POLLINATION STUDIES ON THE cumvmen sweeem YACCENIUMCORYMBOSUM LINNAEUS Thesis for Hm Dogma of M. S. MICHIGAN STATE UNIVERSITY Joseph Edward Dorr 1965- WWWWWIllflflllliflfilllflflll 1mm 31293 00649 2197 Michlgan State Universny "JUN 1 7 1992 [$5 1‘ . _ #4 7 4 ‘éfif’ {3‘75 M? m @0044 ABSTRACT POLLINATION STUDIES ON THE CULTIVATED BLUEBERRY Vaccinium corymbosum Linnaeus by Joseph Edward Dorr During the spring and summer of 196% pollination studies on the cultivated blueberry, Vaccinium corymbosum Linnaeus, were conducted in southern Michigan. Objectives of the study were to check the effective- ness of honey bees, Apis mellifera Linnaeus as pollinators and to evaluate any advantages of cross-pollinating varieties. Observations of wild bees in regard to their abundance and varietal preference were also madeo At the research site in Van Buren County near Grand Junction, Michigan, experiments were conducted to test honey bee pollination effectiveness. Cages were set up over cultivated Jersey bushes into which a colony of honey bees was introduced. One cage without bees was used as a check. Experiments of cross-, self- and Open-pollination were also carried out at the research site. To test cross-pollination a bouquet of blueberry blossoms of the varieties Rubel, Coville or Bluecr0p was placed into the cages containing the Jersey plants and honey bees. The cages containing Jersey bushes and a colony of honey bees, but no bouquet, were used to check the effect of self-pollination. Comparisons to cross- and self-pollination were made on randomly selected Jersey bushes in the same test field. Joseph Edward Dorr Field observations were made on cultivated blueberries in open— pollination throughout Michigan's blueberry-growing area. In these, test counts were made on the number of bees observed working on a single bush per minute, with special reference being made to the blueberry variety. Samples of pollinators were collected and identified to give an indication of the most abundant species. At the conclusion of the study the results indicated that: l. Cross-pollination of varieties was not significantly superior to self-pollination when bouquets were used for the pollinator variety. Jersey variety definitely needs some means to insure pollination and honey bees can be successfully utilized in this manner. The amount of pollination increases with higher concentrations of bees, producing a higher total yield and earlier ripening. The percent of large dark seeds in a berry is correlated with increased berry size. While bumble bees play an important role in blueberry pollination, their value must be weighed against the value of present cultural practices. There is a definite degree of varietal attractiveness to bees, which needs further study to determine possible effects in large block plantings. POLLINATION STUDIES ON THE CULTIVATED BLUEBERRY Vaccinium corymbosum Linnaeus by Joseph Edward Dorr A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Entomology 1965 ACKNOWLEDGEMENTS The author is grateful to Dr. Gordon E. Guyer, Chairman of the Entomology Department, for providing financial assistance for this study and for serving on the author's guidance committee. I wish to express my sincere appreciation to my major professor, Dr. E. C. Martin, under whose constant advice, encouragement and suggestions this investigation was undertaken. I am greatly indebted to Dr. Roger A.Hoopingarner and Mr. C. E. Lewis for reading the manuscript and making valuable suggestions, and to Messrs. Robert B. Carlson and Robert Husband for their technical assistance. Special thanks are extended to John W. Nelson, Research Director of the Michigan Blueberry Growers' Association for his interest, assistance and many helpful suggestions. Thanks are also due to the Michigan Blueberry Growers‘ Association for providing added financial aid and to its many members whose farms were used in the investigation. Finally, my personal thanks are due to my wife, Helen, for typing this thesis and for her valuable suggestions and continued encouragement throughout this undertaking. ii TABLE OF CONTENTS INTRODUCTION . . . . . . . . Background History . . . . . Distribution and Economic Importance . . . . . Nature of the Problem . . REVIEW OF LITERATURE . . . . . Pollination Studies . . . . Pollinating Agents . . . . . Correlation Between Berry Size MATERIALS AND METHODS . . . . . Controlled Pollination Study Field Pollination Study . . mm 0 O O O O 0 O O O O O O Controlled Pollination Study . Field Pollination Study . . DISCUSSION . . . . . . . . . . SUMMARY AND CONCLUSION . . . . LITERATURE CITED . . . . . . . APMDH I O 0 O O O O O O o 0 iii Seed Number . CDCD-F’U) 1h 16 17 17 22 21+ 21+ 39 1+6 52 5h 57 LIST OF PLATES Plate Page 1. Map of Michigan showing blueberry areas . . . . . . . . . . 7 Boundary of high cultivation Line dividing highbush from lowbush Research area 2. Blueberry plantation showing cages used in field pollination tests 0 O O 0 O O O O O O O 0 O O 0 O O O O O 0 0 0 0 O O O 18 3. Close-up of cage used in cross-pollination tests showing hive of honey bees, bouquet and pan of water for bees . . . 20 A. View through the cage screen showing a honey bee working the caged Jersey buSh O O O O 0 O O O O I O O O O O O O O 0 36 5. Bumble bee working a Jersey bush in open-pollination . . . 37 iv LIST OF TABLES Table Page I. Table showing individual plant yield for each picking date of cross-, self- and Open-pollination With and Without bees 0 O 0 0 0 O O O 0 O O O O O 0 O 0 0 0 2S 2. Yields of self- and cross-pollinated plants . . . . . . . . 27 3. Analysis of variance on yields of self- and CI‘OSS-pOllinated plants 0 o e e e o o o o o o o o o o o o o 27 h. A comparison showing the effects of bees on the time of ripening for different pollination treatments . . . 28 5. Relative berry size as indicated by mean cup count in cross-, self- and open-pollination and with no bees . . 31 6. Relationship of the percent of large dark (L.D.) and small light (S.L.) seeds to the mean seed count for 10 berries of cross-, self— and open-pollination of bees and no bees . . . . . . . . . . . . . . . . . . . . . 32 7. Relationship of percent of large dark seeds, mean cup count, and the time of ripening . . . . . . . . . . . . 33 8. Observed bee activity in the Open-pollination test on Jersey 0 O O O I O O O 0 O C C O O O O O O O O O O O O O 38 9. List of bees collected in Jersey test field . . . . . . . . 38 IO. Pbllinating insects observed on different varieties of cultivated blueberries . . . . . . . . . . . . . . . . . 39 ll. List of pollinating insects collected in the Field Pollination Study . . . . . . . . . . . . . . . . . . L3 12. Relationship of the percent of large dark seeds to the mean cup count in the Field Pollination Study . . . AS LIST OF FIGURES Figure l. 2. Page Relationship between the percent of total yield harvested from each treatment for each picking date 0 O O O 0 o o o o o e o o o o o o o o o o o o o o o o 30 Relationship between the percent of large dark seeds and the mean cup count . . . . . . . . . . . . . . . 3% vi INTRODUCTION Vaccinium corymbosum Linnaeus, the cultivated highbush blueberry, has been an economic asset to Michigan Agriculture. Due to climatic and soil requirements that are peculiar to highbush blueberries, their production has been limited to southwestern Michigan. While tremendous advances have been made in agriculture pertaining to blueberry production, certain developments have had some adverse effects. The practice of clean cultivation, while having an initial additive effect, has subsequently been thought of as a cause for the generally mediocre crop taken from the number one variety in Michigan, Jersey. It has been reasoned that while clean cultivation does relieve the competition between weeds and the blueberry plants, it also has destroyed the natural nesting sites of many wild bees. Growers state that wild bees were more numerous in most blueberry plantations before the general use of herbicides as a tool of clean cultivation. Growers have also noted a decline in wild bee p0pulations since the wide use of insecticides in the blueberry cultural program. While insecticides are not applied during the actual time pollination is in'progress, an extensive program is employed throughout the rest of the growing season, having a possible deleterious effect on various stages in the life cycle of wild bees of the area. This drop in yield of Jersey variety has apparently coincided with the decline of wild t... in the blueberry fields. In past years growers.felt that wild bees, both solitary and l social, were mainly responsible for blueberry pollination. Since experiments by Merrill (1936) and Merrill and Johnston (1939) indicated that cross-pollination was not necessary under Michigan A conditions, large blocks of single varieties have been planted. This type of planting is quite different from what is found in many orchard fruits where yields from self-pollination are commercially negligible. Many growers and their representatives in the Michigan Blueberry Growers' Association felt more information was urgently needed regarding the pollination needs of Michigan blueberries. This information was needed to establish which species of bees were most effective, the value of cross-variety pollination and the extent of economic loss due to insufficient pollinators. During the 196A growing season, tests were run using honey bees as pollinators on blueberries. The objectives were to check the effective- ness and potentialities of honey bees, Apis mellifera Linnaeus, as pollinators of highbush blueberries, and to evaluate any advantages of cross-pollinating blueberry varieties in Michigan. The relative value of wild bees for blueberry pollination was also studied. Background History While the cultivated blueberry industry is relatively new to the United States, the blueberry itself is not. Many wild species of Vaccinium were indigenous to the United States, where the wild fruit was cherished and picked extensively by both the Indians and early settlers. While at present there is an industry in some states based on the wild blueberry, it is not as prominent as the industry based on the highbush cultivated blueberry, Vaccinium corymbosum L. As long as wild blueberries were available in adequate quantities, no attempts were made to cultivate them. Johnston (1959) stated that in 1906 Dr. Frederick V. Coville of the United States Department of Agriculture initiated fundamental research on life history and selection which he thought might lead to the development of a cultivated blueberry industry. Mrs. Elizabeth White, also credited with improving the high- bush blueberry, collaborated with Dr. Coville in selecting the first commercial planting of highbush blueberries on her farm in Whitesbog, New Jersey. Distribution and Economic Importance The highbush blueberry is native from northern Florida to southern Maine and westward to southern Michigan. jMichigan and New Jersey lead in the production of cultivated blueberries, with substantial plantings also found in Indiana, North Carolina and Washington. Smaller areas of blueberry cultivation are found in New York, Pennsylvania, Ohio, several New’Ehgland states and Oregon. In Michigan the line which separates the lowbush from the highbush blueberry is fairly distinct. The dividing line runs from Bay City, southwest across the state to a point Just north of Grand Rapids, then northwest along the isothermal lines to the shore of Lake Michigan. The reason for this slight change in direction stems from the moderating influence of the lake (Plate 1). More specifically in Michigan the area where cultivated highbush blueberries are grown extensively is confined to five counties: Berrien, Van Buren, Allegan, Ottawa, and Muskegon (Plate 1). According to the Michigan Blueberry Growers' Association, the total land devoted to blueberries in these five counties is over 9,000 acres. Conditions for growing highbush blueberries are very suitable in this area. Johnston (1959) states these generally include a long grow- ing season, a moderate winter temperature, a soil which is sandy and contains a high degree of organic matter, a pH between h.0 and 5.1, and a water table that can be maintained between lh-22 inches below the surface. Total acreage of blueberries is increasing annually in Michigan. In 196A the cultivated blueberry industry was valued at nearly five million dollars (Holbein, 196A, personal interview). Nature of the Problem Growers feel that Jersey, the predominant commercial variety, has been decreasing in expected yields over the past few years. In examining the possible causes for this decline it appears likely that the number of native pollinators has been decreased by clean cultivation and the increased use of insecticides. This decline of native pollinators is not necessarily limited to the blueberry fields alone. Waste areas surrounding the blueberry fields, which make natural nesting sites for many native bee pollinators, are also affected. Growers, by extending their insect and weed control programs into these areas, have apparently reduced wild bee populations. With bumble bees being the only exception, all are solitary bees building their separate nests close together, often producing large populations in some areas. Dr. E. C. Martin (1965, personal communication) relates that the prevalent opinion among growers in 1951 was that wild bees were predom- inantly responsible for blueberry pollination and that coverage was adequate fer maximum yield. However, it appears that in the last five to ten years the deleterious effect that insecticides and herbicides have had on wild bee populations has most seriously affected the yield from the Jersey variety. Another factor that might conceivably influence the yield of the Jersey variety is the practice of large block planting. The investi- gations of Merrill and Johnston (1939) showed that cross-variety pollination was not necessary for a normal set under Michigan conditions. These investigations led to the planting of large blocks of single varieties on Michigan plantations. If cross-variety pollination does increase yield, then the trend toward larger and larger blocks of just a few varieties may be reducing yields. Therefore, instead of planting large blocks of one variety as is currently practiced, several varieties should possibly be interplanted. Shoemaker (1955) points out, however, that if several varieties are planted together to obtain the value of cross-pollination, too many may complicate the cultural and harvesting operations. This could cause a serious problem in the blueberry industry. Merrill (1936), however, noted the greater importance of insuring some means of pollination over the importance of self- or cross-pollination. Merrill also reported that honey bees were incapable of effective pollination in some varieties and that this would leave the task of pollination to bumble bees. He further stated that because bumble bees cannot be handled as honey bees are, natural conditions must be main- tained to conserve the bumble bees. Although Jersey, the number one variety in Michigan today, was not referred to in Merrill's original work, his results have been applied to this variety. Due to the inconclusive results of previous studies and the lack of any definite impact on production practices, the situation seemed to warrant a renewed and comprehensive study of the pollination problem in Michigan. Tests were set up to check the possibilities of: l) substituting honey bees for natural field bee pollinators, and 2) cross-pollinating varieties as a method of increasing blueberry yields. These tests were set up on the John W. Nelson blueberry plantation in Grand Junction, Van Buren County, Michigan (Plate 1). LAKE rfmfi K W8 3 MEI" ,. mix‘. / om. '\ I \ \ -' 'i ' mama: ! LEELANAU \\ -‘ --‘L.-.T.-. I Q 1 -. l . \0 mam 1:315:50 M| W \ _J . m -._. ..._ _r__ __|_.__. aszI-Twuo |WIW0RU OSCODA! ALCONA _....:.R5EL.._ i_ .._. _._l__._. ' inosco Musmi room ma 1 MAW wax J rmj—octw L —L . o - IT —[ fii I J—IARENAG ““ONI LAKE iosctow cum: Maj. i ' J! "r luv , —'-1. Mam m u ._, “W m%\ chgxw ' ’IMMNN»--4 GENES“ l. .11.i__j_ -1 Imam mum m1 «mo . ooj CALHOUN i JACKSON IWASHTLNij mm: i L. .i. _. .._. -—-r 3"!" -[ — ans isuomiamcn imusuu| LENAWEE gum ._.....L.._J_-._.L L__,._L lEflEIfl= _IIIIIII‘I II III. NIIIINIII [Ill/l llll IIIIIIII IIIIIISI "ll IIIIISI Q IISIIMI If" PLATE 1 -—Map of Michigan showing blueberry areas REVIEW OF LITERATURE Pollination Studies Darwin (1885), instrumental in the early investigations concerning plant fertilization, drew from his observations that cross-fertilization is generally beneficial and self-fertilization injurious. Darwin stated further that "there is weighty and abundant evidence that the flowers of most kinds of plants are constructed so as to be cross- fertilized." One of the first publications concerning blueberry pollination was by Coville (1910) where he stated after extensive observations that, "fruit was produced from flowers pollinated either with their own pollen or with pollen from another flower." This suggests that Coville felt either self- or cross-pollination could be employed effectively to set fruit. However, Coville (1921) stated after further investigations that, "when blueberry flowers are pollinated with pollen from their own bush the berries are fewer, smaller, and later in maturing than when the pollen comes from another bush." Coville (1921) continued that because of the pollination problem it is very important not to have a blueberry plantation made up "wholly from cuttings from one bush." Robbins (l92h) further states on the blueberry pollination problem that, "if pollen from the same plant is used in pollination, the fruit that is formed remains small and green and later drOps off. This fact serves to emphasize the need in the prepagation of blueberries by cuttings, and of making the plantation from cuttings of a number of 8 different bushes." Beckwith and Coville (1927) collaborated and stated emphatically that blueberries must be cross-pollinated for best results. To allow for better pollination by bumble bees and other insects Beckwith and Coville suggested rows of at least two different varieties alternating throughout the field instead of a solid block of one variety. Beckwith (1930) observed after further close study that self- pollinated blueberries under New Jersey conditions were smaller and matured much later, "even when the pollen was carried promptly from flower to flower by bumble bees." Beckwith fUrther noted that "berries of commercial size and quantity could not be produced in this way." Other researchers in other areas were also working on the pollination problem. Crowley (1933) in washington claimed that while fruit did set on "White" variety by selfing, larger and higher yields were obtained when cross-pollination was realized. Merrill (1933, 1936) under Michigan field conditions generally secured a higher percent fruit set with cross-pollination than with self-pollination, but lower than Open-pollination. Merrill (1933) however, while observing a planting of #8 Rubel bushes growing by- themselves at the South Haven, Michigan, Experiment Station, noted they bore an exceptionally heavy fruit crop for three years with no other bush nearer than two miles. Merrill stated after this observation that, "open-pollination, which must have been entirely self-pollination, can induce heavy yields", and also that, "selfed berries were as large in every case as the crossed berries and matured in their normal season." Merrill (1933, 1936) also noted that bumble bees and honey bees are 10 the principal pollinating insects in a blueberry plantation. The method used by Merrill (1936) was to remove all the blossoms on the clusters that were already open or too small and then to emasculate the rest for both the self- and cross-pollination tests. He used his finger to apply the pollen. Bailey (1938) in Massachusetts worked with several varieties including Michigan's number one and two varieties, Jersey and Rubel respectively. Bailey's experiments differ from Merrill's in that Bailey did not remove any of the blossoms on the cluster he used, and his pollinating was done with a camel's hair brush. Bailey's findings indicated that while open-pollination proved in most cases satisfactory, self-pollination did not set enough fruit for a commercial crop. White and Clark (1938) reported that in many cases tests on field grown plants showed, "the percent of total flowers self-pollinated which set fruit was not significantly different from the percent set with cross-pollination." They also stated, however, that the berries resulting from self-pollination were consistently smaller than those resulting from cross-pollination, and in some cases ripened later. White and Clark used the same techniques as those used by Merrill (1936) with the exception of using a small metal spatula for applying the“ - pollen. Marrill and Johnston (1939) conducted new experiments in Michigan to determine the degree of self-fertility of varieties not previously tested, including in this series of tests the variety Jersey. They concluded from these tests that the percent set was not significantly different between self- and open-pollination. They further reported that 11 the berries resulting from slef-pollination were of equal size when compared to open-pollinated berries. Again the hand pollination in this experiment was done either by a camel's hair brush or a finger. In North Carolina, Schaub and Bauer (l9h2) showed that in comparisons of cross- and self-pollination on varieties of Scammell, Weymouth and Dixi, cross-pollination gave earlier ripening by a week and increase in set up to twenty percent. Schaub and Bauer (l9h2) and Shoemaker (1955) also noted that in North Carolina, mixed planting would give a better set and also provide an earlier ripening and better grade berry, providing a greater commercial advantage. Mbrrow (l9h3) conducted experiments in North Carolina in a green- house using both emasculated and unemasculated flowers. His technique of hand-pollination consisted of using a blotter with no thinning of blossoms from the clusters. His results indicated that cross-pollinated berries were larger and ripened earlier than those developing from self-pollination. Meader and Darrow (19h?) in Maryland used the techniques and conditions initiated by Morrow (l9h3) and found that while self- pollination often produced as large a percent fruit set as cross- pollination, the berries produced were smaller and later in maturing. In an earlier test, Meader and Darrow (l9hh) employing the same technique on Vaccinium ashei acquired a wide variation in self-fertility. They reported that the set and size of berry was greater while days to maturity generally decreased with cross-pollination than with self- pollination. Cremmins (1952) in a greenhouse in Oregon conducted a pollination 12 study on hybrid blueberries using both emasculated and unemasculated blossoms, a glass rod as the pollination instrument and no thinning of blossoms. He obtained a smaller set and a smaller berry plus a later-maturing berry from self-pollination than from cross-pollination of the Dixi variety. Tests conducted on Jersey bushes in the field resulted in no significant difference in the percent set obtained between self- and cross-pollination, although self-pollination did result in a smaller size of berry and a delay in maturity. Thomas (1962) conducted a pollination study at Wooster, Ohio and found that cross- and open-pollination resulted in earlier ripening, greater fruit set and larger berries than self-pollination of the varieties Jersey and Berkeley. He also noted that while a commercial set could not be obtained on either Jersey or Berkeley by hand- pollination, it was obtained on the Jersey when it was self- or open- pollinated by bees. The latest recorded findings comparing the pollination requirements of highbush blueberries was done by Filmer and Marucci (1963) at the New Jersey Experiment Station. They performed not only self;, cross- and hand-pollination tests, but cage tests as well, where honey bees were used to cross-pollinate the blossoms. Filmer and Marucci noted that the conditions were so favorable in the cages that contained bouquets of unlike pollen that the benefits from cross-pollination could not be demonstrated. In hand-pollination tests, however, a vigorous response was obtained when unlike pollen was used. The hand- pollinating was done by use of a camel's hair brush. In conclusion they stated that the lack of cross-pollination does not prevent setting of fruit as was shown in their hand-pollination tests and in the cage tests where bouquets of unlike blossoms were not provided. However, "the great increase in size, rapidity of ripening and uniformity of size which are inherent in cross-pollination are all commercially desirable features." In summary, most of the experiments indicated that a higher percent fruit set could be obtained from cross-pollination than from self- pollination. However, White and Clark (1938), Merrill and Johnston (1939), Meader and Darrow (l9h2) and Cremmins (1952) indicated that very little to no difference could be realized in the percent of fruit set due to self- or cross-pollination. Merrill (1936) and Bailey (1938) concluded that open-pollination consisting of both self- and cross-pollination gave a better fruit set than did self-pollination, while Merrill (1936) also found open-pollination to be superior to cross-pollination when 8 referring to percent of fruit set. Thomas (1962) and Filmer and Marucci (1963) further indicated that their cage tests containing honey bees showed a desirable set could be obtained from self-pollination. In reference to the time needed for ripening and the size of the berry developing from cross-pollination versus self-pollination, the majority of the investigators generally agreed that cross-pollination both increased size and shortened the ripening time. The only disagree— ment was from Merrill (1936) who claimed that in Michigan the berries were Just as large from selfing as from crossing. Merrill (1936) also claimed after observing h8 isolated Rubel bushes in the open, assumed to be self-pollinated, that the berries matured during the normal growing season. Young (1952) claimed that there appeared to be no correlation between maturity and the location of the fruit within the cluster, but noted earlier ripening and larger berries from those maturing in the sun to those growing in the shade. Shutak, Hindle and Christopher (1956) reported no correlation between berry size and location in the cluster, 8 but noted final size attained by the berry increased considerably after the start of the blue coloration. Due to variations in results by the many investigators, some of the recommendations for specific areas also varied. Bailey, Franklin and Kelley (l9hl) in Massachusetts, Christopher and Surtleff (1952) in Rhode Island, Shoemaker (1955) in North Carolina and Darrow (1962) with the U. S. D. A. recommended, based on the results of experiments, that under most conditions two or more varieties should be planted to insure cross-pollination. Darrow further noted, however, that growers have received heavy crops from solid blocks of Rubel and Jersey. In Michigan, Johnston (1959) and Bell and Johnston (1962) recommended that large blocks could be planted Ior, "cultivated blueberries are self-fertile so they need no cross-pollination." Johnston (1959) further recommended, however, that more than one variety should be planted to insure against the risk of crop failure. Pollinating Agents Coville (1910) stated that without insects or other pollinating aids, selfing of blueberries happened only occasionally. Coville (1921) continued that for a crop to be set under field conditions, insects were l5 needed to carry the pollen from flower to flower. Chandler and Mason (1935), Chandler (l9h3) and Lee (1958) referring to lowbush blueberries and Merrill (1936) referring to highbush blueberries agreed with Coville (1910) that some means of insuring pollination is necessary for production. Phipps (1930) recorded forty different species of insects visiting the lowbush blueberry, Vaccinium lamarckii (Camp), in Maine. Phipps (1930) and Chandler and Mason (1935) and Chandler (l9h3) noted in remote areas blueberries in small acreages were successfully pollinated by native wild bees, but in large blocks there were insufficient numbers of wild bees to adequately pollinate the crop. Phipps (1930) suggested that honey bees might be provided to restore the balance between bloom and pollinators. Beckwith (1930) and Shaw, Bailey and Bourne (1939) obtained a good set when honey bees were used alone and little or no set in areas where they were excluded. Thomas (1962) and Filmer and Marucci (1963) found that a highly desirable set was obtained in self-pollination tests using honey bees only. Coville (1921), Crowley (1933) and Merrill (1936) stated however, that while honey bees did work blueberry flowers to some extent, the bulk of the pollination is left up to bumble bees. Coville (1921) and Merrill (1936) further stated that honey bees are incapable of pollinating certain varieties because a honey bee's tongue is too short to fit into the narrow, deep blossoms. Coville (1921) also noted that some blueberry pollination is done by solitary bees who are small enough to crawl into the flowers. {.4 OK Correlation Between.Berry Size and Seed Number Merrill (1936) reported that there seemed to be no relationship between the size of the berry and the number of seeds. White and Clark (1939) and Darrow (19h0), however, stated that their tests showed that the larger berries had either more seeds or the seeds in the larger berries weighed more than the smaller berries. Morrow (l9h3) and Meader and Darrow (l9h9) reported that berries developing from cross-pollination contained a greater percent of fully develOped seeds than berries from self-pollination. Results by Thomas (1962) indicated that the final size of the berry is related to the large brown seeds it holds but not correlated with the total seed number. Thomas (1962) also noted that the "number of large brown seeds in a berry is considered to be a possible index to the adequacy of the pollination." Filmer and Marucci (1963) stated that the size of the berries was directly proportional to the number of seeds. In summary, all investigators except Merrill (1936) believed that usually the larger berries contained more large seeds and probably more total seeds than the smaller berries. They also correlated greater weight of the seeds with greater size of the berry. Thomas (1962), while not agreeing completely with the other investigators, did believe that the number of large seeds gave an indication of the completeness of pollination. MATERIALS AND METHODS Controlled Pollination Study The study was carried out during the 196% season on a blueberry plantation at Grand Junction, Michigan. The plantation was owned and operated by John Nelson, Research Director of the Michigan Blueberry Growers' Association. A total of twenty-two six year old, mature, bearing bushes of the Jersey variety were utilized in the study. The bushes were chosen by visual observation on the bases of uniformity in size and number_of buds. Four, h' by 8' by 6' cages were constructed using 16 mesh fiberglas screen supported on a wooden frame of l by 2's. Five other slightly larger cages previously constructed, four measuring 8' by 6' by 8' and one h' by 20' by 6' were also used. The cages were placed over the bushes on May 9, prior to bloom, to exclude all pollinating insects (Plate 2). The cage studies were set up utilizing two separate cages for each individual cross- and self- pollination test. The cages were constructed so that the larger ones containing two equivalent bushes provided as much area per bush as the smaller cages which contained only a single bush. The two cages used in each test, containing a total of three bushes, constituted three replications. In the cross-pollination tests bouquets of Rubel, Coville or Bluecrop were used with the field grown Jersey bushes, each cross being replicated three times. 17 l8 .mumou cofiuucadaoa pHon :« pom: momma wcaaonm coHuauceHa xuuonosfinun.~ MHwc wcaaozm mumou seguecgaaoasmwouo ca pom: owes mo asuomoHouu.m mHoo x season He-ss nae n-o e-m e-e Ne-e seesaw - sadness x season mm~-a x-o »H-o ma-e H-n ama-n season - Hosea a season an.~. mN-Ha woo e-o ~H-H H-e H-m season - noses x season m~-ms «-6 a-e m-~ was-“ e-a banana - Hesse x assess and: fiuuoa amuOH an wa< NH ws4 <.ms< an haan ed hash coauunaaaom one: seasons not anon» mung IBM: noon usonuwk use £uq3 acauucaaaoa gamma pom twamm .ueuowo mo muse waaxuwa some new panda usuam Huspw>apau waaaofia «Hamauu.H quou x H ,ooaqoe oo.~¢e oo.wmo oo.~¢~ oc.omw «.9» o.mH H.wn m.H~ H.no m.¢m m.wm m.~¢ o.~m e.w¢ Honnm N H n e m N H .A.m .95 3&6 .nd .A.m .95 .A.m .95 .q.m 5.4 eoxufim magnumm 0H How assoc poem and: an w=< ma mn< e wn< «N hash- ca hash :Owueaafiaom IIIJI mama mhus moon on ene-nadn.wo-doHuun«HHOAInoao.p=u undue .umnouu mo newness 0H How usSOO puma some one ou spoon A.A.mv unmwa Hanan use A.n.qv sump owunfi mo unmouoa ecu mo nannsOHunHomu-.c Mgn<fi 33 The results of this study indicated that the final size of the berry (cup count) is related to the percent of large dark seeds it contains (Table 7), also shown graphically in Figure 2. A consistent correlation between the size of the berry and the total number of seeds can also be made. TABLE 7.--Relationship of percent of large dark seeds, mean cup count, and the time of ripening. WITH BEES Pollination July 16 July 2h Aug h Aug 17 Ana 31 L.D. Cup L.D. Cup L.D. Cup L.D. Cup L.D. Cup Count Count Count Count Count J X Rubel h8.h 102.3 hl.3 117.3 3h.9 156.0 21.9 215.h* 13.6 703.3* J X Coville h3.9 110.7 38.h 118.3 3h.1 1h2.0 25.1 182.0 22.3 306.0* J x Bluecr0p h8.7 110.0 u2.h 115.0 30.7 1h6.3 17.2 229.8* 17.1 398.0* A A J x Self h1.9 102.7 32.2 118.7 22.0 1h7.7 16.8 200.8* 15.2 68h.7* J - Open 3h.2 97.3 22.7 119.0 17.2 197.8 12.1 275.3 11.8 338.0* WITHOUT BEES 328.2* --- h95.5* Jersey --- --- 21.1 3hl.3* 9.2 290.8 9.1 .- A *indicates the cup count was calculated ‘- 3A 550- 475-- No Bees 400- I I 3251 Jersey - open I 250‘ Jersey X Bluecr0p Cup Count Jersey X Rubel Jersey - self 175‘ ..___ Jersey X Coville I 100‘ who T O I i t ‘- t 10 20 30 4O 50 Percent Large Dark Seeds FIGURE 2.--Relationship between the percent of large dark seeds and the mean cup count 35 Bee Activity in Cage Studies The bee activity in the cages was observed each day from May 15 (when the colonies were introduced into the cages) until May 26 (when blossoming was finished and the cages removed). 000d bee activity was noted throughout most of the day, the only exception being from 12 noon to 2:00 p.m. when the highest temperatures occurred. Honey bees could be observed through the screen working both the naturally growing Jersey bushes and the bouquets (Plate h). The bees could be observed at this time busily performing both self- and cross- pollination. Bee Activity inngen-Pollination Study Bee activity was also observed for the six Jersey bushes used in the open-pollination study. Observations began on what was estimated as beginning of bloom, May 11, and continued until petal fall, May 26. May 19 was estimated as full bloom (Plate 5). Bee activity was recorded as the total number of bees observed working each of the six bushes for one minute between the hours of 10:00 a.m. and 12 noon (Table 8). A sampling of the bees using a sweep net was made. The bees were identified and listed from the most to the least abundant (Table 9). No attempt was made to collect those specific bees observed in pollinating the six Jersey bushes. It should be mentioned that Mr. Nelson, operator of the test plantation site, noted much higher honey bee activity in the Jersey field during the 196R blossoming season than in previous years. This increase .nmsn >mmuon powno onu weaxuoa own mono; a weaaonm season ammo osu swaounu 3ofi>-.e mH