecal stamen contains four microsporangia and the monothecal stamen two microsporangia. Each stamen is folded in a horizontal S-shaped structure, twice folded upon itself (Figures 1-3). Using a different interpretation, Heimlich (1927) stated that there were two and one-half stamens. He summarized the deveIOpment of the structures in the following manner: during the development of the stamen, the primordium of each stamen develops a broad filament. Con- nective tissue forms at the apex and serves as a base for the theca. Each stamen develops a long extension of connective tissue called the Prolongation, usually five in number, which mass together at the center 0f the dorsal part of the stamen due to their angled inner surfaces. L. ‘. , Figure l. 1' Overall ....13 . .’ view of staminate flower. 't 3. Monothecal stamen. #. Pistillodia. 2. Bithecal V stamens. 6 At the base of the staminate flower there is a 3-lobed structure which is similar to the nectary of the pistillate flower. However, Heimlich considered this structure the undeveloped pistil, and called the structure the pistillodia (Figure #). The five, yellow, lobed petals and the five sharply pointed pubescent sepals are fused into the recep- tacle of the flower. The entire flower is situated on a slender pedun- cle. In the beginning of the anther deveIOpment, each theca appears as a mass of meristematic cells which eventually become two-lobed along its entire length. The theca consists of these two lobes separated by the tapetum and surrounded by papillate epidermal cells called trichomes (Heimlich, I927). The tapetum serves as the nutritive layer for the deveIOping microspore mother cells and subsequent microspores. It is almost completely digested during the growth and maturation of the pollen. At maturity, a central suture called the stomium splits open and exposes the pollen grains (Hayward, 1938). A cucumber pollen grain is relatively large, spheroid, sticky, and trilobate. The lobes are often viewed so as to give the pollen grain a somewhat triangular appearance (Figures 5 and 6). Pistillate flower structure Judson (l929a) presented a careful study of the development of the pistillate cucumber flower from the earliest visible stages. The ovary of the flower appears first as a tri-merous structure (scale-like) at the base of the receptacle of the flower. Growth is most active at the sides of the three scales, resulting in structures whose edges fold inward. Further growth causes the structures to curl away from each Figure 5. Cucumber pollen mounted in glycerine gel. Figure 6. Fresh cucumber pollen as removed from anthers. 8 other as they reach the center of the arc, and turn toward the center of the ovary. The edges of the scales do not fuse, but remain T-shaped. The infolded edges within each arc form the placentae from which the ovules arise. The three scale-like structures represent the three car- pels in the cucumber fruit (Figure 7). Three fused styles result from growth of the upper extremities of the carpels. On the inner surface of each style there is a groove, which represents the continuation of the T-shaped openings of the infold- ed edges of the carpels. The three stigmatic lobes are separate, each correSponding to a division of the style, and thus a carpel. The lobes of the stigmas are broad and grooved down the middle (Figures 9-11). The nectary tissue was found by Judson (l929a) to differentiate from tissue at the bottom of the corolla after the three carpel lobes had extended upwards (Figures II and 12). Growth took place within the perianth tube and circled the styles. The nectary cells are large with large nuclei and coarsely granular cytoplasm. Number of ovules Tillman (1906) observed that the carpel of the cucumber IIhas three placentae, and the ovules are developed in six rows which are usually double, but this is somewhat irregular.” Mann and Robinson (1950) stated that the cantaloup was ”...tricarpellate and has six longi- tudinal carpel edges whose ovule-bearing surface extends nearly the full length of the ovary. Each of these edges appear in cross-section to bear 2 or 3 ovules, the number per edge averaging somewhat over IOO.'l Judson (I929a) observed that cucumber carpels each bear four rows of ovules, two rows on each placenta. One row is found at the extreme edge Saga; ef6fm iagssiz aleItgsuah.evatxeaavaI;9621%ge.The '“fo'ded Figure 8. Longitudinal section of ovary at anthesis. Ioding shows the obules and the conductive tissue forming the path of the pollen tubes. Figure 9. Overall view of petals, sepals, and ovary of pistillate flower. 10. View of the three stigmatic lobes and five petals. ll. Pistillate flower with corolla and calyx removed, showing 3 stigmatic lobes and nectary. 12. Cross section through pistillate flower showing pollen tube path, nectary tissue, and stigma and style. ll of the placenta and the other nearer the receptacle. Mann and Robinson (1950) found that ovules of cantaloup and cucumber were not arranged in straight longitudinal rows, but in short transverse rows, containing alternately two and three ovules. The pat- tern was not always perfect and could be further distorted as seeds en- larged. An approximate idea of how many ovules were present in an ovary could be obtained by multiplying the number of ovules along one carpil- lary edge by six. Materials and Methods A number of pistillate flowers from variety Spartan Progress were examined for the number of ovules they contained at anthesis. The length of the ovule and the number of ovules along a longitudinal section of the ovary were recorded (Figure 8). Care was taken to make sure the longitudinal sections were cut through the middle of the fruit. This way, one carpel was cut in half, and the exposed ovules along the carpel edge could be seen. To test the technique, both halves were counted, and only a slight variation was found. Results and Discussion It was found that there was a positive correlation between length of the ovary and the number of the ovules per longitudinal section (Table 1). Ovaries 9 mm long averaged 23.# ovules per section, or l#O ovules per fruit. Ovaries with a length of 13 to 17 mm produced 38.3 .to #8.8 ovules per section or 239 to 292 ovules per fruit, respectively. l2 TABLE 1 Correlation of ovary length at anthesis with ovules per longitudinal section Ovary length Number of Average number (in mm) ovaries examined of ovules 9 5 23.# 10 I# 27.7 ll 15 3I-3 12 6 33-5 13 8 38.3 1# 8 37.0 15 # #l.5 l6 5 #6.# l7 # #8.8 Differences significant at 0.01 probability level. The ovules were found to exist in short transverse rows as stated by Mann and Robinson (I950) (Figure 8). With the increase in length of the ovary, the width also increased. Additional thickness reflected increased numbers of ovules in the short transverse rows. Shorter ova- ries possessed only two ovules per short row, while the longer ovaries possessed three, four or more per transverse row. Due to poor soil mixture used for the plants referred to above the plants were small and were only able to set one fruit per plant which was often mishapen pro- bably due to nutrient deficiency. The average ovary length of these flowers was approximately 12 mm (Table I). In a group of vigorous plants of the variety Piccadilly grown during Spring I968, the average ovary length at anthesis was 21.5 mm. POLLINATION PROCESS When a honey bee visits a staminate cucumber flower its primary objective is to collect nectar. Although bees were observed in the greenhouse with a small amount of cucumber pollen on their corbiculae (Figure #1), this occurred only when other pollen sources were absent. In order to reach the nectar of a staminate flower the bee inserted its proboscis between the central mass of five anther lobes and the wall of the highly pubescent corolla. In many flowers, pollen was found on these pubescent hairs. Thus, pollen comes in contact with both sides of the proboscis, mouthparts and head of the honey bee. Cucumber pollen grains are relatively large and sticky,readily adhering to the body hairs of the honey bee. The anthers are attached to the wall of the corolla, providing an obstacle to a bee trying to get nectar from the flower. In addition to this, nectar in the staminate flower encircles the pistillo- dia at the base of the flower and is between it and the corolla wall, thus forcing the bee to move around the anthers several times, inserting and withdrawing her proboscis. This tends to ensure that available pollen will be uniformly removed from the anther lobes. Similar bee behavior was observed on pistillate flowers. Because the style is located at the center of the base of the corolla, the bee does not have any barrier to the lower wall of the corolla. However, bees were observed to insert and retract the proboscis several times in circling the nectary, resulting in uniform distribution of pollen on the surface of the stigmatic lobes. FERTILIZATION PROCESS Path of the pollen tube;.pgllen placement on stigne Vasil and Johri (l96#) observed pollen tubes of Fritillaria roylei grow along the tri-radiate canal formed where the three carpels came together. Mann and Robinson (1950) observed that flowers of cantaloup (Cucumis melo L.) possessed three radial arms from cross section views of the style, each of the arms corresponding to the center line of the carpel. ”The tips of the three radial arms of this conducting strand are slightly enlarged in the style and greatly enlarged in the stig- matic region. At its upper extremity, the conductive tissue spreads out onto the stigmatic surface. The pollen tubes take an intercellular path down the conducting tissue, most of them following the enlarged tips of the three arms. Slightly below the level of the upper most ovules, the solid tissue gives way to distinct epidermal surfaces, and from this point the pollen tubes travel over these surfaces to the micropylar ends of the ovules.’I (Mann and Robinson, 1950). Poole and Porter (1933) found that the pollen tubes of the water- melon were confined to the conductive material of the stigma, then ad- hered to the cavity walls of the stigmatic lobes. Mann (l9#3) observed that radially asymmetrical watermelons resulted when pollen was placed on only one or two of the three stigmatic lobes. Seed counts of the in- dividual carpels showed that most pollen tubes grew into the carpel di- rectly below the stigma on which the pollen grains were placed, while 21 percent grew into each of the other two carpels. Teidjens (l928a), however, concluded that the placement of pollen grains on one lobe of the stigma of cucumbers did not alter the subsequent distribution of seeds in the three carpels. I5 Materials and Methods During the fall of 1967 attempts to produce asymmetrical cucum- bers by placing pollen only one stimatic lobe failed. Even seed distri- bution was observed in every case. Therefore, during the spring of 1968 flowers in the greenhouse were hand pollinated using a camel's hair brush. One stigmatic lobe was removed from one half of the flowers with a razor blade. The remaining flowers served as controls. Mature fruits were harvested and the carpels separated and the seeds counted. Using the highest and lowest number of seeds in the three carpels in each fruit, the average difference was recorded for the two groups, as shown by the expression: Highest seed count - Lowest seed count = Difference. Results When the range between the high and low seed count of the three carpels for the two groups were compared using an unpaired t test, the differences were not significant at the 0.05 probability level. The average range between the carpel with the greatest number of seeds and the carpel with the smallest number of seeds was l#.l for the control and 12.# seeds for the experimental group with one lobe removed. This implies that pollen tubes from one stigmatic lobe crossed over into either of the two other carpels. Therefore, pollen placement on all three stigmatic lobes does not appear to be necessary for uniform dis- tribution in the three carpels of a cucumber (Table 2, Figures 13-16). l6 ._m_coumE mc_uo:pcoo mc_3o;m >Lm>o mo xmam caucus“ cc_uoom .o_ ._m_cmumE mc_uonvcoo voc_mym mc_zoLm was“ Lucm_coa e0 omen LagoLLu co_uoom .m_ .CLmuuma mum_meI_Lu .mcocom c_ _m_coumE mc_uo:vcoo vmc_mum ec_oo_ mo uzIomo_u .J. .m__m3 Locc_ co mL_m; 6cm o_>um c_ _m_LoumE mc_uo:ocoo .mEm_pm oumno__cu mCWZOLm maze Lucm_cwa we omen o>cgm >_uLm__m co_pom .m_ ocsm_m I7 TABLE 2 The effect on number of seeds formed per carpel after removing one stigmatic lobe from a fully pollinated flower. Variability in -. M.Number of . Number of Group flowers fruit cucumber seeds pollinated developed per carpel. a b Control #8 #2 l#.l One lobe a b removed A7 3h 12°h l Difference not significant. 3 x2 = 0.3799, d.f. = 0.8927: Difference not significant. b Unpaired t test Cutting one stigmatic lobe appeared to be detrimental to fruit formation, as 3# fruit (72.5%) deveIOped in the experimental group while #2 (87.5%) developed in the control group. Removal of two stigmatic lobes with a razor blade in a separate group resulted in no formation of fruit. TTIne from pollination to fertilization . Literature The length of time reQUired for a pollen tube to germinate and grow into the ovule is poorly understood and apparently influenced by "Hafiy factors. Poole and Porter (1933) found that cucumber flowers pol- I irwated at 9 a.m. had pollen tubes pass the first ovules in the_ovary shc>I't.ly after l2 noon. Pollen tubes of grains placed on the stigma after 9 a-m. were slower in growth. Very high temperatures of 1030 to 117° F dacreased pollen tube growth, but did not stOp it. Fertilization of the ovules had not occurred 7 hours after pollination. Cummings (l90#) found 18 that pollen tubes of Cucurbita pepo L. required six days to reach the ovules. Kremer (l9#3) noted that the pollen tube in cucumber required four days to reach the ovules near the stem end of the fruits at 65°F. Any interference would cause a constriction in the fruit. Material and Methods Removing the perianth of a pistillate flower with attached style and stigma after pollination provided a method of studying the time need- ed for the pollen tube to penetrate the ovary of the flowers. The re- sultant number of seeds would provide an indication of the degree of pollination and fertilization after different time periods. Six replications of five plants each were grown under identical conditions and uniformly pollinated at 9 a.m. using a camel's hair brush. .All work was done in the greenhouse with the minimum temperature set at o :70 F. All plants were growing vigorously and had not previously set 'Fruit. One group served as a control and the perianth tubes were not rennoved. Other groups had perianth tubes removed at O, 3, 12, 2# and 48 hours intervals after pollination. Results As shown in Table 3, no fruit development occurred in the zero ‘ar‘Cl 3 hour interval groups. Two of the five flowers in the 12-hour 9"0Up developed fruits which averaged 6# seeds. With five fruit each, time, 2#-hour group produced an average of 328 seeds and the #8-hour group proCiuced an average of 351 seeds. The four fruit in the control averaged 377 seeds . I9 TABLE 3 Time from pollination to fertilization Style removed __ hrs. Average number Number of fruit after pollination of seeds which developed 0 O O 3 O 0 12 6# 2 2# 328 5 #8 351 5 Control 377 h Five flowers in each original group. Significant at 0.05 probability level. Under the conditions found in the greenhouse, at least 12 hours “mare needed for any fertilization to occur, with 2# hours needed for a Seed count approximating that of the control. These observations were made under favorable conditions. Under variable field conditions, it i S likely that the rate of growth would be slower, as indicated by K remer (l9#3) . INFLUENCE OF THE PLANT 0N FRUIT FORMATION W Literature dealing with variation in flowering patterns and ChanQes in fruit set and development is limited. Working with the 20 Hubbard squash, Bushnell (1920) found that there was a cyclic setting pattern which could only be attributed to a physiological condition of the plant. Environmental factors did not play a major role in the set- ting of fruit. He noticed that the average diameter of the ovary in- creased as the plant age increased. However, no correlation was found between ovary size and set of the flowers. Materials and Methods Investigations were made into the effect of stage of growth of the cucumber plant on the yield and shape of the fruit. Three varieties were studied in the greenhouse: Piccadilly, MSU 356, and Spartan Progress. MSU 35G is a special gynoecious inbred line used as the female parent in the gynoecious Fl hybrid, Spartan Progress, both of which produce very 'few staminate flowers. Piccadilly is a commercial variety technically (:lassified as predominately female by plant breeders and geneticists, ert could be called subgynoecious or intergrade monoecious based upon i ts intermediate nature between a gynoecious and a monoecious line. F’i<:cadilly was selected for use in this experiment due to its popularity Wi th growers . These three varieties differed greatly in the stage of plant deVeIOpment at which first female flowers were preduced. Gynoecious MSU 35G and Spartan Progress produced female flowers at an early stage ch l3lant development, usually at the first node, and continued to pro- duce female flowers. Piccadilly usually produced staminate flowers on ti‘EB 'First 5 or 6 nodes, and thereafter only an occasional female flower. MSU 356 and Spartan Progress were divided into three replications 21 of 13 plants each. Piccadilly was divided into two replications of 13 plants. One replication of each of the three varieties was pollinated when the first pistillate flower appeared; the second group was pollinat- ed one week later. The third group of Spartan Progress and of MSU 35G were pollinated three weeks after the first flower. As implied above, the actual dates of these events for the three varieties did not occur at the same time. Results Spartan Progress and MSU 35G produced significantly more seeds as the age of the plant increased. Piccadilly did not exhibit this in- crease, largely due to the delay in appearance of the first female flo- I~er. In MSU 35G and Spartan Progress, the number of seeds per fruit increased by about 100 seeds for each step in pollination delay (Table # and Figure 17). All 3 varieties produced significant increases in the length/ vvim 6cm 0mm “cm—a wo a_£mco_um_mm .m_ mczm_m an 3.0 3...: ...:qu 5:32.. m.omwm_w gnu a w._< m o< e 24: m0< h2<4m 2# TABLE 5 Effect of age of the cucumber plant on length/width ratioa of mature fruit. m a - Flower Flower Variet First pistillate pollinated one pollinated three y flower on vine week after first weeks after first pistillate flower pistillate flower SPARTAN PROGRESS A. _ 2003 2053 2.83 aF(2,32) - 20.#2 Number in sample l3 l3 9 MSU GYNOECIOUS 35 I. 20141,, 2.8 *F(2,3|) = 228.0 97 5 Number in sample II I3 I0 PICCADILLY 2 12 2 5| "‘F(I,|8) = 28.70 ° ' Number in sample 13 7 " aLength/Width Ratio: A measurement of the length of the fruit without the stem at the longest point, divided by the width of the fruit at the widest point. *Differences of row values all significant at 0.01 probability level. ..... .o_umc Luv_3\;umco_ 6cm omm Hem—a mo a_;mco_pm_om .w_ mesm_m 3.3.”. “MIG 3W2 mmnmnmuwm anthOhm cabana” 035. : 3.2;.3023 0." nd 9N n; Lo.o._ 25 3.00; 00:: .0.0a 3003 0:0 bO’O-L a .3: 0:5. Ieo_3\:eozfl .m> mo #24:: 3. :3. 26 Node Position Materials and Methods Field observationswere made in the summer of 1968 on the influence of the node position of a cucumber upon its length/width ratio. The length, width, shape, constriction, and curvature of the fruit were re- corded according to the node number of the fruit. Only fruit three inches or longer were harvested. The plot were harvested frequently to increase the number of fruit produced per plant during the season, and to utilize the greatest number of pistillate flowers possible. Results The length/width ratio of fruit from the main vine was greatly influenced by position on the vine (Table 6). Very short cucumbers were found at the first node, with the length/width ratio averaging 2.33. At the tenth node, the value had increased to 2.78. As shown in Table 7, short fruit were found more frequently on low nodes while long fruit were found on the high nodes. Fruit constriction and curvature could not be correlated with the node number of the fruit. The length/width ratio of fruit on lateral vines was not influenced by position. Table 8 shows the lowest ratio of 2.60, and the highest value at 2.69, a non-significant difference. As shown in Table 9 no pattern is shown in the shape, constriction, or curvature as due to position on the lateral vine. Figure I9 shows the increase in the length/width ratio for fruit from different nodes on the main vine, and consistent ratios of fruits from the lateral vine. 27 TABLE 6 Effect of node position on cucumber shape average length/width ratioa of main vine fruit. Node Number Sample size Average length/width ratio 1 101 2-33 2 1#5 2.33 3 l9# 2.#3 # 267 2.#9 5 308 2.52 6 250 2.61 7 I93 2.62 8 155 2.83 9 79 2.67 10 #7 2.78 aLength/width ratio: a measurement of the length of the fruit (without stem) at the longest point, divided by the width of the fruit at the widest point. Differences significant at 0.05 probability level. 28 .o_omc cuv_3\zumco_ 6cm co_u_moa 660: $0 a_;mco_um_om Lonfiauau we Lonfiaz Oval 0— a o n o n V n N — o.._>. c _o.z 05> .03....— co:_mon_ evoZ xm vmfscmemo 0:3. f_3>>\fmc.: .m_ mt:m_u —.N Nd n4. id Wu 9N Nu ad ..0 oeIoa HIPLM/UI5“¢T 29 TABLE 7 Effect of node position on shape of fruit from the main vine. Grading percentages. N:::EEHE:::pIe Shape Constriction 7 Curvature Number 5'23 Short Normal Long None Slight Severe None Slight Severe l 101 59 34 5a 73 zh la 6l 29 la 2 l#5 39 52 8a 68 25 5a 50 38 Ila 3 l9# 28 61 10a 63 19 17a 50 37 12a # 267 20 66 12a 66 25 8a 5# 35 9a 5 308 2l 65 13a 69 17 12a 6# 26 9a 6 250 I7 61 21b 68 20 11a 5# 29 16a 7 I93 10 67 21b 70 23 5a 57 29 12a 8 155 8 71 20b 67 21 lla 51 3I 17a 9 79 6 64 28b 75 17 6a 52 #0 6a 10 A7 u 70 25b 79 1# 10a 57 36 6a Differences significant at the 0.05 probability level are shown by different small letters in each group. 30 TABLE 8 Effect of node position on cucumber shape(average length/width ratio of lateral vine fruig, Node Number Sample size Average length/width ratio 1 65 2.65 2 189 2.69 3 235 2.63 # l#l 2.66 5 58 2.60 6 30 2.68 No statistical difference at 0.05 probability level. TABLE 9 Effect of node position on shape of fruit from lateral vine. Gradlng percentages. w Node Sample Shape Constriction Curvature Number Size Short Normal Long Normal Mild Severe Normal Mild Severe l 65 9 60 30 72 2| 6 53 35 IO 2 I89 9 59 30 67 28 3 #7 39 I2 3 235 ll 6# 23 69 2# 5 52 37 9 # I#1 8 58 32 71 I8 10 #7 37 15 5 58 6 7# 18 86 12 l #3 #3 l3 6 3O 10 #6 #3 83 IO 6 S3 #3 3 No statistical difference at 0.05 probability level. 31 Discussion The following points summarize the effect of the node position of a cucumber fruit upon seed counts and fruit shape. 1. Fruits from a higher node position, and thus older plants, con- tained more seeds and were longer than fruits from lower nodes (i.e., younger plants). 2. Effects of the node position were most noticeable in highly gynoecious varieties. Variety Piccadilly showed little increase in number of seeds as a result of delayed pollination, largely because the plants did not produce pistillate flowers until a week or more after the initial staminate flowers appeared on the plant. 3. The number of short cucumbers was most frequent on low node numbers while the amount of constriction and curvature was not influenced by node number. Flowers on the first and second nodes were usually small and fre- quently aborted. Thus fruit usually did not form until the third or later node. This is shown in Table 7, where 101 fruit were obtained from node I, and 308 were obtained from node 5 from this group of plants. It is possible that there was a physiological inability of the plant to set fruit at this early growth stage. No cyclic pattern in fruit setting was observed on cucumbers as described by Bushnell (1920) on squash. As indicated in the review of flower anatomy, there is a positive correlation between the length of the ovary at anthesis and the number of ovules. The consistency of these reported observations and our ex- perimental results indicates that flowers early nodes are smaller, have smaller ovaries with less ovules, and develop to shorter fruits. 32 Parthenocarpy ingpickling cucumbers Literature The cucumber is capable of producing fruit parthenocarpically, i.e. without either pollination, fertilization, or seed formation. So prevalent is this characteristic, that selected commercial varieties of parthenocarpic fresh market cucumbers have been in use for years in Europe. In recent years, plant breeders have deveIOped parthenocarpic strains of pickling cucumbers (Pike, 1968, 1969). Results of research dealing with the pollination of pickling cu- cumbers can be confused by the development of fruit parthenocarpically, particularly under certain environmental conditions. A review and Study of factors contributing to parthenocarpic fruit deveIOpment would pro- vide information helpful in evaluating results of pollination studies. The inheritance of parthenocarpy was reported due to an incom- pletely dominant pair of single genes and perhaps several modifier genes (Pike, I968, I969). Plants homozygous for the parthenocarpic charac- teristic produce fruit without pollination under proper growing conditions, whereas, plants heterozygous for this characteristic produce fruit parthe- nocarpically only on older plants. The genetic lines mentioned in Pike's work were selected for their genetic parthenocarpic tendency. They did not evaluate the parthenocarpic tendency of commercially available pickling varieties. Teidjens (l928a) noted that high light intensities reduced parthe- nocarpic development of cucumbers. A reduction in the amount of light resulted in a vegetative response which brought about parthenocarpy and ”easier fruit set”, implying that the percentage of flowers which 33 developed was higher. Takashima and Hatta (1955) and Wong (1936) found that parthenocarpic development of fruit could be induced by treatment with Napthaline acetic acid (N.A.A.), a plant growth hormone. McCollum (193#) found that seed forming fruit inhibited the de- velopment of other fruit on the vine to a greater degree than partheno- carpic fruit. Munson (1892) reported that some varieties of cucumbers produced parthenocarpic fruit with withered, shrunken, and often hollow apex ends. Berkel and Vriend (1957) indicated that established parthe- nocarpic fresh market varieties reSponded to bee pollination with the deveIOpment of seeds in the flower end of the fruit. This resulted in misshapen fruits. Koot (1960) added that the occurence of these I'bull necked fruits“ was also influenced by the variety, the age of the plant, and the male to female flower ratio. Observations and Discussion In the field plots of 1967, the variety Spartan Dawn developed no parthenocarpic fruit with the exception of a few produced under cages with bees excluded. These fruits were small, curved, and often hollow, fitting and description of Munson (1892). In the field plots of 1968, no parthenocarpic fruit were observed during July or August in the variety Spartan Progress. During September and October, however, plants of this variety grown in isolation and containing no monoecious plants developed large numbers of cucumbers parthenocarpically. Most of the cucumbers Were normal in shape. Plants grown under stress in the greenhouse rarely developed Flarthenocarpic fruit, and when they did, they were on older plants and ‘Shaped like those described by Munson (1892). On vigorously growing 3# plants, the strongest parthenocarpic expression was observed in the fall of 1967 and the spring of 1968. Many of these fruit were normal or nearly normal in Shape. During the summer of 1968, parthenocarpic fruit appeared on a few greenhouse plants after the glass had been whitewaShed. These instances indicate that environmental factors influence the degree of parthenocarpic tendency of a plant. On one day a flower may require 200 pollen grains to develop a fruit, while on another day only one pollen grain may be needed. Changes in the degree of parthenocarpic tendency as affected by environment may have an influence upon fruit shape. If a cucumber start- ed to develop parthenocarpically under environmental conditions favoring parthenocarpic expression, and the favorable weather conditions changed, a poorly-shaped fruit might result. The commercial value of a partheno- carpic tendency in gynoecious varieties Should be tested more completely as it could have an important bearing on pollination requirements as well as the problem of growing large numbers of well shaped cucumbers per vine. Presence of inhibitive fruit Literature McCollum (193#) reported that cucumber fruit growing on a vine inhibited the set and development of subsequent fruit. He found that the uppermost fruit on the vine has a physiological advantage due to its Physical nearness to photosynthetically active leaves. Upon inhibition 0f vegetative growth, fruit disintegrate with the phloem and other tissues collapsing. In contrast, when the cucumber fruit reached #-5 inches in length, vegetative growth of the plant ceased. 35 Mann and Robinson (1950) found that maturing cantaloup strongly inhibited subsequent fruit production. When no inhibitive fruit were present 66% of the newly pollinated flowers set fruit, compared to 9% with other fruit present. They also found that removal of inhibitive fruit up to # days after pollination resulted in normal fruit deveIOp ment. Similar removal eight days after pollination resulted in a re— duction in the level of fruit setting, although the number of seeds was unaffected. Materials and Methods During the summer of 1967, flowers were tagged and subsequent fruit development observed to identify the inhibitive effect of other fruit. The position of the marked flower relative to the remaining fruit and flowers was recorded. Position obServations were based upon the closeness to the crown (root end) of the plant. Observations were made on young and old plants of the variety Spartan Dawn. The same plants were used several times for different investigations. Vines were selected and all fruit removed. Only one flower per plant was tagged and observed. From these tagged flowers the percentage of fruit development was determined for different positions on the vine and with different numbers of inhibitive fruit present on the vine. The resulting fruit were graded into the classes of either perfect, neck, nub or crook depending on the degree of constriction or curvature. Results Of the developing fruit on a cucumber vine, the fruit closest to the roots deveIOped 39.5% of the time (Table 10 and Figure 20). Fruit 36 in second position deveIOped 29.1%, third position 18.5%, and fourth position 15.8%. Fruit in the fifth position did not deveIOp. The change in the quality of the fruit in the different positions is shown in Table 10. The relative percentage of neck and nub shaped fruit increased only slightly in the second position. Curved or crooked cucumbers were only found in the position closest to the roots. As shown in Table 11, and Figure 20, the influence of the number of other fruit on the vine played a major role in the percentage of fruit development. As the number of cucumbers on a vine increased, the per- centage of flowers which developed dropped from 33.3% for fruit without competition to 21% for fruit with two competing fruits. No more than four fruits per plant were obtained in this study. The frequency of necks and nubs decreased very slightly as the number of other fruit in- creased, whereas, the frequency of crooked fruit increased slightly. Discussion The advantage of fruit located closest to the crown of the plant was observed by McCollum (193#), Mann and Robinson (1950), and Cunning- ham (1939). Fruit closest to the crown had the greatest probability of developing and the greatest probability of being well-shaped. Neck and nub-shaped fruit were most common farthest from the crown of the plant. From an economic standpoint, a great step forward would be the development of a cucumber capable of developing # or 5 fruit at the Same time rather than the present one or two. With present varieties, One or two cucumbers develop on the vine and most of the remaining P0llinations are wasted effort. This is a significant factor in trying to assess pollinator requirements for maximum yield in a field of cucumbers. 37 ._ LonEJc co_u_moa c_ m. .LonEJc 660: mu_ *0 mmo_cemmmc oc_> osu co u_:ew Loguo oc ;u_3 u_:Lm < .u_:cw ozu mo umo>emz wo oE_u ox“ um oc_> one :0 ucomoca u_:cm cozuo mo Logan: oz“ >n emc_Ecouoo m_ co_u_mom _ o.e 0.0 0.0 0.0 o.o e.eo_ e o e e o e m w.m_ 0.0 0.0 ~.m m.o_ ~.em m. o o _ N o. e m.m_ o.o m.o m.e m.e_ m._w me. o _ _ k. em m _.mN e.o _.e m.m m.k_ m.o~ ~e~ e e. e. o: om_ N m.mm o.~ :.: :.m :.m~ m.om mum m. .4 mm NNN mmm _ o>__< x xooto x 632 x. xooz e.oooetoe x eeoe x .oooe xooco n32 xuoz “cameo; vmoo _co_u_m0m mo:_m> ummucouLom Rem. home eo>tomoo mamsm one How u_scw co oc_> egg :0 co_u_moa mo mucm:_wc_ o~ mem__<_x xooto x 632 x xooz x Homecoe x_eooo x _oeoe xooto eez xooz eooeeoe eooo emoewm mo:_m> ammucouLom mom_ I name vo>gomno .oamcm u_3ew new uom u_:Lw co u_zcm Loguo mo Logan: wo «econ—w:— __ u4moco£ eo Ccouuma u;m__e 6cm >mo eo oE_u wo co_um_om .NN ocam_m mwh32_<< on 2. m¢m>>04m o— 0... mh_w.> mun ".0 «uni—47. m0<¢w>< v m a so _ eooeu _ _ _ o. o Eco / >._._>_._.U< FIGS“. koo— mm on nu /\ SLISIA :10 UBBWHN 50 Figure 23 shows the flight activity on a day with a warm overnight temperature and high temperatures during the morning hours. Flight acti- vity was underway at 8 a.m., and peaked at 9 a.m. Figure 2# shows the opposite extreme. When temperatures were cold overnight, and morning temperatures were low, the bee flights did not start until 11 a.m. How- ever, greater numbers of bees were observed on this day, indicating that pollination of cucumber flowers in the field on that date did not suffer due to this slow start. Figure 25 shows the flight activity on an average day, resembling the activity in Figure 22. The average number of visits per flower for ten 30 minute obser- vation periods was #5.1 visits. Because observations were made for 30 minutes out of the hour, the observed values were doubled to reflect 10 hours of pollination. On this basis, the average flower received 90.2 visits from 8 a.m. to 6 p.m. On an average, each flower was visited once every 6.65 minutes (600 minutes/90.2 visits) at a rate of 9.02 vi- ‘Sits per flower per hour (V/F/H). The bee p0pulation in the plots was estimated to be three colonies per acre. As implied above, the time of the first bee flights on cucumbers appeared to be related to temperature and the general growth pattern of the plant. As shown in Table 15, the time of the first flights varied considerably with temperature. The average time of the first visit to cucumbers for the entire season was 8:30 a.m. EDT in East Lansing. Bee visits and colonies per acre Commercial cucumber fields supplied with one colony of bees per acre were used to compare with plots in East Lansing supplied with three colonies per acre. Table 16 shows the detailed weather information for 15 E 12 ; w- 9 O .3 6 e E 3 Tlme 9o 1‘ O 2 2 2 7o 0 l .0 so 51 Bee Visits/Flower/Hour 8 910111212 34 Temperature /‘/\/\ (Overnight low 60°F) August9.l967 Figure 23 Number of Visits 15 I2 52 Bee Visits/FIower/Hour Time 8 910111212 3 4 Temperate re (°|=) 90 80 70 60 50 Temperature (Overnight Low 50°F) August25.l967 Figure 2# N umber of Visit: 3e 25 2O 15 IO 53 Bee Visits/Flower/Hour Time 8 910111212 3 4 Temperature (°F) V e 8 o. 0 Temperature (Overnight Low 38°F) August23J967 Figure 25 the dates of these visits. TABLE 15 Air temperature at the time of the first flights of the day on cucumbers. Date Time of first bee visit Temperature 0F at time (Time in a.m.) of first flights August 1 9 63 2 Before 8 68a 3 8 6l 9 Before 8 70a 11 10:30 62 l# 10 62 17 Before 8 65a 22 8:30 60 23 11:30 65 25 9:30 60 Seasonal Average 8:30 a.m. 62°F Range: Daylight to 11:30 a.m. 6O - 65°F aOvernight low higher than range of other values. These values were excluded from computations. 55 TABLE 16 Weather observations for central Michigan during comparative counts of bee populations on three dates. August 9 _ August'23 August 25 Temperature: High 85°F 76 83 Low 60 38 50 Average 73 57 67 Wind velocity 13.5 mph 6.2 mph 2.6 mph Percentage of. 66% 100% 77% Sunllght recelved Precipitation 0.02 in. none none Comments Fog, haze in Fog in a.m. Heavy fog a.m. and haze Prepared from ESSA Weather summary for Lansing, Michigan. 1967 Counts at Springport, Michigan: On August 9, 1967 each flower received an average of 9.75 bee visits during seven 30-minute counting periods in a 20 acre field. This compare to 29.0 visits per flower during the identical counting periods in East Lansing. This information is shown in Table 17. When these values are converted to V/F/H, the Sprjrlgport field received 2.78 V/F/H while the East Lansing plots receiv- ed 8.28 V/F/H. The Springport field, with one colony per acre, had 33.6% of idle pollination activity of the East Lansing plots where three colo- nies per acre were present. Plants in Springport were near the end of the first week of bloom and flowers were abundant. Plants in East Lansing had been in bloom 56 for several weeks but were producing flowers in approximately the same numbers as the Springport field. TABLE 17 Bee flight counts for Springport and East Lansing, Michigan August 9, 1967. Time Range Visits per Flower Springport East Lansing 9-10 a.m. 0.9 5.7 10-11 2.# 5.7 I 11-12 2.1 5.3 12- | p.m 2 O #.5 l- 2 2.2 3.9 2- 3 0.9 23 3- LI 0.3 l 6 Visits per flower in 210 min. 9.75 29.0 V/F/H 2.78 8.28 Number of plants in sample 20 10 When plants were harvested in both fields to duplicate a destruc- tive mechanical harvest, with no previous fruit removal, the number of <:ucumbers produced per plant under these different bee populations did ruat differ statistically at the 0.05 probability level (Table 18). The East Lansing plot produced an average of 1.#7 fruits per plant. In 57 Springport, the average yield per plant was 1.5# fruit. According to the field manager for the Springport field, the overall cash value of the cr0p was the highest obtained during 1967 by the processor (Schwarz, 1967). TABLE 18 Plant yield data from three fields for the first harvest duplicating mechanical harvestinga. '1 . Number of fruit per plant East Lansing Springport Cedar Springs 1.33 1.72 1.59 1.#9 1.56 1.19 1.59 1.12 1.37 Average 1.#7 1.5# 1.38 One way analysis of variance indicated no significant difference at the 0.05 probability level. aThis table serves to compare the first harvest from different plots at different times of the summer. It is provided to indicate that similar yield results were obtained under differing conditions and should not be used as an absolute comparison of yield. Counts at Cedar Springs, Michigan: On August 23, 1967 each flower observed in the 16 acre Cedar Springs field received an average of 2.# visits per flower in nine hours, compared to 50.3 visits per flo- ‘Ner in East Lansing during the same time period (Table 19). These re- Present 0.53 V/F/H in Cedar Springs and 11.7 V/F/H in East Lansing, and_only l+-8°/c'.of E. LanSing activity in C.Springs. Sixteen colonies of bees were 58 located in the Cedar Springs field but blossoming was in an early stage which undoubtedly contributed to the low bee count. On August 25, I967, 10.# visits per flower were observed in the Cedar Springs field while 30.8 visits per flower were observed in East Lansing (Table 20). The Cedar Springs field received 2.95 V/F/H while the East Lansing field received 8.68 V/F/H. Thus the bee visit activity in the Cedar Springs field was 3#.2% of that in East Lansing. The East Lansing plots produced 1.#7 fruit per plant, while the field Cedar Springs produced 1.38 fruit per plant, a non-significant difference (Table 18). TABLE 19 Bee flight counts for Cedar Springs and East Lansing. August 23, 1967. Time Range Visits per Flower Cedar Springs East Lansing 8:30 - 9:30 0.0 0.0 9:30 - 10:30 0.0 0.0 10:30 - 11:30 0.2 0.1 11:30 - 12:30 0.5 1.1 12:30 - 1:30 0.# 15 1 1:30 - 2:30 0.9 13.1 2:30 - 3:30 0-3 9 9 3:30 - #:30 0.3 8.0 #:30 - 5:30 0.0 2.6 Visits per flower in 270 minutes 2.# 50.3 V/F/H 0.5 11.2 Plants in sample 20 10 59 TABLE 20 Bee flight counts for Cedar Springs and East Lansing. August 25, 1967. Time Range Visits per Flower Cedar Springs East Lansing 8:30 - 9:30 0.0 0.0 9:30 - 10:30 0.0 0.3 10:30 — 11:30 0.8 #.2 11:30 - 12:30 2.# 5.8 N 12:30 - 1:30 1.7 8.# 1:30 - 2:30 2.1 5.6 2:30 - 3:30 3-2 6-5 Visits per flower in 210 minutes 10.# 30.8 V/F/H 2.95 8.68 Plants in sample 20 IO On August 23, the Cedar Springs plants produced an average of 0.56 pistillate flowers per plant while on August 25, 0.92 pistillate flowers per plant were produced. This represents an increase from 28 to #6 thousand pistillate flowers per acre during a two day period. This had an important effect in increasing the bee visit rate from 0.53 to 2.95 V/F/H. No correlation between bee visits and yield can be inferred from this data due to variability in maturity of plants at the time counts were made. These data are of interest in pointing out the rapid build-up of bee activity with a greater intensity of bloom. 60 Length of individual bee visits The length of a bee visit to a pistillate flower decreased as the number of the visit increased. As Shown in Table 21 and Figure 26, the first visits to a cucumber flower in 1967 lasted an average of 36.2 seconds, while the average length of subsequent visits dr0pped sharply. The average length of the fourth through twelfth visits to a flower was 10.6 seconds in 1967 and 7.9 seconds in 1968. Observations in the field indicated that bees settling on a flower worked around the stigma and style for 5 to 10 seconds even if unsuccessful in finding nectar before moving on to another flower. TABLE 21 Number of bee visit and average bee visit length Number JLV I967 1968 of the Sample Average length Variance Sample Average length Variance visit Size in seconds Size in seconds I 585 36.2 0.99 #78 39.2 1.16 2 510 1#.8 0.91 399 12.9 0.7# 3 #03 11.# 0.53 326 10.1 0.53 # 295 10.8 0.62 261 8.1 0.50 5 201 10.6 0.76 l9# 7.7 0.51 6 116 12.5 1.#0 132 8.3 0.52 7 72 9.3 0.8# 90 8.3 0.62 8 50 9.1 1.11 55 7.0 0.61 9 27 8.7 1.38 #0 8.0 0.9# 10 15 8.3 1.6# l# 5.# 0.89 11 8 10.6 1.72 9 . 8.3 l.## 12 -- -- -- 6 #.5 0.50 61 .u_m_> ox“ mo camco_ exp 6cm Lozo_e m Op u_m_> eon ox“ wo LonEJC oz“ mo a_;mco_um_om ._._m_> LO mmmZDZ N— —— o. o m k 0 n v. m N / I/ I 4 .l.ll.l mnvo _ I IHOZmA .:m_> w.» Em; Lo 3332 :3 I .om oczm_m _ 0— ON on 0V JO H19N31 llSIAI (SONODBM 62 Length of the first visit to a flower varied with the time of day of the visit. As shown in Table 22 and Figure 27, early morning visits were shorter than those occurring after ll a.m., indicating that a greater supply of nectar had accumulated. The length of the first visit of the day varied on different days. In I967 the average length of visit varied from as low as 22 seconds to as high as 50 seconds. Table 23 shows the length of these visits as well as a summary of weather data for the period of the observations. No single weather factor could be directly correlated with the differences in initial bee visit length. TABLE 22 Effect of time of day upon length of initial bee visit 7‘3:y°f Sample Aver;::7length Variance Sample Aver;::8length Variance Size in seconds Size in seconds 8- 9 a.m. 6 l0.5 3.ho -- -- -- 9-lo l6 8.8 .95 ll 27.0 u.h8 lO—ll 36 30.3 .22 Sl 34.2 3.58 ll-l2 86 38.3 .Ah 8] #6.5 2.97 l2- 1 p.m. l6 38.5 .37 -- -- -- l- 2 .M9 39.9 .85 6h h3.l 3.l9 2- 3 ‘36 37-3 .98 97 37.1 2.5“ 3- h 87 37.9 .8h l2h 38.3 2.36 h- 5 “7 37.2 .67 38 h0.l 3.l5 5- 6 3 45.0 .58 -- -- -- 63 Abouv > ._<_._._Z_ m> > umc_w osu mo Lumco_ oLu ocm >mo on“ $0 oE_p ozo wo o_meo_um_om .NN ocsm u on }O ngua1 OBDJOAV Inlllul l!’!lA spuoaos u! 6h Discussion Flight activity Bee flights were found to start on the average at 8:30 a.m. EDT at an average temperature of 6OOF in East Lansing. These observations agreed with those of Seaton and Kremer (I938). The peak activity of honey bees on cucumber flowers was found to occur at noon, the time Barnes (l9h7) found to be optimum for greenhouse pollination. Maximum bee flights in the field extended from 10 a.m. to 3 p.m. This agrees with Nemirovich-Danshenko (l96#) who found that nectar secretion, pollen abundance and bee visitation were at their highest levels from ll a.m. to 2 p.m. during greenhouse observations. An average of 9.02 visits per flower per hour occurred over a ten hour period on cucumbers during 1967, in East Lansing. Mann (I953) found that there were an average of 5.3 V/F/H on cantaloup during 10 hours. The difference between the two species was possibly due to the greater number of colonies per acre in the cucumber observations. McGregor and Todd (I953) found that bees usually followed rows of cantaloup rather than jumping from row to row. However, in cucumbers planted at high populations, the vines are so intermingled that bees were not observed to follow a single row, but moved in an erratic pattern, although usually in one general direction. In greenhouse work, where plots were arranged in rows, bees worked down the rows more often than moving from row to row. They would tend to follow a single vine rather than move from vine to vine. 65 TABLE 23 Weather data for Lansing, Michigan and the average length of the first bee visit to cucumber flowers. L Temp. Temp. Temp. Rain Days Average Percent Average l3; Date Maximum Minimum Variance f" from Wind Sunshine Visit Length In. Rain JULY I967 l2 80°F 59°F - 2 0 l 12.h 95 26.# I3 69 50 -l2 t 2 9.6 93 27.3 ]7 8h 57 - ] 0 6 8.8 72 29.9 18 8] 62 0 0.]8 7 9.] 90 22.3 ]9 80 6] - l 0 l 5.0 5] 33.0 20 82 57 - 2 0 2 3.9 87 9].] 2| 86 54 - 2 0 3 3.3 99 36.3 AUGUST ]967 l 86 55 - ] 0 3 5.9 93 3].8 3 88 59 - ] t 5 8.2 9] 38.3 9 85 60 ] 0.02 ] l3.5 66 28.8 I8 83 6h 9 0.0l 8 8.9 hh 90.9 2] 79 55 - 3 t ] ]].] 8] 93.9 22 73 97 - 9 0 2 8.2 9] 43.8 25 83 50 - 2 0 5 2.6 77 50.0 28 79 56 0 0 3 l0.“ 9] 39.3 29 8] 52 0 0.04 3 9.“ 60 47.2 AUGUST I968 2 80 49 ‘ 7 0 2 5.2 92 32.9 5 8] 70 h 0.9] 5 6.3 7 2].7 6 88 72 8 0.0] 0 9.9 25 92.6 7 87 68 6 0.08 0 7.] 70 42.3 8 86 65 h 0.30 0 3.9 53 50.4 9 8h 68 h 0.09 0 6.2 56 3A.] 1“ 79 52 - 5 0 5 6.0 94 37-9 l5 8| 50 - 5 0.0h 6 8.8 79 h3.h I9 89 68 9 0.75 l ]0.9 8] 42.9 2] 92 66 9 0 3 6.0 65 29.6 23 9] 75 lh 0 5 9.] 92 98.2 28 7h 9] -]0 0 9 3.3 IOO 27.5 66 Bee visits and colonies per acre Three times as many bee visits to flowers occurred where there were three times as many colonies of bees. It is felt that it is unsafe to relate bee visits to number of colonies as closely as was indicated by these data. Attractiveness of other plants in the area could have been a factor, as well as the length of time past bloom for bees to ad- just their flight patterns to the field. The fact that the number of visits per flower per hour increased in preportion to the amount of bloom present in the field located in Cedar Springs, might indicate that cucumbers were relatively attractive to honey bees and that a given number of colonies could provide quite a range of bee visits depending on the number of flowers and their relative attractiveness. Length of individual bee visits A great amount of variation was observed in the length of time bees spent on individual cucumber flowers. The first visit to a flower by a honey bee lasted an average of 36 to 39 seconds. Subsequent visits dropped sharply in length. Visits after the fourth averaged from 8 to II seconds. This point of bee behavior is of considerable significance in assessing pollination ecology of cucumbers and has apparently been missed by previous observers on other vine crops, if it exists on them. The length of the first visit of the day to a flower varied with the time of day. Early morning visits were shorter than noon or late afternoon visits. This would indicate that the nectar content of the flowers increased during the early part of the day. 67 Some days during the observation periods of I967 and I968 produced a wide range in the length of first visits. No explanation can be offer- ed for this variation. It was hypothesized that the length of the visit was directly linked to the amount of nectar available in the flower. Shuel (I967) correlated nectar secretion with solar radiation, air tem- perature, atmospheric humidity, soil moisture, soil fertility and the genetic constitution of the clone. Table 23 provides information on the climate for the days when observations were made, but no close correla- tions could be made with any single factor. No evidence was found that the stigma of the pistillate cucumber flower was receptive for just a short period of time in Michigan. This condition was observed in cucumbers in India (Chaudhury and Phatak, l96l) and on cantaloup in Arizona (McGregor, Levin and Foster, I965). In East Lansing, successful pollinations were obtained at many separate intervals throughout the day, implying that whenever bees were present in the field, fruit could apparently be set. Air temperatures in Michigan rarely ex- ceed 95°F and the relatively humidity is usually high. These factors may result in the long period of receptivity of the cucumber flower stigma. The priod of optimum flower exposure to pollination is discussed in Chapter III. High wind velocity was often associated with poor bee visit acti- vity. However, bees were able to work cucumbers at relatively high wind speeds due to the dense protective cover provided by the plants. Bees were able to fly beneath the canopy of leaves, visiting flowers for long periods of time. Bees were also observed working under the canopy of leaves during light showers, without any major change in their behavior. 68 Attractiveness of cucumber to honey bees The attractive nature of cucumber flowers to honey bees was shown in a move of commercially-owned bees into established cucumber fields. One week after the bees were introduced into paired fields, the bees were switched in order to note any change in numbers of foraging bees. No change was noticed, indicating that the cucumber plants were attractive enough to keep bees flying to them regardless of the length of time the colonies had been placed in the field. FRUIT SET AND DEVELOPMENT AND SEED NUMBERS AS INFLUENCED BY THE NUMBER, LENGTH AND TIME OF DAY OF BEE VISITS LlTERATURE Number of bee visits and length of visit Shemetkov (I957) reported that 8 to ID bee visits to a flower were needed for fruit formation in Russian hothouse cucumbers. Fruit formation occurred with as few as two bee visits, but the average weight and number of seeds produced per fruit increased with additional bee visits. He recommended 8 to l0 visits in order to provide adequate pollination. There was no increase in the number of fruits per plant resulting from more than l0 bee visits. Adlerz (I966) found that a pistillate watermelon flower needed at least eight separate bee visits in order to develop normal fruit. The average time spent visiting a female flower was 8.0 seconds and a male flower 5.7 seconds. He computed that one colony per acre provided bees in excess of those needed for adequate pollination. Mann (I953) found that staminate and perfect cantaloup flowers were visited at a different frequency. Perfect flowers were visited 53 times per day, in comparison with #2 visits to staminate flowers. This difference was attributed to the larger corolla size of the perfect flowers. 69 70 t al. (I965), working with cantaloup, reported that a McGregor 3h.7% set was obtained with 8 or less bee visits and a 5h.3% set with more than 8 bee visits. Flowers exposed to open pollination in Arizona developed 38.3% of the time, producing 22.N% marketable fruit. Fruit set and marketability improved with additional bee visits up to l3-lh visits per flower. It was observed that the average visit time of a honey bee on a cantaloup flower was ten seconds. They concluded that stigmatic receptivity was short under high temperatures and recommended one bee per ten perfect flowers in order to insure good pollination. Sakharov (I957) reported that a bee visited the flowers of certain Cucurbita species for an average of l0.9 seconds per visit. Number of seeds and shape of cucumbers Tiedjens (l928a) reported that he could find no correlation be- tween the number of seeds and the shape of cucumbers. Fruit with as few 2 seeds did not vary in shape from those possessing 200 seeds. He report- ed that the number of seeds found in the cucumber never exceeded the number of pollen grains placed upon the stigma of the flower. Fruit constricted at the stem end could be produced by delaying pollination 2h hours after anthesis. Tiedjens theorized that constriction at the flower end of the fruit resulted from ovule starvation following inhibi- tion by other developing fruit. Seaton (I937) working in Michigan, divided cucumbers into three shape groups: (I) straight, trilocular fruit (perfects), (2) fruit con- stricted at the stem end (necks), and (3) fruit constricted at the flower end (nubs). Seeds were counted in mature fruit and significant positive correlations were found between fruit weight and the number of seeds, 7I especially imperfectly shaped fruit. He concluded that the number of deveIOping seeds determined the amount of fruit tissue that developed, implying that more seeds resulted in better fruit shape. Perfectly shaped fruits had twice as many seeds as fruits constricted at the stem end and three times as many seeds as fruit constricted at the flower end. Berkel and Vriend (I957) reported that parthenocarpic fresh mar- ket cucumbers produced seeds in the flower end of the fruit when the flower was partially pollinated. The presence of seeds results in re- duced market values for parthenocarpic varieties. Verdieva and Ismailova (I960) reported that the average weight of squash (Cucurbita sp.) in- creased with honey bee pollination over those flowers pollinated by other unspecified means. Honey bees were observed visiting squash blossoms for nectar and not for pollen. McCollum (l93h) reported that up to 500 seeds were produced in individual white spine variety cucumbers. Tiedjens (I926) correlated the poor shape of cucumbers with wilting of plants resulting from exces- sively rapid transpiration due to over-watering. He recommended that cucumbers be grown in light soil with adequate fertilizer and that water be applied sparingly. Time of day and length of time for optimum pollination of cucumbers Barnes (l9h7) working on cucumber breeding methods, reported that cucumber anthers dehisced between 9 and 9:30 a.m. in the greenhouse. While pollen remained viable until i or 2 p.m., the optimum time for utilization of the pollen was l2 noon. Hayase (I960) found that in certain Cucurbita species, the stigmas of pistillate flowers were first receptive to pollination the day prior 72 to anthesis; increased in receptivity until the morning of the day of anthesis; then receptivity decreased. The pollen was viable beginning at I2 noon the day before anthesis reached a peak at midnight and tapered off by 9 a.m. on the day of anthesis. After 9 a.m. it was no longer effective in pollination. Hayase (I955) observed that cucumber pollen was viable prior to anthesis. The temperature of maximum viability was found to be 200- 25°C. Seaton and Kremer (I938) found that temperature was the most important factor influencing the dehiscense of the pollen sacs of cucum- bers. Minimum temperature for anthesis was 580 to 60°F, and for anther dehiscense, 620 to 63°F. Optimum temperature for anther dehiscense was between 650 and 700E. It was found that if the previous day had been cool, higher temperatures were required for anthesis and anther dehiscense. In August and September in East Lansing, Michigan, Seaton and Kremer (I939) noted that anthesis and dehiscense usually occurred between 6 a.m. and 9 a.m. On a very warm day dehisced anthers were found as early as h a.m. When temperatures did not reach 550F., anthesis and dehiscense did not take place. Mann and Robinson (I950) found that bees usually depleted canta- Ioup pollen supplies by mid-morning. Pollen grains were found to germi- nate within a few hours of being placed upon the stigma. Mann (I953) calculated that at anthesis both perfect and staminate cantaloup flowers possessed 8 to 13 thousand pollen grains. This number was reduced to 2 or k thousand by l p.m., and to one thousand by 5 p.m. Adlerz (I966) found the best fruit set in watermelon occurred when pollination took place between 6 a.m. to II a.m. 73 MATERIALS AND METHODS Observations of individual flowers Observations of honey bee visits to cucumber flowers were made during July and August I967 and I968 on research plots located on the Michigan State University Farm, East Lansing. The gynoecious F hybrid l Spartan Dawn, supplied with l0% monoecious Spartan 27 was planted in I967. They gynoecious hybrids Piccadilly and Spartan Progress were planted in I968. Plants were spaced six by eighteen inches in I967 by using a V- belt planter, and later thinning. In I968 plants were precision drilled at a 5 x l5 inch spacing. Plants were sprayed with Malathion to control bacterial-wilt, and weeds were controlled by mechanical cultivation and hoeing. Pistillate flowers were bagged the afternoon prior to anthesis or very early in the morning of the day of the bee visits (Figure 28). The bags were removed at the time of the visit observations. From one to I2 bee visits were allowed on each flower. After the specific rquber of visits were completed, the flowers were clipped shut with paper clips. Each flower was dated and numbered with a tag. These tags were used for identification at harvest. All fruit and flowers other than those observed were removed from the cucumber vines to prevent competi- tion (See Chapter I for a discussion of the effects of inhibitive fruit). Male flowers were present in sufficient numbers to provide adequate pollen. The following data were collected for each flower and its result- ing fruit: date of anthesis, number of bee visits, length of bee visits in seconds, node position of the flower, time of day, fruit shape, 7h constriction and curvature, number of seeds (I968 only), length and width of fruit at harvest, and date of harvest. During January and February of I969, limited observations were made in the greenhouse using Spartan Progress with SMR 58 as pollinator. Zero, two, ten and twenty visits were allowed to flowers. Bees were supplied from a small colony. Variable exposure of honey bees to cucumber flowers The effect of honey bee pollination at different times of the day was investigated by eXposing cucumber flowers for restricted periods of time during the day. It was possible to expose flowers for different periods of time in hours and at different one hour intervals throughout the day. The three methods used were: Exclusion Method - Bees excluded except for specific times. Uni- form groups of pistillate flowers were closed mechanically with paper clips at hourly intervals throughout the day to prevent pollination of flowers during these specific periods. Groups excluded late in the day received a full day's exposure to pollination, while those covered early in the day were pollinated from the time bees began to fly until they were covered. Inclusion Method - Flowers were available to bees except for specific time periods. Groups of pistillate flowers were covered with 2.5” square organdy bags prior to anthesis. 0n the day of anthesis, identical groups were uncovered at different times of the day and were left uncovered. Groups uncovered early in the day received a full day of pollination, while groups uncovered late in the day were pollinated only from the time of uncovering until the bees stopped visiting. 75 One Hour Exposure Method: In the third group, flowers were cover— ed prior to anthesis as were flowers in the inclusion method. These flowers were then uncovered for a period of one hour, after which they were closed with a paper clip. This method compared the effectiveness of pollination at different times of the day. The variety Spartan Dawn was used for the above studies. Spartan 27 was supplied for pollinator as 20% of the total plant population. All plants were fertilized and irrigated. Only one flower qu studied on each vine, which eliminated the problem of inhibition by other deve- loping fruits. Each flower was recorded by tagging the vine and bee- excluding clips were placed gently on the corolla tips to prevent mecha- nical damage to the developing ovary. RESULTS AND DISCUSSION Single Flower Observations Number of bee visits and percentage of fruit set During I967 and I968 single bee visits were found capable of producing fruit 53% of the time. As shown in Table 2h, there was little difference in the percentage of fruit set between single bee visits and multiple bee visits until nine or more visits, where upon fruit deveIOped 79% of the time. Statistical treatment of these data do not provide significant differences for the total column. However, a Chi-square comparison of flowers receiving single visits and those receiving 9 or ID visits was significant at the 0.05 probability level. 76 TABLE 2h Number of bee visits and percentage of fruit set 1 Number of Number of Number of Percentage Visits Flowers observed Fruit Set Fruit Set l A9 26 53.06 2 66 39 59.09 3 SI 26 50.98 a 45 30 66.66 s 38 23 60.52 6 l8 l0 55.55 7 IA 8 57.lu 8 ll 6 su.su 9 9 7 77.77 lO-l2 IS l2 80.00 X2 test of total table not significant at 0.05 probability level, however, significance was found when testing difference between I visit against lO-l2 visits. In January and February I969 greenhouse observations (Figure 3l), flowers received either 2, IO, or 20 bee visits. As shown in Table 25, those receiving 2 or IO visits did not differ from each other, develop- ing 3l to #0 percent of the time and producing an average of hh to #8 seeds. Flowers that received 20 visits developed 55% of the time and produced an average of 6h seeds. The controls, flowers open to pollina- tion the entire day, developed 93% of the time but did not contain any more seeds per fruit (Figure 32-35). The plants were all small and pro- duced small flowers, limiting the number of ovules per fruit (See the 77 .EmcmOLQ Lucmomoc m_;u wo mucmo .mco>om kw pom: .moo_:um ucoEo>oe co__oo 6cm moo—o Leumc___oo ommucoocoo mc_o:_oc_ .om .__mo c_ oo__aa:m Loom: .co_umc___oo embasoso co$ ommo c_ .mu_m_> won aco>oca cu oommmn mcozo_w oum___um_m .mN oc:m_m mu:m_a omsoccmoco ..m .muO_Q Locmmmmc v_m_w mom— NoEmcwimv mo>_; moo—0:2 .mN ‘I' 78 .mm .mu_m_> 663 o_ EOLw ucoan_o>oo boom __:w :u_3 topazozo “swam—aeoo boom ..3u .dm .co_umc___oa mo >mo __:m m mo u_:moc m mm ucoEoo_o>oc comm __:w ;u_3 V .u_m_> own oco Eccm mc_u_:moc mvoom 30m >co> cu_3 conE3osu .u_m_> 663 0:0 50cm .mm consaozu .Nm oc:m_u n wok—m; mun oub¢13.5 SLISIA 330 on; an" .33 .33.; o -oo.. 79 discussion on thé length of ovary and number of ovules in the first chapter). There was not a parthenocarpic tendency, as indicated by the absence of fruit development in the non-pollinated group. TABLE 25 Percentage fruit set and average seed counts from 2, IO, and 20 bee visits. Winter I969. Number of Total Number Number Percentage Average Number Visits Fruit Dead Alive Alive Seeds 2 IO 6 h #0 #8.3 a l0 l3 9 5 3| hh.3 a 20 . 9 h 5 55 6h.6 b Open l2 I ll 92 58.0 b Never Open 5 5 O O 00.0 c Values with the same letter are not significantly different. Total F(3,20) = '93.8 Number of bee visits and fruit shape Table 26 provides data on the influence of varying numbers of bee visits on fruit shape for I967 and I968. There was no correlation between the percentage of perfectly-shaped cucumbers and the number of bee visits to flowers. This points out the difficulty of isolating factors which cause poor fruit shape and the role of pollination in affecting fruit shape. More comprehensive tests, possibly in growth chambers, might be able to correlate seed numbers and fruit shape under 80 specific environmental conditions. TABLE 26 Number of bee visits and shape of fruit, I967 and I968. Number of Visits Perfect Fruit Imperfect Fruit Percent Perfect l 23 3 89 2 36 3 92 3 23 3 89 h 22 8 73 5 29 h 88 6-8 2| 3 87 9-l2 I8 I 95 X2 comparison indicates no significant difference at the 0.05 probability level. Number of bee visits and seed count Table 27 shows the effect of the number of bee visits to a flower on the number of seeds in the resulting fruit. No significant differences were observed except for an increased number of seeds with more than l0 visits. This trend was found by Shemelkov (I957) on cucumber, Alderz (I966) on watermelon, McGregor E£.21' (I965) on cantaloup, and Verdieva and Ismailova (I960) on squash. In the field, one or two bee visits produced fruit with an average of 258 seeds. This indicated that a great amount of pollen was placed on 8l the stigma in one or two visits, resulting in high fruit set and good grading quality. In the greenhouse, where less pollen was available, flowers which received one or two bee visits produced fruit which con- tained a range of 6 to 9| seeds, although all fruit were well shaped. All fruit in this study were short, indicating fewer seeds than plants observed in the field. This implies that while two visits produced fruit, the number of seeds which developed was dependent upon the num- ber of pollen grains actually transferred by honey bees. The dependa- bility of pollen transfer could vary with (l) the staminate to pistil- late flower ratio, (2) localization of staminate and pistillate flowers in the field or (3) the timing pattern of staminate and pistillate flowering. If the availability of pollen was inadequate in time or location, the number of visits needed would be much greater than with a constant, balanced proportion of staminate:pistillate flowers through- out the field. TABLE 27 Effect of the number of bee visits on the number of seeds per fruit Number of Visits Numgigm?:e:rUit 2:2529: 3:??zgczf l- 2 8 258 j; 27 a 3- ll 13 257 3: 31s a 5_ 6 II 2714-: I6 a 7_ 8 9 23h :- 32 a 9-l0 5 219155 a ll-l2 5 320 i 33 b Values followed by the same letter are not significantly different at 0. 05 probability level. 82 Correlations between number of seeds and shape were determined by counting seeds of cucumbers grown under field conditions. Table 28 shows the average number of seeds found in perfect, nub, and neck-shaped cucumbers during I967. In agreement with Seaton (I937) the perfectly- shaped fruit had approximately twice as many seeds as the necks, and three times as many seeds as the nubs. Table 27 indicated that even one or two bee visits produced more seeds than found in nubs or necks. Higher seed counts increased the probability of the fruit being perfect- ly shaped. However, fruit receiving a controlled II or l2 visits in the field averaged 320 seeds per fruit, significantly less than the average 393 seeds found in the open-pollination fruits used for comparison. I TABLE 28 Average number of seeds obtained from three shapes of fruit . Fifteen fruit randomly selected in each class. Shape Average Number of Seeds Perfect - No constriction 39k a Neck — Constricted at stem end 205 b Nub - Constricted at flower end I36 c IThis table does not imply that the number of seeds determines shape of the fruit. ‘ Values followed by the same small letter are not significantly different at the 0.05 probability level. 83 We can summarize with two basic observations. One or two visits was capable of supplying enough pollen to form a normally-shaped cucum- ber. Up to l2 bee visits per flower did not produce as many seeds as perfect fruit picked at random, which represented a full day of pollina- tion. Length of bee visits and resulting number of seeds The length of time individual bees visited cucumber flowers varied greatly. Data based upon the total length of time a flower was visited, disregarding the number of visits, are shown in Table 29. Comparison of the total length of bee visits with resultant fruit de- velopment provided no significant differences. In assessing these re- sults it must be kept in mind that the first visits averaged much longer than subsequent visits. TABLE 29 Total length of bee visits and resultant fruit development W m Length of Visits Total Number of Fruit Percent in Seconds Observed which Developed Developed 0- 30 25 I7 68 30- 6O 96 SI 52 60— 90 l6h IOO 6| 90-I20 20 IO 50 120 + l3 ll 85 Developed fruit £3. Total - no significance X2 = l.5l with h degrees of freedom. Dev-eloped fruit x_s_. Non-developed - no significance X2 = 5.35 with ll de- grees of freedom. 8h The average number of seeds produced is correlation with the length of visits, as expressed in 30-second time intervals, shown in Table 30. No important differences were found in the resulting number of seed counts except for the O to 30-second interval. The relatively poor yield and low seed count in the O to 30-second interval group would again indicate need for more than one visit per flower. TABLE 30 Total length of bee visit and average number of seeds 't6flgth of ViSits in Seconds Number of Fruits 5 'Average Number of seeds 0 - 30 3 l33-3 a 30 - 6O 20 262.l b 60 - 9O lh 266.0 b 90 - l20 7 26h.7 b l20 + 7 266.3 b Values followed by the same letter are not significantly different at the 0.05 probability level. Time of day of pollination Table 3l summarizes I967 and I968 data on the effect of time of day on pollination. It was hypothesized that the volume of pollen in staminate flowers could be depleted or could lose viability during the afternoon, consequently late afternoon pollinations might not as succes- ful as early day pollination. The percentage of fruit development 85 dropped after 2:30 p.m. There was little difference in the average number of seeds obtained from fruit pollinated during different time intervals. Fruit pollinated between l0:3O a.m. and 12:30 p.m. did pro- duce slightly more seeds, but the differences were not significant. Thus, it appeared that the amount of pollen or the loss of viability were not limiting factors in later afternoon pollinations. Either poor receptivity of the stigma or less favorable time for pollen tube growth or some other factor may have been responsible for reduced fruit set. As shown in Table 3|, cucumber pollen was evidently available as soon as bees started morning flights. This agrees with observations of Seaton and Kremer (I938), who indicated that staminate flowers reached anther dehiscense at approximately the same time that bees began to fly. Also see Figure 36. Variable Exposure of Honey Bees to Flowers Minimum hours of exposure to pollination for maximum fruit set The first method of excluding bees provided information on flower exposure time needed for effective pollination. This test seemed to in- dicate rather conclusively that the number of fruits set in any one day varied directly with the number of hours that bees had access to the flowers (Table 32). No clear-cut influence on shape was evident, again indicating the difficulty of determining the influence of pollination on shape and relationships with other stress factors affecting shape. Table 33 provides further information on the influence of length of pollination time on set of fruits. The period from 5:30 p.m. to 8:30 p.m. (three hours) resulted in h0% fruit deveIOpment. The level of fruit set increased as the exposure period increased, reaching 7l% in 86 IO hours. The rate dropped to 23% at ll hours, but this may have been resulted from confusion by field help of day-old flowers. In general a full day of pollination favored fruit development. TABLE 3| Effect of time of day of pollination on fruit development and the number of seeds per fruit ‘ Time of day Number Number Percent Number Fruit Average of pollination Flowers Fruit whi h DeveIOped with Seeds Number Pollinated Developed Counted Seeds 8:30- 9:30 ll 6 55.0* 2 I08 9:30-l0:3O 60 I9 3l.5 9 297 IO:30-ll:30 I33 39 29.2 8 25h ll:30-l2:30 8O 39 #8.8 7 272 l2:30- l:30 lhl #6 32.5 0 -- l:30- 2:30 262 79 30.0 ID 238 2 30- 3:30 155 2| 13.5 l0 ‘ 23h 3:30- “:30 I29 22 l7.0 h 276 #:30- 5:30 IOI 29 23.8 0 -- *Days when overnight temperature did not drop below 65°F, visit activity began before daybreak. l Values significantly different at 0.05 probabglity testing number pollinated 22' number devel0ped with X test. 2Values significantly different testing number of seeds with time of day of pollination. F(6 #3) = -l9.66 87 TIME OF DAY OF POLLINATION AND 9° PERCENT DEVELOPMENT 2 80 § gm 3 .o 3 ,f 40 f. E30 \ \ 20 / ID 830 930l030 ll30 l230 I30 230 330 430 Time of Pollination Figure 36. Relationship of the time of day of pollination and the PErcentage of fruit deveIOpment. 88 TABLE 32 Controlled exposure of flowers to bees. Effect of time of day and length of exposure to pollination on fruit formation. Inclusion Method Tlme F‘°We' # Flowers # Alive % Alive # Perfect % PerfeCt was opened* on Total 5:30 p.m. 10 h #0 2 20 #:30 I9 9 #8 7 37 3:30 19 7 . 37 7 37 2:30 l8 ll 6l 9 50 l:30 l9 9 #8 7 37 12 30 20 l2 60 8 ho ll:30 a.m. 20 ll 55 II 55 lO:3O l7 l2 7l IO 59 9:30 l7 A 268 6 2h *Flowers were not reclosed. aConfusion of old day-old flowers with new flowers may have taken place during the morning hours. 89 TABLE 33 Controlled exposure of flowers to bees. Effect of time of day and length of time of exposure to pollination on fruit formation. Exclusion method. Time flower Hours of Number Number %.Alive Number % % Perfect was closed Exposure Flowers Alive Perfect Perfect of the to Polli- of total Developed natiOn. 8:30 a.m. 0-2 25 O O O 0.0 0.0 9:30 3 25 O O O 0.0 0.0 IO:3O A 4| 3 7.3 2 h.8 66.7 ll:30 5 2h 3 l2.5 O 0.0 030 l2:30 p.m. 6 hl 9 22.0 6 lh.5 66.7 l:30 7 2h 6 25.0 3 l2.5 50.0 2:30 8 hi l9 h6.5 l5 36.5 79.0 #:30 IO A6 32 70.0 l9 lil.5 59.5 5:30 ll 26 2l 8I.O l5 58.0 7|.5 90 80 7O 60 50 4O 3O PERCENTAGE FRUIT DEVELOPMENT 20 10 Figure 37. 9O EXPOSURE PERIOD AND FRUIT DEVELOPMENT EXCLUSION INCLUSION -—— f I I — 2 3 4 5 6 7 8 9 IO NUMBER OF HOURS FLOWERS WERE EXPOSED TO BEES. Relationship of thelength of exposure to bee pollination and the degree of fruit set and development. 9l Time of peak_pollination activity Table 34 shows results of one-hour periods of exposure to bees. The period of lO:30 to 3:30 produced a range of h2.9% to 68.8% fruit development. Early morning and late afternoon pollinations were less effective (7.l to l8.8 %). The period of greatest fruit set coincided with the period honey bees were observed making the greatest number of visits to cucumber flowers. TABLE 3h Time of day of pollination - Timed open study. Time of one Number of Flowers Number of Flowers Percent hour of Exposure in Sample that Developed Developed Anthesis to 8:30 l9 0 0.0 8:30- 9:30 l6 7‘ h3.8 9:30-IO:30 l6 3 l8.8 lO:30-ll:3O lb 3 6 h2.9 ll:30-l2:30 l6 9 56.2 l2:30- |:3O l7 ll 57-9 |:30- 2:30 l8 II 68.8 2:30- #:30 ID I l0.0 #:30- 5:30 14 I 7-' 92 100 TIME OF DAY OF POLLINATION AND ’° PERCENT DEVELOPMENT 80 .— Z I” 2 m 70 O —I I“ > HI 6‘) o t 2 50 II- .— 12 Ill 40 u 8 I“ " 30 20 10 830 930 IO3O ”301230 I30 230 330 430 530 TIME AT END OF ONE HOUR EXPOSURE TO BEES Figure 38. Relationship of one hour of exposure to bee pollination at different times of the day and the degree of fruit deveIOpment. .o_o_w mg“ :2 oo>c6mno .oo>OEoL x>_mo 6cm m__0coo Os“ mo “Ema cu_3 yo: mm: m_ck .om_:o_ncoo co co__oq Longsoso Lozo m Opmc_Emum Sop» cmoooc mc_>oeoL oou wo “cacao __mEm m £u_3 Omno£c00cm c_ mom ._ «V V .. . p. In... mow": . . ‘ o 9‘. ‘1 n ... I o .. .. ... . ... . \ N). s . .u_m_> co_omc___oa o>_uoommo Lo» co_u_moo mom .0: .cozo_w Oomc_Emom um mc_>_ccm owm .mm oL:e_o . ‘ 3 9 IV PERCENTAGE STAMINATE FLOWERS OPTIMUM FOR FRUIT YIELD AND SEED COUNT LITERATURE When Peterson released the gynoecious hybrid variety Spartan Dawn in I962, he was faced with the problem of providing pollen for the predominately female plants. Following tests with bees in cages, he recommended blending l0% seed of a normal monoecious variety (e.g. Spartan 27) with the hybrid seed to act as a pollinator (Peterson and De Zeeuw, I963). Commercial varieties developed since the introduction of gynoecious hybrids have followed this recommendation (Figure #3 and ##). A more comprehensive study of pollen supply needed to provide optimum pollen distribution in cucumber fields was needed. The practi- cal solution of the problem has been complicated by the fact that most zuammercial varieties commonly termed gynoecious cannot in fact be called lgynoecious because of ineffective breeding methods resulting in many staminate flowers. Counts of varieties in commercial fields produced mixtures with #5 to 60% of the plants monoecious in nature. Any dis- cussion of the percentage pollinator to add to hybrid seed should con- sider this factor. Large, variable deviations from the gynoecious cha- racteristic Observed in commercial varieties made it necessary to count 9# 95 .mp0: >co>o >_LmOc um .co_uo:60co c630.» Oum___um_o ucmc_EoU ozu ooosooca Ocm mcozo_m Oumc_Emuw .OmJOLCOOLm Ouoz .ommo o_o_ m c_ moms uno;u_3 czocm c_ “cm—a m:o_ooocoE _mo_a>h .J: oL:m_u czmo cmucmam o_cn>; u m:o_600c>o .ma Ocnm_u .li fir . . ..«g ..‘VR. . . . WV . \J - LL ,3 v . I o t. F \J u 4.7. 96 the actual staminate to pistillate flower ratio for certain aspects of this research. Several factors influence the staminate to pistillate flower ratio of cucumber. Genetic, varietal, and environmental differences were reported by Currence (I932), Atsmon E£.El' (I965), and Miller and Ries (I958). Plant age affects flower production in monoecious varie- ties (Currence, I932). Typically, the first flower of mOnoecious varie- ties are staminate, gradually shifting to a stronger pistillate expres- sion. Seasonal variation in the flowering pattern was observed by Edmond (I930). During the period of December to April, plants produced a staminate to pistillate flower ratio of 5:l to 8:l. During the summer this ratio increased to 50:I to 59:I. Light intensity was believed responsible for this shift in sex expression (Thadjens, l928b). In a monoecious variety, the staminate:pistillate ratio shifted from l#:l to 7#:I when plants were moved from minimum to maximum light. MATERIALS AND METHODS ‘l967 Cage Study Group I: A series of duplicate plots were planted containing 5, l0 and 20 per cent monoecious variety Spartan 27, in the gynoecious hybrid Spartan Dawn. The plots were spaced at l8'I x 6”, which was equivalent to #2,000 plants per acre. Each of the three sets of plots contained: (I) four plots with no cages and exposed to open pollination (estimated at Ifllree colonies per acre), (2) two plots caged with bees throughout flowering, and (3) two plots caged to exclude bees. 97 The plots were l2'x8'X6' with approximately lOO plants per cage. Initial flowering started during the first week in July. All fruits were hand harvested. Yield data were recorded on the basis of the average number and weight of fruit per plant. All fruits were graded for shape into one of the four classes: perfect, neck, nub or crook. Group II: Tests were repeated on a later group Of plants that began bloom during the first week of August. NO exclusion cages were established, otherwise, the plot design was identical to Group I. I968 Isolation Plots In order to try an alternate method to test the efficiency of different percentages of pollinator, four plots were established in isolation to allow free bee movement without the restraint of cages. Plots were located in a line 500 feet apart; thus 2,000 feet separated the first and last plots. The plots were located in a new planting of alfalfa. This seemed to be sufficient isolation to insure against the transfer of pollen from one plot to another. Each plot was divided into four subplots of two rows each, which measured #2 sq. ft. or l/IOOO of an acre. Each subplot contained approximately 50 plants or a rate of 50,000 plants per acre. The blend of gynoecious and monoecious varieties was recorded and the ratio of staminate to pistillate flowers was determined from flower counts. The number of staminate and pistillate flowers produced each day were counted during the early flowering period of July l7 to 3|. Plant growth was uniform in each of the four plots. 98 Staminatezpistillate flower ratios for different varieties Several commercial varieties of interest in this research were tested in the greenhouse during I967 and I968. All plants were grown under similar conditions of temperature, water, soil mixture, etc., with the main variable being the time of year. Powdery mildew and plant mites were controlled by periodic spraying and fumigation. Daily flower counts were started at the appearance of the first flowers. Each variety was observed for a minimum of IS days. This period represents the time normally needed for fruit to develop which would be harvested in a destructive harvest. Percent fruit development per day of flowering From July l8 to 3|, I968, flowers were selected and tagged daily in each of the four plots planted in isolation. These tagged flowers were later checked for fruit production. In a separate plot 25 flowers were tagged daily to serve as a control, providing information about the level of fruit set under optimum conditions. Only plants without developing fruit were used in both controls and whenever possible in the isolation plots in order to eliminate the effect of inhibitive fruit. However, no fruit were removed from the isolation plots. The purpose of this study was to examine the daily fruit setting patterns of the different staminate:pistillate flower ratios. 99 RESULTS I967 Cage Study Plants in cages from which bees were excluded produced practical- Iy no cucumbers, although the plants produced measurably larger leaves and flowers. Plots with bees produced from 90 to I60 fruit per IOO cu— cumber plants comparted to only # fruit per IOO plants when bees were excluded. No significant differences in yield occurred in plots with bees, as a result of variation of percentage of pollinator level (5, l0 and 20 percent). This was true for weight yeild per plant (Table 35), for number of fruit per plant (Table 36), and fruit shape (Table 37). TABLE 35 Weight yield per acre and percentage pollinator I967. Based upon 50,000 plants per acre. Treatment . Weigth yield in lbs. for different pollinator Open Pollination Five Ten Twenty a Group I I3,350 l2,000 l2,650 Group lla 3,050 3,050 3,l00 Plots Caged with Bees Group Ia IO,#OO 9,950 8,000 Group Ila 2,850 3,200 3.250 Plots Caged without Bees Group Ia #50 0 350 aGroups I and II should not be compared due to lack of homogeniety at harvest time. Statistical variation within rows not significant at the 0.05 probability level. IOO TABLE 36 Average number of cucumbers produced per plant and percentage pollinator. I967 m Treatment Percentage of Pollinator Open Pollination Five Ten Twenty Group Ia l.60 I.57 l.62 Group lla 0.98 0.99 l.33 Plots Caged with Bees Group Ia I.3# 0.96 0.9l Group lla 1.06 0.89 0.89 Plots Caged without Bees Group la 0.0A 0 0.03 aGroups I and II should not be compared due to lack of homogeniety at harvest time. Statistical variation within rows not significant at the 0.05 probability level. lOl TABLE 37 Percentage perfect, neck, nub and crook shaped fruit from plots of different percentage pollinator I967 Open Pollination Shape Percentage of Pollinator Group I Group II Five Ten Twenty Five Ten Twenty Perfect 50% 56% 55% 38% 52% 59% Neck 27 25 26 5 5 6 Nub l7 l5 l6 5# 39 30 Crook 5 # # # l0 3 Plots Caged with Bees Group I Group II Five Ten Twenty Five Ten Twenty Perfect 66 5# 55 57 5# #2 Neck 23 2# 27 6 ll 9 Nub I9 l7 I6 33 30 36 Crook 6 5 2 # 5 3 l02 1968 Isolation Plots Highest yield and dollar value were obtakled from a seed blend of 90% Spartan Progress and lO% SMR 58, which produced an S/P ratio of 0.5:l during the first l5 days of bloom. _These studies provided data (Table 38) on the variability on the staminate:pistillate flower ratio of varieties ostensibly gynoecious hybrids. For example, during a l5 day counting period Piccadilly pro- duced 73% staminate flowers while Spartan Progress produced only 2.7%. During the same period Spartan Progress blended with l0%.SMR 58 yielded 33% staminate flowers, and blended with 50% SMR 58 yielded 5#% staminate flowers. TABLE 38 Staminate and pistillate flower production of different varieties and variety mixtures during the first l5 days of bloom. Variety Plot Pistillate Staminate Percent S/P Number Flowers per Flowers per Staminate Ratio Plant in Plant in IS days l5 days Piccadilly I00% I 3.6 9.7 73% 2.7:l Spartan Progress 50% || 6. . 4% l 2.] SMR 58 50% 7 7 9 5 ' Spartan Progress 90% III 16.] 7.8 33% 0.5:] SMR 58 IO% Spartan Progress IOO% IV ll.8 0.3 3% 0.03:] I03 Table 39 shows the average number of pistillate and staminate flowers produced per plant per day. This data is of value in assessing pollen requirements and pollen availability. It is also of interest in calculating flower pOpulations per acre and the bee population necessary to provide adequate pollination. In plots I, II, and III Of the I968 pollinator tests there was an increase in staminate flower production with the age of the plants (Table 39, #O and Figures #5-#7). Plot IV (IOO% Spartan Progress) con- sisted of a highly gynoecious variety and did not exhibit this pattern. Therefore, the increase was largely due to the monoecious plants mixed with Spartan Progress in plots II and III. The dollar value of each of the four plots is shown in Table #l. The maximum yield was produced in plot III, 90% Spartan Progress and l0% SMR 58, which produced 33% staminate flowers during the first l5 days of flowering. The yield of plot IV (I00% Spartan Progress with 2.7%.staminate flowers) was negligible. Somewhat surprising was the fact that plot I (l00% Piccadilly with 73% staminate flowers), yielded much lower than Plot III in dollar value of the first and second harvest. The number of fruit produced per plant is also shown in Table #2. In the first harvest the highest yield was in Plot III, with l.6 fruit per plant. Plot II produced 0.6l fruit per plant, plot I produced 0.29 fruit per plant, and plot IV produced only 0.03 fruit per plant. The same trend was observed for these plots for the second harvest, with plot III consistently producing more than the other three plots. Yield data of Table #I and #2 and Figure #8 show the variation in the number of fruit produced per plant in the I968 percentage polli- nator plots. Plot III with approximately l0% monoecious variety and IO# —n o\oo_ on :3. Xooooiom on on an .co_uo:OOLQ L63o_w Oumc_Emum co oco_n comm mo moses—mc_ f on 3;...“ o Koondoiom X81533... Nu 0N nu Va nu «N pm on o— @— 8:63 c2229.... Ooom 205:5 *0 30.x. E scion—OPE $30.“. 305805 .m: Oc:m_m k. :3 mm on wk 00— na— 92 3: RN laqwnN ’0 “Duo” atougwotg I05 .co_uo:OOLa Lozo_w Oum___um_a co oco_n comm mo oozes—we. .0: Oc:m_m o\oo_ on «so o\oo... a... as; x0e 3252.. cutoam AXqu 3300i cutoam XOO— 3.233... pm on ON on Nu cu nu 1N nu «N 3 ON O— a. N— _\\\\ ll. \3 \\.\\ ..l.u4.\... IIIIIII7/ \ / x on o / ‘\ / \ esco- coo 00.0 .I‘ o qqqqq / Os .. / \ .. Os a. on. o... co co. /\ so. can .00. o on o s... 0.00 .000 0° on . OO— cOZuonOE COBOEBOZZmE .a. 0: Oc— mmou ”Plume Io “QWON 106 .o_umL Lm30_w mum___um_auoumc_smum co tam—n vOMm mo mucus—m:— .m: mL:m_u ....I....OI‘.‘. ' ' '-- o\oo_ :52... o\oo... on :3 o\oOO_ ....ooi cotoam W\oOm 30.00.... eaten» o\oOO— 2:10qu __n arm ‘nfl n.N nufl pa 9 — k — >uqan Ju.¢aoow..oo\ \ x _ x / x \ / \ u /\\//l\ 0:3. 326: fuzzmauflosefim n 0H”! 4:9 107 .mLum Lma u_3Lw mo LonEJC mcu co o_umL Lm30_m mum_._ym_anmumc_Emum $0 Hummwm .m: mL:m_m o..u<\2:o.._ ooodn .a 0.05:.m Antennae... c: 0.0( .... :3:— &o 5.36:2 on. ma— 00— an on ma nnnnnnnnnnnnnn ccccccccccccc .............. nnnnnnnnnnnnn oooooooooooooo ------------- oooooooooooooo ....... u a o > .. O I n c o u. m : o> . a I u a ._ _ u H.H.H.H.H.N.n... ......... . — .. n 0.0 ................................................................ “mnu."mum”...u..n.u..u...u.nu.nwwwwwwwwww"Maw.“u."”mum.m.H.”Wm."n.um.nn.um.nm.um.m.um.nm.um.mm.Wm."m."mmmmm."mm."mmmmmmwwwwwwwwmmm — "n . o ------------------------------------------------------- nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo ---------------------------------------------------------------------------- ----------------------------------------------------------------------------- ---------------------------------------------------------------------------- ----------------------------------------------------------------------------- uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo .................................................. . ...“.Emm. — u N .— c ............................................... nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn -------------------------------------------------- ooooooooooooooooooooooooooooooooooooooooooooooo .............................................. ooooooooooooooooooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooooooooooooooooo ............................................... ................................................... .............................................. nnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn ................... . cocoa-coo coco-coco. oooon ..... (J’JKIJ’JI? ffvfrvvf??? .ftddq5555 9........... mv_m_> c0 Ozom .030: 0:3. ._\m 30.:3E635E2m ...o .uoim 108 33% staminate flowers had a significantly higher yield than other plots. This indicated that under machine harvesting conditions, a highly gynoe- cious type of plant blended with 10% monoecious seed produced 33% stami- nate flowers and provided the best yield for the four seed ratios tested. As shown in Table #3, plot III, which was superior in other yield data, produced the highest percentage of short, non-curved, non-constricted cucumbers for both the first and second harvests. As discussed in the first chapter there was a correlation of ovary length and the number of ovules, and younger plants tended to produce shorter fruit, correlated to smaller flower and ovary size at anthesis. The shorter fruit were less constricted and less curved in shape. These results indicate that Plot 111 had a larger number of pistillate flowers pollinated during an early stage of flowering, while the other plants did not have pistillate flowers present. Thus, the shorter, more perfectly shaped fruit were most common in Plot III. Staminate: pistillate flower ratios for different varieties Table 4h shows the extent of the variation in sex expression of flowers of several gynoecious hybrid and monoecious varieties. The S/P ratio was found to differ with the time of year. For instance, Piccadil- 1y planted May 20 in the greenhouse produced an S/P ratio of 5.93:1 while plantings two months earlier produced an S/P ratio of 1.20:1. This change was characterized by a reduction in pistillate flower production, with staminate flower production remaining constant. Edmond (1930) and 'Tie djens (1928) related this change to increased light intensity. l09 oo. mm. :m. 0:. mm. :N. we. ow. .m oo. mm. mm. :m. nm. :N. mo. Nu. om oo. mm. :0. mm. on. :N. mm. w_. mm «0. mm. o:._ Nm._ :~._ mo._ ~_._ wN. mm .0. mm. m:._ No._ m... o... m... mu. RN .0. :m. w:. .m. oo. 0:. mm. mm. mm mo. mm. mm. mm. mm. mm. mm. MN. mu 5. «2 mm. 2.. mm. mm. mo; mm. .3 No. mm. mm. o:._ ::. No. mm. Nm. mu No. _m. __. mm. om. mm. mm. om. NN :o. mm. m_. mm. w_. mm. mm. .m. _N m_. mo._ ::. _N._ mw. :N. :m. mm. om o_. mm. m_. mm._ m_. mm. m:. .N. m— oo. mm. :_. .m. mo. mu. m_. o_. m. oo. 05. oo. om. _o. mm. oo. oo. N. mum:_Emum oum___um_m oumc_Emum oum___um_m mumc_Emum oum___um_m oumc_Emum oum___um_m >_:s Xb_ mm mzm me mm mzm Nbo. mmbcmOLm cmucmom Xbm mmocmoLm cmucmam Xbm mmocmOLm coucmam Xbo_ >___omuu_m mama >_ ___ .Eoo_n mo m>mp m. umcwm osu mc_csv >mv sumo mmczux_e >uo_cm> pcm mm_um_cm> «cocowm_u mo co_uu:voca L030.» oum___ym_a cam oumcmEmum mm 59:. llO oo.o o... mn.~ m~.m .m oo.o :w._ nm.~ nu.~ om oo.o :~._ Nm.m mu.: mm 00.0 _m.o w... mm.m mu No.0 Nm.o wo._ cm.: mm 00.0 mm.o w:._ um.m mu mo.o om.o ~m.. mm.m mm 00.0 mm.o om._ mN.m :N No.o _~.o mm.o ow._ mu no.0 ...o :m.o nm._ «N mo.o m_.o m:.o m... _N m_.o w~.o mm.o No._ om .o.o ~_.o mN.o no.~ m— oo.o m_.o mu.o ou._ m— oo.o oo.o mo.o oo.o n. :2, xbo_ mmmcmOLm cmucmam Rom mmocmomm_cMMLMnm xbm mmocmOWMmcmmcMum $00. >___nmuo_m puma >_ ___ __ _ .mOLaume >ummLm> Ucm m0_u®_.._m> ucmmewmfl r: m0_um.h LQZO—w Qum—__um_n_umumcmEMHm >—_mo o: wu—m<.._. lll TABLE #1 Dollar value of percentage pollinator tests in isolation plots - 1968. Dollar Value per Acre Based upon 50,000 Plants per Acre Variety Mixture First Harvest Second Harvest Total Value Piccadilly 100% 3h.h0 226.08 260.hl Spartan Progress 50% SMR 58 50% 68.10 286.80 35h.90 Spartan Progress 90% SMR 58 10% 126.90 282.78 h09.68 Spartan Progress 100% 3.h5 51.75 55.20 Differences significant at 0.01 probability level. TABLE #2 Number of fruit per plant of percentage pollinator tests in isolation plots - 1968. 4 L L = 3 Variety Mixture First Harvest Second Harvest Total Fruit/Plant Piccadilly 100% 0.29 1.12 1.#1 Spartan Progress 50% SMR 58 50% 0.61 1.66 2.27 Spartan Progress 90% SMR 58 10% 1.60 2.59 h.19 Spartan Progress 100% 0.03 0.35 0.38 Differences significant at 0.01 probability level. l12 m_ __ N m ocm>mm :m mm m. N v__z _m mm mm mm 6:02 ocsum>cau : m : : oco>om «w m. w : p__z .2 mm mm om ocoz co_uu_cumcou N. m o. a 95.. mm .5 mu .m .mecoz o_ :N :_ :_ acocm mamgm mo:_m> ommucouLom umo>cmz pcoumm II —N w— 0N 0L0>6m .. mm m: m: u__z in 0: mm mm 6:02 ocsum>cau u- e um mu oco>om in _N m. mm p__z .1 Nu RN mm 0:02 co_uu_cumcou i t .3 .4 m8.. -- mm om mm .mecoz 1 o. m m :2; mo:_m> ommucoocom umm>cmz umc_m mamsm Nbo. mmoLmOLm cmucmam >_ xa. mm aim New mmocmoca coucmam xbm mm aim me mmmcmocm cmucmam fico. >___omou_m u_um_couumLmzo mamcm mom. i mucoEumoLu >um_cm>\vco_m Loumc___oa ucocmwm_v csow EoLw u_:cw mo mamzm m: m4mmz : mm mzm _~.m_ m.o~ a._ o. mom. .N >mz mso_um0coz RN cmutmam No.o ..o o.m mN mom. .oN >mz m:o_000c>u omm :m: 00.0 o.o m.o ow wwm. .ou >mz : mmmcmocm cmucmam :N.o “.6 ..m o. mom. .QN >mz : czmo cmutmam oo.o 0.: 0.0 mm mom. ..ma : czmo cmutmam mm.m m.m m._ 6N mom. .oN >mz : >___umoo_a ~o.~ ..o_ .o.m o_ mom. .N >mz : A.ou Nuoc_m:v >___nmuu_a _~._ m.~ ~.m o. mom. .N >mz : Aummm _ e:_umzv >___vmuu_a o~._ m.m N.“ ow mom. mc_cam : >___emuu_a mo.~ :.m ..s o: mom. mc_cam mmwflmwnchm >___umuu_a m>mo m>mo o_umx m. c_ “cm_a m. c_ acm_a mucm_a mum___um_m Loo mcozo_m coo mcozo_u mo consaz oumo mc_ucm_m oa>h xom >uo_cm> guessoou "oumc_Emum oumc_Emum oum___um_m .nnuuuunununnuuui lipui ucmcmmw_p c_ soo_n mo m>mp m. umc_m mzu mc_c:p pouavOLQ mcmzo_w oum___um_a pew oumc_Emum .cmo> ecu mo mms_u ucmcowm_o um omsoccomcm ecu c_ czocm mo_uo_cm> :: m4mmp mzu co» 4<_hzmhom uz.h_:xu esp oumu_oc_ mucm_a omoch .oo>oeoc who; u_:cm o>_u_a_;c_ .muo_a cozuo ox__c: .cOumc___oa $0. gum; cZOLm mucm_a who; mcozo_w _0cucoUm as. o:_ as. ca. com ummmmu .mHOP o. o. o. o. mu >mu cog vommmu whose—m o 0. cm 0N Nu .m o o. o: o_ mm om 0 cu o: 0. mm mm o ow ON ou in mm o c: on cm 1- RN 0 ON o: ow mm ow o o. om om om mm o om om o~ oo. :N o o. o_ 0: cm mm 0 cu oo 0. cm mm o. om on om ow _N o. om om om :w on c on ow 0m um m. o co ow o ow . m— >_:w No. mzm xom mm arm m_ocucou ---: Nbo. mmmcmOLm cmucmam me mmmcmoLm cmucmam me mmocmOLm cmucmam goo. >___pmuo_m wou_n_cc_c: oumo g mom. . mummy ecumc___oa ommucoocma c_ >mo Loom u_ncw Ouc_ ooao_m>mo zo_cz mcmzo_m oum___um_a mo ommucoume m: u4m0<< cm__on_ .m: 0L:m_u nu. On. J OC'UJn hi nun )0 00;. urns nu.— 19.5 on.— nu.— IU"|d 12# plant in the plot 0-25 feet from the pollen source, compared to 0.#1 fruit per plant produced in the plot 200-225 feet from the pollen source. Furthermore, the rate of decrease was rapid, with the 25-50 foot plot producing 1.09 and the 75-100 foot plot producing 0.57 fruit per plant. This indicates that it would be most advantageous for maximum yield to have the pollen source within 25 feet of flowers needing pollination. TABLE #7 Pollen Gradient Study 1968. Dollar value of fruit of 225 plants and average number of fruit per plant at incmeased.distances from pollinator source. Distance from Dollar Value Dollar Value of One Average Number Pollen in Feet of 225 plants Acre at 60,000 of Fruit/Plant Plants/Acre 0- 25 $1.05 $281 1.57 25- 50 0.77 206 1.09 50- 75 0.75 201 1.00 75-100 0.39 103 0.57 100-125 0.#2 111 0.60 125-150 0.#7 12# 0.6# 150-175 0.39 103 0.59 175-200 0.50 133 0.69 200-225 0.33 89 0.#1 225-250 0.#7 126 0.72 Isolation Plot 500 0.21 56 -- 125 Fruit quality: Although no quantitative data were collected to conclude this on a mathematical basis, there was a tendency for the fruit harvested from the plots farthest from the pollen source to be curved in shape. Often, the extent of this curvature was slight, making the factor difficult to classify. No other significant differences in quality were recorded. Seed counts: Table #8 and Figure 50 show seed counts from groups of ten cucumbers from plots located at different distances from the pol- len source. Fruit at zero feet averaged 216 seeds, at 75 feet averaged 96 seeds, at 150 feet averaged 78 seeds, and at 500 feet averaged 56 seeds. TABLE #8 Average seed counts from mature fruit grown at increasing distances from pollinator plot. a Distance from Pollen Source in Feet Average Number of Seeds 0 216 75 96 150 78 500 56 Ten cucumbers per sample Level of fruit set: Table #9 shows that there were two major days (July 19 and 20) of fruit set in tagged plants in the gradient strip. 126 .ooL30m cm__oa oz“ EoLm moocmum_o mc_mmococ_ um u_:L» c_ pouzooLa moomm 00 Lucas: 0:» .0m mcam_u oo.:om cmzom EOE”. 3020.0». ::.". *o mommm *o LonEaz o m a o o n o n _ o o . o n ::.:ng no; UOIIOd 127 On other days, zero or one fruit deve10ped out of 25 flowers tagged. Because several male flowers were noticed on July 19 and 20, it is sus- pected that these dates contributed largely to the fruit set and the seed production observed throughout the gynoecious plants located at distant points from the pollen source. TABLE #9 Percent fruit set for 25 flowers tagged daily in one group planted with pollen source and second group planted without pollen source. Date Adjacent to Pollen Source 0-250 feet from Pollen Source July 16 12 17 80 # 18 80 # I9 92 12 20 8# 2# '21 80 # 22 8O # 23 60 t, 2# 100 # 25 80 0 26 76 0 27 -- # 28 -- # 29 52 O 30 76 0 31 - 72 0 August 1 6O 0 128 DISCUSSION Singh (1950) and other researchers have shown conclusively that honey bee gatherers visit flowers within a rather limited area of any one of a sequence of trips to the same field. Yield data indicated that there was no consistent movement of pollen much beyond 25 feet. Polli- nation beyond this distance likely took place with pollen from staminate flowers on the gynoecious plants although it might be theoretically possible for bees to retransfer pollen from stigma to stigma and thus extend the area of limited pollination. Beyond 25 feet from the pollen source, there was a reduction in the dollar value, the number of fruit per plant, and the number of seeds per fruit. No definite correlation could be made with fruit shape and distance from pollinator. In light of these results a new cultural method for planting gynoecious F1 hybrid cucumbers is suggested. That is the use of wide rows, possibly #0 feet wide, of a gynoecious variety, or an ethrel treated monoecious variety, alternating with narrow rows of a monoecious pollinator variety. Under these conditions, it is expected that Adequate pollination could be provided and the grower could obtain a more satis- factory assessment of the coordination of staminate and pistillate flowering as well as the overall performance of his gynoecious variety. VI DELAYED POLLINATION LITERATURE Cucumber growers and beekeepers frequently ask when to introduce bees into a commercial field. They wonder if they should move bees in- to a field prior to flowering, at the first sign of flowering, or after a delay of several days. Several lines of evidence have been mentioned which suggest that a delay in introduction of bees into cucumber fields would produce higher yields on a single harvest basis than in fields where bees are introduced at the first bloom. By way of summary they include the following: 1. Older, more vigorous plants produce fruit with a greater length/width ratio. 2. Older plants tend to produce larger flowers at the time of anthesis. Larger ovaries contain more ovules, and thus more potential seeds. The additional increase in number of seeds implies longer, more desirable fruit. 3. Commercial predominantly female varieties produce staminate flowes first, then mainly pistillate flowers. An appropriate delay in pollination would allow time for the development of more pistillate flowers and a more uniform fruit set. McGregor and Todd (1952) delayed pollination until the crown flowers of cantaloup had largely disappeared. They observed that ”In the post-crown bee treatments no melons were set until bees were 129 130 introduced, when melons set rapidly and at harvest time the production was not significantly different from that in the other bee cages.” How- ever, they observed a significant increase in the number of culls pro- duced by delayed pollination. Because of these observations, studies were carried out in 1967 and 1968 dealing with delayed pollination and its effect on yield and fruit shape. MATERIALS AND METHODS Preliminary investigations - August 1967 Three plots containing a total of 153 plants of variety Spartan Dawn with 5, IO and 20 percent pollinator (Spartan 27), were caged when the first flowers appeared. Bees were introduced seven days later. Ad- jacent plots without cages but constantly exposed to bees served as con- trols. Each plot was harvested twice and the weight and the number of fruit per plant recorded. Flower removal of different varieties Eight different varieties were tested for this response to delayed pollination by removing flowers at anthesis for the first four days of flowering. The varieties were planted in rows which were divided into three sections. Plants in one section had all flowers removed for the first four days of flowering (July 11-15), a total of 156 flowers. No flowers were removed from the other two sections. Each section was har- vested twice and the resulting number of fruit produced per plant tested by a paired t test. l3l Delayed pollination cage plots -l968 During July and August of 1968 an extensive study of the effect of delayed pollination was carried out using plots covered by plastic screen cages 8' x 12' x 6'. Bees were not allowed to enter the cages. Each plot contained approximately 75 plants growing in a 5” x 15“ spac- ing. Plants were counted and yields determined on a plant basis. The variety Piccadilly was used for this study. A decision on the number of days to delay pollination was based on growth of the vines and obser- vations of the most appropriate staminate:pistillate flower ratio. Five different types of manipulations were made with three plots involved in each one. They were: 1. Caged Pollination Control - Bees were caged with plants throughout flowering until harvest. 2. Open Pollination Control - No cage was placed over the plants and they were exposed to the field p0pu1ation of honey bees throughout the pollination period. 3. Delayed Pollination I - Bees were excluded by cages for 11 days, then introduced until harvest. #. Delayed Pollination II - Bees were excluded for 11 days, then the cages were open to field bees. 5. Delayed Pollination Ill - Bees were excluded for 11 days then placed within the cages for 6 days, and then excluded until harvest. The cage plots were harvested only once in order to duplicate a destructive harvest. Then the cucumbers were counted, graded according tc> their shape, curvature and constriction, and a dollar value determined according to Appendix 1. One way analysis of variance was used to test 132 for differences. RESULTS Preliminary Investigation - Summer 1967 Table 50 shows a significant increase in the number of fruit per plant in the delayed group, but no increase in the average weight of fruit produced per plant. The group of plants pollinated throughout the flowering period produced 2.18 fruits per plant, while the group delayed one week produced 2.85 fruit per plant. Weight of fruits per plant were 0.136 and 0.137 fruits per plant respectively. Flower removal of different varieties Results of this test as shown in Table 51 indicate that delayed pollination increased yield for the first harvest but not for the second. A significant increase in the yield of second over first harvest was found in the control group, while no significant yield differences were found in the delayed pollination group. In total, the delayed groups produced a higher yield than the control group. Delayed pollination in caged plots - 1968 Fruit per plant: Delaying pollination by excluding bees with cages resulted in increased yield (Table 52). Control I, where bees were caged throughout flowering, produced an average of 0.39 fruits per plant in a single harvest, while control 11, where plants were uncaged and exposed to field population bees, produced 0.75 fruits per plant. 133 The comparable delayed pollination groups produced l.5#, 1.21 and 1.#8 fruits for the first harvest. The highest yield occurred in the group delayed 11 days and then exposed to bees until harvest, although yield was not significantly different to that of the group exposed to bees for only 6 days. TABLE 50 Delayed Pollination - preliminary investigation August 1967 m Control Group: Bees present throughout blooming period Percentage Pollinator # Plants Weight/Plant Fruit per Plant Five 113 0.132 2.28 Ten 129 0.107 1.8# Twenty 105 0.169 2.#2 Total 3#7 Average 0.136 2.18 Delayed Group: Bees introduced seven days after first bloom Five #3 0.151 2.91 Ten 60 0.126 2.29 Twenty 50 0.135 3-34 Total 153 Average 0.137 2.85 Differences between the two groups in weight per plant not significant at 0.05 probability level. Differences between the two groups in fruit per plant significant- ly different at the 0.05 probability level. 13# TABLE 51 Effect of delayed pollination on yield of different varieties of cucumbers - July 1968 Number of fruit per plant. First Harvest Second Harvest Total Yield Variety Controla’c Delayeda’ Controlb’c Delayedb’d Controle Delayede Spartan 27 0.00 0.50 1.70 2.57 1.70 3.07 Spartan Dawn 0.83 2.#1 2.#1 2.83 3.2# 5.2# Piccadilly 1.03 3.28 2.70 2.6# 3.73 5.92 (Med. I seed) Piccadilly 0.72 2.00 1.92 2.53 2.6# #.53 (Med 11 seed) Spartan Progress 1.16 2.09 #.03 #.36 5.19 6.#5 Gynoecious 35 0.96 1.11 2.3# #.82 3.30 5.93 Piccadilly w/ pollinator 1.#5 1.50 2.87 1.75 #.32 3.25 SMR 58 0.96 1.61 1.88 1.92 2.8# 3.53 Average "umber 0.88 1.80 2.#5 2.91 3.33 #.71 of fruit/plant Results of paired t tests: aFirst harvest: control vs. delay, t=-3.53, significant at 0.05 blevel. Second harvest: control vs. delay, t=-l.29, no significance. cControl: lst harvest vs. 2nd harvest, t=-7.95 significant at 0.01 level. Delayed pollination: lst harvest vs. 2nd harvest, t=-2.22, no significance. eTotal control vs. total delay, t=-3.33, significant at 0.02 level. 135 TABLE 52 Effect on cucumber yield of delaying pollination by excluding bees from the flowers. Number of cucumbers per plant. Replicate Replicate . Replicate Type of Exposure 1 II ||| Average . a 1 Control - Bees in cage 0.38 0.37 0.#3 0.39 11 Control - Open field bees 0.59 1.06 0.61 0.75a Ill Delayed 11 days, b then caged with bees 1.28 1.75 l.6# 1.5# 1V Delayed 11 days, b then exposed in field 1.21 1.15 1.30 1.22 V Delayed 11 days, Bees b 6 days, then without 1.28 1.62 1.56 1.#8 bees until harvest F of I to V: (#,10) = 18.00; Significant at 0.01 probability level F of I to 11: (l,#) = -3.65; No significance F of III to V: (2,6) = 2.86; No significance. Dollar value per plot: As shown in Table 53, a delay of 11 days in the pollination of variety Piccadilly resulted in a highly signifi- cant increase in the estimated dollar value per acre. The values of the control groups (pollinated throughout flowering) were $26.70 and 78.30 per acre for the first harvest. The three groups delayed for 11 days produced yields of $230 to 285 per acre, with comparable harvests. These values are based upon 50,000 plants per acre. No significant variation in yield was found in the three different methods of delaying pollination. This indicates that in favorable weather sufficient bee pollination of cucumbers occurred in as short a period as 136 six days. There was no significant difference between the two control groups. TABLE 53 Effect on cucumber yield of delayed pollination by excluding bees from the flowersa. Dollar value per acre - estimate based upon 50,000 plants per acre. Replicate Replicate Replicate Average Dollar Treatment 1 II III Value/Acre 1 Control - Bees in cage $ 30 $ 35 $ 15 ’ $ 26 II Control ' Open field 55 '25 55 78 bees III Delayed 11 days, then caged with bees 250 315 290 285 IV Delayed 11 days, then exposed in field 200 250 250 230 V Delayed 11 days, bees 6 days, then no bees 235 295 305 278 until harvest aSingle harvest F of I-V: (#,10) #3.21; significant at 0.01 probability level F of 1-11: (1,#) #.59; no significance F of III-V: (2,6) = 2.57; no significance. Grading results: (Table 5#) Shape - More medium and long cucumbers occurred in the delayed pollination treatments. The number of nubs or short cucumbers was constant in both groups. The major differences were reflected in the added numbers of well-shaped fruit. Constriction - 137 There was a tendency for a smaller proportion of fruit to be constricted in the delayed pollination plots, although the actual number of severely constricted fruit remained constant in both groups. Curvature - A small- er proportion of curved fruit occurred in the delayed pollination group, although the actual numbers increased. TABLE 5# Effect of delayed pollination on cucumber shape. Number of fruit per shape per 1,000 plants. Shape Short Normal Long Control I 153 508 92 Control 11 85 213 102 Delay 1 L 119 1182 252 Delay 11 109 9#2 l7l Delay III 122 1155 123 Analysis of variance Row F not significant, Column F significant at 0.01 level. Constriction None Mild Severe Control I 653 137 68 Control 11 30# 61 35 Delay 1 12#l 253 59 Delay 11 1018 125 79 Delay Ill 1279 1#8 59 Analysis of variance: Row F not significant, Column F significant at 0.001 level. Curvature None Mild Severe Control I 212 316 230 Control 11 175 125 100 Delay 1 ' 930 #73 150 Delay 11 853 339 69 Delay III 1077 8#2 81 .Analysis of variance: Row F not significant, Column F significant at 0.025 level. —’_, CONCLUSIONS 1. Pickling cucumbers grown in Michigan rely entirely upon honey bees for pollination. With an adequate bee population density, uniform development of adequate numbers of staminate and pistillate flowers, and satisfactory weather conditions, bee pollination for present yields could be completed in a short time, probably less than a week. 2. Both pistillate and staminate flowers produce nectar which attracts bees. Flowers bloom one day only. Hand pollination of second day flowers was successful on rare occasions only. 3. The ovary of the pistillate flower varied in both size and number of ovules. A direct correlation was observed between the length of the ovary at anthesis and the number of ovules it contained. #. Cucumber vines grow over a prolonged period of time and frequent- ly produce lateral vines near the crown (root end). It was observed that pistillate flowers produced at early nodes on young plants were shorter and contained fewer ovules, which influenced the shape of the fruit. 5. Seeds were produced in all three carpels of cucumbers even when one lobe of the stigma was removed after pollination, indicating that pollen tubes from any of the three stigmatic lobes may enter any carpel. The removal of one lobe did increase the mortality of pollinated flowers. 6. Pickling cucumbers of several varieties sometimes produced fruit parthenocarpically' (without pOIIination). This phenomenon was environ- lnentally induced and observed most frequently in the fall and spring. 138 139 Parthenocarpic fruits were obtained most profusely on older plants lacking normally-pollinated fruit. The shapes ranged from crooked, hollow and severely tapered to those indistinguishable from the fully-pollinated fruit. 7. There was some correlation between number of seeds and the commercially important problem of pickling cucumber shape, although perf- ectly-shaped fruits were found with one to 520 seeds. Genetic make-up, nutrition and various stress factors such as drought, heat, cold, etc. affect cucumber shape. More detailed studies will be needed in order to fully understand the relative importance of these factors and their interrelationship with pollination. 8. Flowers closest to the crown were most likely to develop and be well-shaped. 9. As the number of fruits present on the vine increased, the probability of additional fruit deveIOping and being well-shaped decreased. 10. In a comparison of levels of fertility of the soil, low levels caused plants to flower more slowly, develop fewer flowers (both staminate and pistillate), set a higher percentage of pistillate flowers, and have more short fruit than plants grown in high fertility plots. 11. Honey bees began flights at the same general time cucumber flowers reached anthesis. Temperature was found to be the most import- ant factor in determining these events. During 1967, the average time of the first bee visits to cucumber flowers was 8:30 a.m. EDT, excluding -those nights when the overnight temperature did not fall below 70°F. “The average temperature at that time was 62°F. Bee flights rarely 1#0 became abundant until the temperature reached 70°F. 12. Honey bee visits to cucumber flowers varied greatly in length. 0n the first visit, bees spent 36 to 39 seconds. This dropped to 8 to 11 seconds for the #th through 12 h ViSits. Early morning visits were short, with the visit from 8 to 10 a.m. lasting 9 - 11 seconds, 10 to 11 lasting 30 seconds, and 11 a.m. to 6 p.m. lasting 38 to #5 seconds. Variations were observed, ranging from 22 to 50 seconds for the first visit on different days. Although these variations were due largely to weather factors no detailed analysis of these factors was attempted. 13. Honey bees were most active in the field from 10 a.m. to 3 p.m. for the 1967 season, with peak flights near noon. Very warm mornings produced peak activity between 9 and 10 a.m., while some cold days delayed the peak until 1 to 2 p.m., with no flights until 11 a.m. Three colonies pen acre produced 9.02 visits per flower per hour in a 10 hour period as the seasonal average for 1967. I#. Comparison of different numbers of colonies per acre indicat- ed that the number of bee visits per flower per hour increased with additional colonies of bees per acre, within limits of numbers tested. 15. Single visits of honey bees to pistillate cucumber flowers produced well-shaped cucumbers. The number of seeds increased with additional visits, implying that the fruit was not fully pollinated with just one visit. Data indicated that at least 10 bee visits were needed to insure pollination under a variety of variable conditions. 16. The longer the total honey bee visit time to pistillate flowers, the greater the number of seeds which developed in the fruit. 17. Under field conditions flowers exposed to honey bee pollination the entire day set the most fruit. Shorter exposure periods reduced the 1#1 level of fruit set. Comparing effectiveness of different one hour periods of exposure to bees throughout the day, the highest level of fruit set occurred during any one hour period from 10:30 a.m. to 2:30 p.m. There was good evidence that under normal Michigan conditions, stigmas remained receptive to pollen tube growth throughout the entire first day of bloom. This provided a longer period of effective pollin- ation than may be characteristic of hotter climates 18. Based upon the actual staminate:pistillate flower ratio prod- uced by different seed blends, a S:P ratio of 1:2 proved the best of those tested. 19. The percentage pollinator in plots influenced fruit pro- duction and fruit shape indirectly because some blends produce staminate and pistillate flowers at different stages of the growing period. The S:P ratio of 1:2 produced more short, non-curved, non-constricted fruit than other blends due to fruit setting on the first few nodes of the plant, where such shapes most frequently occur. 20. The staminate:pistillate flower ratio was influenced by light. Depending upon the time of the year, greenhouse plants (variety Piccadilly) varied in S:P ratio from 1.2:1 to 5.9:1. 21. Yield of a homozygous gynoecious line decreased 35% twentye five feet from a pollen source, although yields indicated some pollen movement took place up to 125 feet. 22. Delaying the introduction of bees and thus preventing fruit formation on plants for 5 to 11 days resulted in more fruit, with larger length/width ratios, and containing more seeds per fruit. It also allowed plants time to complete an early cycle of staminate flower pro- duction prior to the appearance of higher proportions of pistillate flowers. 1#2 23. Rotating colonies from one field to another during the flowering period did not appear to be advantageous. 2#. Several variables are involved in estimating the number of bees needed for optimum cucumber pollination. Plant populations for mechanical harvesting range from less than 20,000 to over 200,000 plants per acre. Varieties differ greatly in the number and time pattern of development of staminate and pistillate flowers. Environmental factors cause shifts in the number and ratio of S:P flowers. A colony of bees can vary in population from a handful to 60-65 thousand workers. Colonies can vary in the proportion of hive:fie1d bees at any one time. The ratio of bees working as foragers in early July could change drastically by August or September. The egg-laying pattern of the queen normally changes during this period. Furthermore, if the percentage of foraging bees could be established, it would still be difficult to determine the percentage that were working cucumber and those working other crops. Recognizing the variables and using information gleaned from this research, we now attempt to assess factors concerned with the beezflower:plant population dynamics of a cucumber field and to trans- late this into a practical recommendation on bee use. We start with a hypothetical situation of 50,000 cucumber plants per acre containing one new staminate and one pistillate flower each day. Assuming that bees make no major discriminations between flower types and that each pistillate flower must receive ten bee visits to be fully pollinated, 20 visits per plant per day would be needed. Assuming a minimum of #0,000 bees per colony with 20,000 field bees: if competing plants in the area attracted half the field force 1#3 on any one day, there would be approximately 10,000 bees pollinating 100,000 cucumber flowers. If each flower received 10 visits there would be a total of one million visits. If each bee visited 100 flowers each day (divided between several trips), one colony would provide 10,000 x 100 or one million flower visits. On this basis and ignoring obvious variables one colony of at least #0,000 bees should provide an adequate pollination force for one acre of 50,000 cucumber plants. We have no data on the interaction of plant density in this eco- logical situation. We might assume that the number of colonies should be increased proportionate to the number of plants per acre. This might imply the need for 1.5 colonies for 75,000 plants per acre and 2 colonies per 100,000 plants per acre. It is possible that the need for colonies would be less than proportionate with further increase in plant popula- tions per acre. LITERATURE CITED LITERATURE CITED Adlerz, W. C. 1966. Honey bee visit numbers and watermelon pollination. J. Econ. Entomol. 59: 28-30. Alex, A. H. 1957. Honey bees aid pollination of cucumbers and cantal- oupes. Progr. Rep. Texas Agric. Exp. Sta. No. 1936. Atsmon, D., Galun, E., and Jakob, K. B. 1965. Relative time of anthesis in pistillate and staminate cucumber flowers. Ann. Bot. 29: 277-283. Barnes, W. C. l9#7. Cucumber breeding methods. Proc. Amer. Soc. Hort. Sci. #9: 227-30. Berkel, N. van, and Vriend, S. 1957. Zaadkoppen bij komkommers. (Seed-heads in cucumbers) Jversl. Proefst. Groenten Fruitt. Naaldwijk 1965: 117-120. Abstract seen in Apic. Abst. 1963: 9#8. Bushnell, J. W. 1920. The fertility and fruiting habit of the cucumber Proc. Amer. Soc. Hort. Sci. 17: #7-51. Chakravarty, H. L. 1958. Morphology of the staminate flowers of the cucurbitaceae with special reference to the evolution of the stamen. Lloydia 21: #9-87. Chaudhury, B., and Phatak, S. C. 1959. Sex expression as affected by Gibberellin. Indian J. Hort. 16(#): 233-235. Cummings, M. B. l90#. A study of reciprocal crosses. Maine Agr. Exp. Sta. Bull. 10#. Cunningham, C. R. 1939. Fruit setting of watermelons. Proc. Amer. SOC. Hort. SCio 37: 81]-“. Currence, T. M. 1932. Nodal sequence of flower type in the cucumber. Proc. Amer. Soc. Hort. Sci. 29: #77-9. Edmund, J. B. 1930. Seasonal variation in sex expression of certain cucumber varieties. Proc. Amer. Soc. Hort. Sci. 27: 329-332. Foster, R. E., and Levin, M. D. 1967. F] hybrid muskmelon, 11. Bee activity in seed fields. J. Ariz. Sci. #: 222-5. 1## 1#5 Hayase, H. 1955. Effect of temperature before and after anthesis on pollen viability in cucumber and egg plant. J. Hort. Assoc. Japan. 2#: l##-8. Seen in Hort. Abst. 26:172#. . 1960. Physiological studies on the pollen of the genus Cucurbita with special reference to the fertilizing power of pollen and anther dehiscense. Hokkaido Natl. Agric. Exp. Sta. Rept. 5#: 7# pp. Haynes, H. K., and Jones, D. F., 1916. First generation crosses in cucumbers. Conn. Agric. EXpt. Sta. Rpt. 1916: 319-22. Hayward, H. E. 1938. Structure of Economic Plants. MacMillan Co. Chapter IV. Heimlich, H. F. 1927. The development and anatomy of the staminate flow- er. Amer. Jour. Bot. l#: 227-37. Knysh, A. N. 1958. Pollination by bees of varieties of cucumbers. Sad i Ogorod (6): 13-16. (In Russian) Abstract seen in Apic. Abst. 1958: 372. Koot, Y. van. 1960. The effect of bees on the production of mis-shapen cucumbers. Meded. Du. Tuinb. 23(80): 735 - 7#9. (In Dutch; English summary). Abstract seen in Apic. Abst. 1962: 338. Jenkins, J. M. l9#2. Natural self-pollination in cucumbers. Proc. Amer. Soc. Hort. Sci. #0: #11-2. Judson, J. E. 1929a. The morphology and vascular anatomy of the pistillate flower of the cucumber. Amer. Jour. Bot. 16: 69-89 0 . 1929b. The floral development of the staminate flower of the cucumber. Papers Mich. Acad. Sci. 9: 163-8. . 1935. Floral development of male flowers of the honey rock muskmelon. Proc. West Va. Acad. Sci. 8: 93-8. . l9#9. The pistillate flower of the muskmelon. Proc. West. Va. Acad. Sci. 20: 79-8#. Levin, M. D., Kuehl, R. 0., and Carr, R. V. 1968. Comparison of three sampling methods for estimating honey bee visitation to flowers of cucumbers. J. Econ. Entomol. 61(6): 1#87-8. Markov, 1., and Romanchuk, I. 1959. Pollination of cucumbers by honeybees. (In Russian). Sel. Khoz. Sibiri (2): 53-#. Abstract seen in Apic. Abst. 1961: 237. Mann, L. K. l9#3. Fruit shape of watermelon are affected by place- ment of pollen on stigma. Bot. Gaz. 105: 257-62. 1#6 . 1953. Honey bee activity in relation to pollination and fruit set in the cantaloup, Cucumis melo L. Amer. J. Bot. #0: 5#5-53. Mann, L. K., and Robinson, J. 1950. Fertilization, seed development and fruit growth as related to fruit set in cantaloup (Cucumis melo L.). Amer. J. Bot. 37: 685-97. McCollum, J. P. 193#. Vegetative and reproductive responses associated with fruit development in the cucumber. Cornell Agr. Expt. Sta. Memoir. 163: 1-27. McGregor, S. E., and Todd, F. E. 1952. Cantaloup production with honey bees. Jour. Econ. Entomol. #5: #3-7. McGregor, S. E., Levin, M. D., and Foster, R. E. 1965. Honey bee visitors and fruit set of cantaloups. J. Econ. Entomol. 59: 968-70. McMurray, A. L., and Miller, C. H. 1968. Cucumber sex expression modified by 2-Chloroethanephosphonic Acid. Science 162: 1397-8. . 1968. Changing the sex of cucumbers by chemical means. PicklePak 28:3 only. Miller, C. H., and Ries, S. K. 1958. The effect of environment on fruit development of pickling cucumbers. Proc. Amer. Soc. Hort. Sci. 71: #75-9. Morrison, F. D., and Ries, S. K. 1967. Cultural requirements for once- over mechanical harvest of cucumbers for pickling. Proc. Amer. Munson, W. M. 1892. The secondary effects of pollination. Mich. Agric. College. Ph.D. Thesis. Nemirovich-Danchenko, E. N. l96#. Concerning the nectar yield and floral biology of cucumbers. lzv. tomsk. Otd. vses. bot. Obshch. 5: 127-32. Abstract seen in Apic. Abst. l96#: Nevkryta, A. N. 1953. Insect pollinating cucumber crops. (In Russian). Kiev: Izdatel'stvo Akademii nauk ukrainskoi SSR 92 pages. Seen in Apic. Abst. 1961:153. Peterson, C. E., and Weigle, J. L. 1958. A new method for producing hybrid cucumber seed. Mich. Agr. Exp. Sta. Quarterly Bull. ‘10: 960-5 0 Peterson, C. E., and Anhder, L. D. 1960. Induction of staminate flowers on gynoecious cucumbers with Gibberellin A3. Science 131: 1673-#. Peterson, C. E. 1960. A gynoecious line of cucumber. Mich. Agr. Exp. Sta. Quarterly Bull. #3: #0-2. 1#7 Peterson, C.-E., and DeZeeuw, D. J. 1963. The hybrid pickling cuc- umber, Spartan Dawn. Mich. Agr. Exp. Sta. Quarterly Bull. 1‘6: 267-73 0 Pike, L. M. 1967. The inheritance and related studies of the partheno- carpic character in cucumber Cucumis sativus L. Mich. State Univ. Ph.D. Thesis. (Unpublished). Pike, L. M. 1969. Inheritance of parthenocarpy in the cucumber, Cucumis sativus L. Euphytica (In press). Poole, C. F., and Porter, 0. R. 1933. Pollen germination and develop- ment in watermelon. Proc. Amer. Soc. Hort. Sci. 30: 526-30. Robinson, F. A. 1952. The use of honey bees in production of cucurbits in Flordia. Amer. Bee. Jour. 92: 326-8. Sakharov, M. K. 1958. Pollination by bees on vegetable seed plots. (In Russian). Sad i Ogorod 96: 21-3. Abstract seen in Apic. Abst. 1961: 1#7. Sanduleac, E. 1961. Data on the entomophilous pollination and the selection of cucurbitaceae. In Romanian. Lucr. sti. Stat. cent. Seri. Apic. 1: 129-32. Abstract seen in Apic. Abst. '96]: “3]. Schwarz, R. 1968. Field Manager, Heifetz Pickle Company, Eaton Rapids, Michigan. Personal communication. Seaton, H. L., Hutson, R., and Muncie, J. H. 1936. The production of cucumbers for pickling purposes. Special Bull. No. 273. Mich. Agric. Exp. Sta. Seaton, H. L. 1937. Relation of number of seeds to fruit set and shape in cucumbers. Proc. Amer. Soc. Hort. Sci. 35: 65#-8. Seaton, H. L., and Kremer, J. C. 1938. The influence of climatological factors on anthers and anther dehiscense in the cultivated cucurbits. A preliminary report. Proc. Amer. Soc. Hort. Sci. 36: 627-31. Shahin, Abd El-hay Abbas. 1967. The influence of pollinators, especially honey bees on fruit and yield of some cucurbits. cairo University, Ph.D. Thesis. (Unpublished) Shemetkov, M. F. 1957. The use of bees for pollinating cucumbers in hot-houses and forcing beds. (In Russian). Byull. nauch.- tekh. Inf. Inst. Pchelovodstvo (2): 21-#. Abstract seen in Apic. Abst. I959: 389. . 1960. Pollinating activity of bees in greenhouses. (In Russian). Pchelovodstvo 37(1): 28-31. Abstract seen in Apic. Abst. 1963: #37. 1#8 Shifriss, O. 1961. Sex control in cucumbers. J. of Heredity. 52: 5-12. Shuel, R.wy. 1967. The influence of external factors on nectar pro- duction. Amer. Bee. J. 107: 5#-56. Sims, W. L., and Gledhill, B. L. 1969. Etherel effects on sex expression and growth development on pickling cucumbers. Calif. Agric. 23(2): #-6. Singh, Sardar 1950. Behavior studies of honey bees in gathering nectar and pollen. Cornell U. Agr. Expt. Sta. Memoir: 288. Stout, B. A., LeLong, M. M., Pettingil, D. H., and Ries, S. K. l96#. Once-over mechanical harvester for pickling cucumbers. Mich. Agr. Exp. Sta. Quarterly Bull. #6: #20-30. Takashima, S., and Hatta, S. 1955. Effects of phytohormones and parthenocarpy in cucurbits. J. Hort. Assoc. Japan. 2#: 59 - 61. Taylor, E. A. 1955. Cantaloup production increased with honey bees. J. Econ. Entomol. #8: 327 only. Tjezdjens, V. A. 1926. Some observations on the response of greenhouse cucumbers Cucumis sativus L. to certain environmental conditions. Proc. Amer. Soc. Hort. Sci. 23: 183-9. 1928 a. The relation of environment to shape of fruit in Cucumis sativus L. and its bearing on the genetic potentialities of the plants. J. Agric. Res. 36: 795-809. 1928 b. Sex ratios in cucumber flowers as affected by different conditions of soil and light. J. Agric. Res. 36: 721-#6. Tillman, 0. I. 1906. The embryo sac and embryo of Cucumis sativus L. Ohio Naturalist. VI: #23-30. Tuljyenkova, F. F. 1955. The effect of pollination by bees on a yield of cucumbers grown under cover in the far north. Sad i Ogorod. #: 2# - 5. Vasil, I..K., and Johri. M. M. 196#. The style, stigma, and pollen tube. 1. Morphological and anatomical features of the stigma and style for 6 families. Phytomorphology. 1#: 352-69. Verdieva, M. G., and Ismailova, M. K. 1960. The influence of bee pollination on the increase of the crop from feed squash. (In Russian). Pchelovodstvo 37: #O-#1. Abstract seen in Apci. Abst. 1963: 951. Warren, L. O. 1961. Pollinating cucumbers with honeybees. Arkans. Fm. Res. 10: 7 only. Weigle, 1#9 J. L. 1956. Evaluation of FI hybrid pickling cucumbers. Ph.D. Thesis. Umichgan State Univ. (Unpublished). Whitaker, T. W., and Davis, D. N. 1962. Cucurbits. John Whiley and Sons. 250 pp. Williams, P. H., and Kauffeld, N. M. 1967. The role of honey bees Wong, C. on pollination of pickling cucumbers. PicklePak. 27: #-5. 1938. 'lnduced parthenocarpy of watermelon, cucumber and pepper by the use of growth promoting substances. Proc. Amer. Soc. Hort. Sci. 36: 632-6. Zobel. M. P., and Davis, G. N. l9#9. The effect of the number of fruit per plant on the yield and quality of cucumber seed. Proc. Amer. Soc. Hort. Sci. 52: 355-8. APPENDIX I Method of determining the dollar value for cucumbers based upon the diameter of the fruit. Harvested cucumbers were dropped through a grading board consisting of a series of holes of the specific diameters. ' GRADES INCHES IN DIAMETER DOLLAR VALUE #1 up to 1-1/16“ $6 per th. #2 1-1/16 to 1-1/2“ $3 per th. #3 1-1/2 to 2“ $2 per th. ## 2” to 2-l/#'I $1 per th. #5 2-1/# to 2-1/2” 50 cents #6 over 2-1/2” 25 cents RATINGS: 1 to 5 l is best, 5 is poorest. Developed by members of Pickling Cucumber Improvement Committee and approved as of 2/19/67. 150 11111111111111 143111111111 11111 1111 ES 131392