HORTICULTURAL ENTOMOLOGY Pollination Efficiencies of Three Bee (: Apoidea) Species Visiting Rabbiteye Blueberry

1 2 BLAIR J. SAMPSON AND JAMES H. CANE

J. Econ. Entomol. 93(6): 1726Ð1731 (2000) ABSTRACT Inadequate bee pollination limits rabbiteye blueberry, Vaccinium ashei Reade, pro- duction in the some areas of the southeastern United States. Honey bees, Apis mellifera L., are currently the only manageable pollinators available for pollinating V. ashei. However, a new adaptable pollinator for rabbiteye blueberry, Osmia ribifloris Cockerell, was successfully reared and ßown in captivity. The bee nested successfully in wooden shelters and conferred superior fruit set to 2-yr-old potted, rabbiteye blueberry bushes. Pollination efÞciency or the percentage of blueberry ßowers to set fruit after being visited once by a female O. ribifloris was comparable to that of the female blueberry bee laboriosa (F.) and worker honey bees. Interestingly, honey bees once thought to be inefÞcient pollinators of rabbiteye blueberry were found to be very efÞcient, especially for ÔClimaxÕ and ÔPremierÕ ßowers.

KEY WORDS Apis mellifera, , Osmia ribifloris, blueberry, pollination, fruit set

BLUEBERRY PRODUCERS HAVE traditionally relied on wild H. laboriosa by active management is difÞcult and bees for blueberry pollination. Many native bee spe- might further be hindered by a tendency of females to cies are highly efÞcient pollinators with single visits to abandon their burrows after disturbances to nesting virgin ßowers setting Ͼ40% fruit (Cane and Payne substrates (Cane 1995). A very promising and new 1990, Payne et al. 1991); repeated or unrestricted visits pollinator for blueberries is a solitary, cavity-nesting can further improve fruit set and size (Danka et al. bee indigenous to the western United States, Osmia 1993, Stubbs and Drummond 1999). For many of the ribifloris Cockerell (Torchio 1990). O. ribifloris co- SoutheastÕs Ϸ10,000 acres of blueberry farmland, pol- coons can be safely harvested and transported and lination problems arise in larger Þelds where there are after a brief courtship period adult females will nest in too few bees to pollinate available blossoms. Clepto- portable wooden shelters. The bee is native to the parasites, pesticides, and competing bloom also con- United States and is subject to less regulation than the tribute to lower pollinator abundance in blueberry A. pilipes and is far easier to manage than H. laboriosa. Þelds (Parker et al. 1987; Payne et al. 1989; Spiers 1990; Osmia ribifloris is a pollen specialist of plants like Torchio 1990; Cane 1993, 1995, 1997; Cane and Payne blueberries that are in the heath family (Ericaceae). 1993; Batra 1995, 1997; Kevan et al. 1997; Pittman et al. Although the single-visit pollination efÞciency of O. 1998). ribifloris females at blueberry ßowers has not been The long-tongued bees Anthophora pilipes Smith determined, when bees were given unlimited and sole (Anthophoridae), Habropoda laboriosa (F.) (An- access to ßowers of rabbiteye blueberry, Vaccinium thophoridae), and Osmia ribifloris Cockerell ashei Reade, average fruit set was 54% (Sampson et al. (Megachilidae) have been evaluated as supplemental 1995). Also, adult emergence, courtship, foraging and pollinators of the major North American blueberry oviposition for this bee can be timed to coincide ex- crops (Cane and Payne 1990, Torchio 1990, Batra 1994, actly with rabbiteye blueberry ßowering. Adult O. Stubbs et al. 1994, Cane 1997, Stubbs and Drummond ribifloris live Ϸ2Ð3 wk (Torchio 1990, Stubbs et al. 1999). A. pilipes nests in portable clods of adobe clay 1994) and their offspring overwinter as preemergent and pollinated both highbush and lowbush blueber- adults in silken cocoons. Thus, there is no need to feed ries (Batra 1994, 1997; Stubbs and Drummond 1999). O. ribifloris beyond blueberry bloom (Parker et al. Originally from Japan, A. pilipes is currently in quar- 1987). antine and unavailable to blueberry producers To justify releasing O. ribifloris into southern blue- (Stubbs and Drummond 1999). H. laboriosa is a wild berry orchards, the beeÕs relative pollination efÞ- and efÞcient native bee pollinator that, although spe- ciency must Þrst be compared with already estab- cializing on southern blueberries, is often rare in many lished and important blueberry pollinators. The two commercial blueberry Þelds. Bolstering populations of important southern pollinators are the native H. la- boriosa, and European honey bee, Apis mellifera L. 1 Small Fruit Research Station, USDA-ARS, Poplarville, MS 39470. The percentage of fruits set after single visits by bees 2 Bee Biology and Systematics Laboratory, USDA-ARS, Utah State to virgin rabbiteye blueberry ßowers is a useful and University, Logan, UT 84322Ð5310. standardized measure of pollination efÞciency (Cane December 2000 SAMPSON AND CANE:POLLINATION EFFICIENCIES OF THREE BEE SPECIES 1727

1997). Xenia pollen effects in blueberry fruit set (Gup- O. ribifloris subspecies between the years 1999 and ton 1984, 1997) should not seriously affect compari- 2000. All females were released during blueberry sons of single-visit pollination efÞciency for this and bloom. We also obtained 565 males, of which one-third other studies, because the interplanting of only a few were released in the screenhouse for mating purposes. highly compatible V. ashei cultivars is consistent Nectar and pollen for the bees was supplied by nearly throughout the Southeast. The three most common equal proportions of 2-yr-old ÔClimaxÕ, ÔPremierÕ, and cultivars ÔTifblueÕ, ÔPremierÕ, and ÔClimaxÕ are largely ÔTifblueÕ plants (n ϭ 300 plants). Portable nesting self-incompatible and currently comprise 80% of the shelters each housed 17 acrylic-coated, untreated pine rabbiteye blueberry production in the southeastern blocks. Each block provided female O. ribifloris with United States (Spiers 1990). Þve or 10 tubular nests lined with paper straws (0.7 cm diameter by 15.0 cm long). Potted strawberry, rose, oak, and southern highbush blueberry plants supplied Materials and Methods female O. ribifloris with suitable leaf material to fash- Fruit set resulting from single visits by bees to 887 ion nest cell partitions and entrance plugs. Additional ßowers occurred on potted, 2-yr-old ßowering V. ashei nest shelters and blocks were provided as needed. plants from the commercial cultivars Climax, Premier, Tanglefoot barriers (Tanglefoot, Grand Rapids, MI) and Tifblue. Single-visit fruit set is a standardized prevented predation of bee brood by the red imported measure of pollination efÞciency that is independent Þre ant, Solenopsis invicta Buren. Single visits by O. of pollinator foraging density. Plants were exposed to ribifloris occurred at 348 virgin ßowers from a total of the same environmental conditions and handling to 134 racemes from 15 plants per cultivar. ensure that management practices did not contribute The second study site was at two nearby rabbiteye to differences in fruit set. Only virgin ßowers on plants blueberry plots (USDA-ARS Small Fruit Research Sta- received single visits by bees; plants were left in the tion in Poplarville, MS). The plots contained Ϸ200 Þeld for only a few hours each day and then returned blueberry plants ranging from 2 to 15 yr of age, hosted to a common area (screenhouse) to mature fruit. All a large foraging population of H. laboriosa, and pro- potted plants received the same amount of water and vided highly compatible pollen sources (ÔClimaxÕ, fertilizing. Intraclonal variability was reduced by se- ÔPremierÕ, ÔTifblueÕ, ÔWoodardÕ, and ÔDeliteÕ). We ob- lecting plants of the same age, plants growing in the served single visits to 434 ßowers by H. laboriosa at 72 same type and size of pot, same soil medium and from racemes from six plants, two plants for each cultivar. the same mother plants and nursery. Plants were returned to a screenhouse so they could On 36 racemes, 228 virgin ßowers were enclosed in mature fruit under the same conditions as plants used Þnely woven mesh bags to completely exclude polli- to evaluate captive O. ribifloris already conÞned inside nators (unvisited control). Caging plants within the screenhouse as well as plants visited by honey bees 1.5-m3 Lumite enclosures (Synthetic Industries, from another site. Gainesville, GA) greatly increased the number of vir- The third and most remote site in 1999 was a com- gin ßowers that were available to bees. Any open mercial blueberry farm, 100 km SE of the Poplarville ßowers that could have been previously visited by an station. No screenhouse existed at this site; O. ribifloris pollinator were removed before caging. were not released and Habropoda were entirely ab- When sufÞcient ßowers had opened, random plants sent. This provided a unique opportunity to test the were brought to each of three study sites. Bags or cages pollination efÞciency of honey bees without native were then removed to expose virgin ßowers to single blueberry pollinators being present, a situation that pollination visits by both nectar and pollen foraging normally necessitates the renting of honey bee colo- female bees. Male bees rarely visited our virgin ßowers nies for blueberry pollination. Commercial honey bee and so were not included in the data analysis. colonies were clustered along the tree line at a stock- The Þrst study site was inside a 133-m2 screenhouse ing rate of 1.2 colonies per acre. Compatible pollen at the USDA-ARS Small Fruit Research Station, Pop- sources at the third site included ßowers of 18-yr-old larville, MS, where potted plants were to ripen fruit ÔPremierÕ, ÔClimaxÕ, ÔTifblueÕ, and ÔBrightwellÕ rabbit- after ßowers received single pollinator visits. Because eye blueberry. One-hundred-Þve single ßoral visits by O. ribifloris was not native to Mississippi, USDA- honey bees at 24 racemes occurred on three plants, APHIS-PPQ and the Mississippi Department of Agri- one plant per cultivar. Unfortunately, we could not culture (MDAC) gave permission only to release O. spare additional plants for evaluating honey bees in ribifloris inside a screenhouse. Adult O. ribifloris were 1999; however. these healthy plants had the same brought out of dormancy by incubating cocoons at genetic make-up as the plants used to assess the pol- 21Ð28ЊC over a 15-d period beginning 19 February and lination efÞciencies of the other bee species. ending the 20 March for both 1999 and 2000. Osmia All ßowers probed by a bee on each raceme were ribifloris ribifloris Cockerell were originally from east- counted and any unvisited virgin ßowers were re- ern Texas in 1994 and propagated for 5 yr in a screen- moved. The percentage of pea-sized, stage I fruit house (Auburn University, AL). Another subspecies (29Ð30 d after pollination, Eck 1986) developing from (O. r. biedermannii Michener) was trap-nested in the all probed ßowers was a measure of pollination efÞ- Catalina Mountains near Tucson, AZ, in 1998, as well ciency for each bee species. We chose not to use ripe as Calaveras County, CA, and the San Rafael Desert, fruit set as a measure of pollination efÞciency, because UT, in 1999. We obtained 181 live females from both the risk of fruit loss is higher for potted plants. How- 1728 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 93, no. 6

Fig. 2. Seed set per fruit resulting from no visits or single pollinator visits to rabbiteye blueberry ßowers. Bars repre- sent mean seed set ϮSE for X number of fruit (shown as white numerals). Letters above the bars that are the same Fig. 1. Mean percent of rabbiteye blueberry ßowers to represent mean seed sets that are not different using Tukey set green fruit after single visits from three bee species or no HSD test (P Յ 0.05). visits (unvisited). Bars represent mean fruit set ϮSE for all cultivars (ÔTifblueÕ, ÔPremierÕ, and ÔClimaxÕ) and each indi- vidual cultivar. Astertisks denote cultivars that produced no df ϭ 3, 254; P Ͻ 0.0001). An overall 4 Ϯ 2% of unpol- fruit (0% fruit set). Letters above the black bars that are the linated V. ashei ßowers produced fruit with few to no same represent overall fruit sets that are not different using Ϯ Ϸ Tukey HSD test (P Յ 0.05). seeds (14 9 seeds per fruit), showing that 96% of ßowers receiving no pollinator visits failed to set fruit after 30 d. Bagged ÔPremierÕ and ÔClimaxÕ ßowers pro- ever, advanced green fruit set should accurately re- duced no fruit. Only Tifblue racemes were self-fruitful ßect optimal ripe fruit set for our potted plants be- among our small potted rabbiteye blueberry plants. cause premature fruit drop caused by poor pollination When we included ßowers that received pollination (i.e., shot-berries) should have occurred well before or not, percentages of fruit set were uniform for our counting (Eck 1988). Ripe fruit were harvested in small potted rabbiteye blueberry plants (F ϭ 0.59; df ϭ June 1999Ð2000; and the number of mature, intact 2, 254; P ϭ 0.5576). Blooms from 80 ÔClimaxÕ racemes seed was counted to assess seed set attributed to single set an average of 39 Ϯ 5% fruit. ÔPremierÕ blossoms pollinator visits. Data from both years (1999Ð2000) from a total of 52 racemes set 38 Ϯ 5% fruit, and 134 were pooled to increase sample size for O. ribifloris. racemes containing ÔTifblueÕ ßowers set 36 Ϯ 3% fruit. We improved the normality of the percent green fruit An interaction between pollinator species and rabbit- set data by transforming them before analysis using eye blueberry cultivar (F ϭ 3.79; df ϭ 6, 254; P ϭ ϭ ϩ the equation YT square root (Y 0.5) where YT is 0.0012) indicated that some bee species were more the transformed percentage and Y was the original efÞcient pollinators at speciÞc cultivars. The peculiar, percentage (Lyrene 1989). Two-way analysis of vari- slipper-shaped ßowers of ÔPremierÕ that sometimes ance showed how the cultivars, different pollinators, develop on younger plants and later in the season the unvisited control and their interactions inßuenced often bafßed foraging Osmia. Single visits by O. ribi- percentage fruit set and seed set per ripe berry. The floris to ÔPremierÕ ßowers resulted in lower fruit set Tukey honestly signiÞcant difference (HSD) test sep- (16%), which was still greater than if ÔPremierÕ ßowers arated means at P Յ 0.05 level of signiÞcance (PROC were not pollinated. Honey bees were less efÞcient GLM, SAS Institute 1985). pollinators of ÔTifblueÕ ßowers in our study; and in another unpublished study, worker honey bees readily avoided or robbed ÔTifblueÕ ßowers in a com- Results mercial blueberry Þeld (B.J.S., unpublished data). H. Overall, O. ribifloris was an efÞcient pollinator of V. laboriosa were the most consistent pollinators at all ashei with single visits to ßowers setting 44 Ϯ 4% fruit three V. ashei cultivars (Fig. 1). (Fig. 1). Overall pollination efÞciency for O. ribifloris Fruit resulting from O. ribifloris visits had fewer at rabbiteye blueberry ßowers did not differ from the seeds than those fruit resulting from single-visits from fruit set levels shown by the established eastern pol- the other pollinators assessed, and it did not differ linators A. mellifera and H. laboriosas (Fig. 1). Fruit set from the unpollinated control (Fig. 2; F ϭ 6.77; df ϭ values resulting from single visits to ßowers made by 3, 168; P ϭ 0.0002). These values were biased by the Osmia, honey bees (32 Ϯ 7%) and H. laboriosa (44 Ϯ fact that Habropoda and honey bees more frequently 4%) were much higher than ßowers from the 36 visited the ßowers of the earlier-maturing and seedier racemes that received no pollinator visits (F ϭ 8.65; varieties ÔPremierÕ (77 Ϯ 3.5 seeds per fruit) and ÔCli- December 2000 SAMPSON AND CANE:POLLINATION EFFICIENCIES OF THREE BEE SPECIES 1729 maxÕ (55 Ϯ 3.8 seeds per fruit) (F ϭ 4.25; df ϭ 3, 168; hanced blueberry fruit set and yield in cages and in the P ϭ 0.0064). Flowers from the later season cultivar Þeld (Shaw et al. 1939, Martin 1966, Marucci 1966, ÔTifblueÕ produced fruit with fewer seeds (35.2 Ϯ 2.4 Whatley and Lackett 1978, Davies and Buchanan 1979, seeds per fruit) and these ßowers were preferred by Goodman and Clayton-Greene 1988, Lang and Danka female O. ribifloris (F ϭ 3.76; df ϭ 2, 168; P ϭ 0.0253). 1991). These Þndings conßict with a study showing honey bees to be less effective rabbiteye blueberry pollinators than native bees (Cane and Payne 1990). Discussion It is hypothesized that in areas abundant with native Bees are essential components to rabbiteye blue- bees like Habropoda, pollen standing crops may be berry production. Without cross-pollination by bees, depleted to a point that worker honey bees are unable 96% of ßowers failed to set fruit. EfÞcient pollinators to passively obtain enough pollen needed for satisfac- such as O. ribifloris are the most desirable because tory pollination. Additionally, no carpenter bee rob- their pollination efÞciency is comparable to the in- bery holes were available to tempt workers into rob- valuable H. laboriosa, and they are manageable. Tor- bing blossoms on our 2-yr-old V. ashei plants. Honey chio (1990) recognized the potential beneÞts of man- bees more frequently robbed or avoided ÔTifblueÕ aging O. ribifloris for blueberry pollination, ßowers in the Þeld, suggesting that this cultivar, al- developing a management program that recom- though popularly grown, is a less suitable ßoral host for mended releasing as few as 300 female O. ribifloris per honey bees (B.J.S., unpublished data). The intriguing acre to pollinate highbush blueberry, Vaccinium possibility that Habropoda and other native bees may corymbosum L. Greater densities of this bee might be have a competitive edge over honey bees at ßowering needed for pollinating rabbiteye blueberries that are V. ashei, especially at ÔTifblueÕ blossoms, and that these less self-fruitful than highbush blueberry (Morrow competitive interactions among pollinating bee spe- 1943, MacKenzie 1997). An estimated stocking density cies can affect pollination efÞciency warrants further from one to two O. ribifloris females per bush (Ϸ650Ð study. Breeding blueberries with ßoral traits compat- 1,300 females per acre) should be enough for polli- ible with speciÞc pollinators or relying on two or more nating V. ashei. The lower estimate is based on the abundant bee species for crop pollination would be number of nest provisioning females that can be sus- optimal strategies for obtaining adequate and uniform tained with the Ϸ10,000 ßowers produced from each fruit and seed production for various rabbiteye blue- mature V. ashei bush. The basis for setting an upper berry cultivars. Conserving and possibly managing H. limit of two bees per bush was to ensure that ßowers laboriosa populations would be the best solution to receive uniform and adequate pollination. A captive promote rabbiteye blueberry pollination in the South- population of female O. ribifloris equivalent to two east because of this beeÕs consistent ability to effec- bees per bush set fruit on 54% of openly visited V. ashei tively pollinate and set seed for the three major V. ßowers (Sampson et al. 1995). This level of fruit set ashei cultivars. would have been even higher if some of the pollinated The pollination efÞciency of H. laboriosa appears to ßowers were not destroyed by Botrytis blossom blight, be within the range (28Ð51%) previously measured an aggressive fungus that is problematic for blueberry for this bee elsewhere (Cane and Payne 1990, Payne ßowers under cool (15Ð20ЊC) and humid conditions et al. 1991, Cane 1997). Thus, pollination efÞciency (Ͼ95%) (Smith 1998). based on percentage fruit set for this oligolectic pol- The stocking density for O. ribifloris could be period- linator is a consistent indicator of this speciesÕ ability ically adjusted to help optimally supplement varying to reliably enhance blueberry crop yield at different background levels of pollination achieved by H. labo- varieties and in different blueberry growing regions riosa, and other wild bees. Where native bee pollinators throughout the southeastern United States. Consis- are increasingly abundant, fewer female O. ribifloris per tency in pollination efÞciency for H. laboriosa reßects acre would have to be released. A visual census showing this beeÕs uniform size, rapid foraging, synchrony be- an average of one or more bumble bees or H. laboriosa tween adult emergence and blueberry ßowering, and females at each bush are desirable bee densities for com- innate preference for wild and cultivated Vaccinium mercial V. ashei pollination (Cane 1993). Average den- (Cane and Payne 1988, Cane 1995). sities below one native bee per bush would justify aug- Although H. laboriosa are currently unmanaged, mentation with O. ribifloris or honey bees. conditions favoring their management include a lack There has been some question about a honey beeÕs of cleptoparasites and a potential for females to adopt ability to efÞciently pollinate blueberry. Despite holes artiÞcially dug into sand beds (Cane 1995). Un- larger colony sizes for A. mellifera (Winston 1987), like O. ribifloris cocoons that can be shipped undam- potential problems with using honey bees as V. ashei aged, the earthen cells of H. larboriosa are perhaps too pollinators include a rarity of pollen-foragers at blue- delicate to be harvested and moved, and even slight berry ßowers, their stronger preference for competing disturbances to the surrounding soil could disorient bloom (Payne et al. 1991), and repeated probing of nesting females (Cane 1995). Instead, unmated adults robbery holes made by male carpenter bees, Xylocopa could be relocated to blueberry Þelds having a nearby virginica (L.). However, our data show honey bees raised sand bed serving as suitable nesting substrate can be efÞcient V. ashei pollinators, especially for for female H. laboriosa to dig their burrows (Cane ÔClimaxÕ and ÔPremierÕ rabbiteye blueberries. Honey 1995). Finding large enough wild populations to min- bees at various blueberry species and cultivars en- imize disruption of populations could be difÞcult. 1730 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 93, no. 6

However, upon successful transplantation of adult honey bee (Hymenoptera:) pollination of berry bees, both male and female H. laboriosa could provide crops. J. Econ. Entomol. 85: 1300Ð1306. unrivaled pollination service to blueberry producers. Danka, R. G., G. A. Lang, and C. L. Gupton. 1993. Honey Should such efforts prove impractical, O. ribifloris will bee (Hymenoptera: Apidae) visits and pollen source ef- soon be available for southern blueberry pollination. fects on fruiting of ÔGulfcoastÕ southern highbush blue- Honey bees can also be effective blueberry pollina- berry. J. Econ. Entomol. 86: 131Ð136. tors. The challenge to releasing honey bees is to keep Davies, F. S., and D. W. Buchanan. 1979. Fruit set and bee activity in four rabbiteye blueberry cultivars. Proc. Fla. them in ßowering blueberry Þelds where more favor- State Hortic. Soc. 92: 246Ð247. able host plants compete with rabbiteye blueberry for Eck, P. 1986. Blueberry: pollination requirements, pp. 75Ð their pollination services. One solution we are testing 85. In CRC handbook of fruit set and development. CRC, is applying chemical lures to entice honey bees to visit Boca Raton, FL. and subsequently pollinate rabbiteye blueberry ßow- Eck, P. 1988. Blueberry Science. Rutgers University Press, ers. Synthetic queen mandibular pheromone is one of New Brunswick, NJ. the most promising lures, which when sprayed on Goodman, R. D., and K. A. Clayton-Greene. 1988. Honey- blueberry bushes around midbloom can enhance bee pollination of highbush blueberries (Vaccinium honey bee visitation, fruit set, and yield. Synthetic corymbosum). Aust. J. Exp. Agric. 28: 287Ð290. queen pheromone has proven to be most successful Gupton, C. L. 1984. Effect of pollen source on fruit char- and cost-effective for pollinating northern highbush acteristics of low-chilling highbush type blueberries. Hort blueberries when honey bee densities are low (Curry Science 19: 531Ð532. Gupton, C. L. 1997. Evidence of xenia in blueberry. Acta et al. 1992). Hortic. 446: 119Ð123. Kevan, P. G., C. F. Greco, and S. Belaossoff. 1997. Log- normality of biodiversity and abundance in diagnosis and Acknowledgments measuring of ecosystemic health: pesticide stress on pol- The authors are grateful to Russell GrifÞn for the use of his linators on blueberry heaths. J. Appl. Ecol. 34: 1122Ð1136. blueberry Þeld; Danita Lewis and Brandy Powell for their Lang, G. A., and R. G. Danka. 1991. Honey bee-mediated workshop, laboratory, and Þeld assistance; and Howard Bond cross- versus self-pollination of ÔSharpblueÕ blueberry in- for helping design and build the Osmia nest shelters. We also creases fruit size and hastens ripening. J. Am. Soc. Hortic. thank Steve Buchmann and Don Viers for helping to screen Sci. 116: 770Ð773. the Osmia cocoons. The authors thank Robert Danka and Lyrene, P. M. 1989. Pollen Sources Inßuences Fruiting of John Braswell for their critical reviews of the manuscript. ÔSharpblueÕ Blueberry. J. Am. Soc. Hortic. Sci. 114: 995Ð999. MacKenzie, K. E. 1997. Pollination requirements of three highbush blueberry (Vaccinium corymbosum L.) culti- References Cited vars. J. Am. Soc. Hortic. Sci. 122: 891Ð896. Martin, E. C. 1966. Honey bee pollination of the highbush Batra, S.W.T. 1994. Anthophora pilipes villosula Sm. (Hy- blueberry. Am. Bee J. 106: 366Ð367. menoptera: Anthophoridae), a manageable Japanese bee Marucci, P. E. 1966. Blueberry pollination. Am. Bee J. 106: that visits blueberries and apples during cool, rainy spring 250Ð251. weather. Proc. Entomol. Soc. Wash. 96: 98Ð119. Morrow, E. B. 1943. Some effects of cross-pollination versus Batra, S.W.T. 1995. Bees and pollination in our changing self-pollination in the cultivated blueberry. Am. Soc. Hor- environment. Apidologie 26: 361Ð370. tic. Sci. 42: 469Ð472. Batra, S.W.T. 1997. Solitary bees for Vaccinium pollination. Parker, F. D., S.W.T. Batra, and V. J. Tepedino. 1987. New Acta Hortic. 446: 71Ð76. pollinators for our crops. Agr. Zool. Rev. 2: 279Ð303. Cane, J. H. 1995. Nesting biology and mating Behav. of the Payne, J. A., J. H. Cane, A. A. Amis, and P. M. Lyrene. 1989. southeastern blueberry bee, Habropoda laboriosa (Hy- Fruit set, seed size, seed viability and pollination of rab- menoptera: Apoidea). J. Kans. Entomol. Soc. 67: 236Ð241. biteye blueberries (Vaccinium ashei Reade). Acta Hortic. Cane, J. H., and J. A. Payne. 1988. Foraging ecology of the 241: 38Ð43. bee Habropoda laboriosa (Hymenoptera: Anthophori- Payne, J. A., D. L. Horton, J. H. Cane, and A. A. Amis. 1991. dae), an oligolege of blueberries (Ericaceae: Vaccinium) Insect pollination of rabbiteye blueberries, pp. 42Ð49. In in the Southeastern United States. Ann. Entomol. Soc. Am. 81: 419Ð427. Proceedings, 1991 Missouri Small Fruit Conference: Cane, J. H., and J. A. Payne. 1990. Native bee pollinates 19Ð20 February 1991. UMC University Extension, Mis- rabbiteye blueberry. Ala. Agric. Exp. Stn. Auburn Univ. souri Department of Agriculture, Division of Marketing Highlights Agric. Res. 37: 1Ð4. and Development, Southwest Missouri State University, Cane, J. H. 1993. Strategies for more consistent abundance SpringÞeld, MO. in blueberry pollinators. Proceedings, 6th Biennial South- Pittman, P. O., J. H. Braswell, and J. M. Spiers. 1998. Survey east Blueberry Conference and Trade Show, Tifton, GA. of major small fruit crops grown in the United States, pp. Cane, J. H. 1997. Lifetime monetary value of individual pol- 1Ð10. In Proceedings, 8th North American Blueberry Re- linators: the bee Habropoda laboriosa at rabbiteye blue- search and Extension Conference. Wilmington, NC. berry (Vaccinium ashei Reade). Acta Hortic. 446: 67Ð70. SAS Institute. 1985. SAS userÕs guide: statistics version 5 ed. Cane, J. H., and J. A. Payne. 1993. Regional, annual, and SAS Institute, Cary, NC. seasonal variation in pollinator guilds: intrinsic traits of Sampson, B., J. Cane, and J. Neff. 1995. Blue bees for blue- bees (Hymenoptera: Apoidea) underlie their patterns of berries. Ala. Agric. Exp. Stn. Auburn Univ. Highlights abundance at Vaccinium ashei (Ericaceae). Ann. Ento- Agric. Res. 42: 12, 13, 15. mol. Soc. Am. 86: 577Ð588. Shaw, F. R., J. S. Bailey, and A. I. Bourne. 1939. The com- Curry, R. W., M. L. Winston, and K. N. Slessor. 1992. Effect parative value of honeybees in the pollination of culti- of synthetic queen mandibular pheromone sprays on vated blueberries. J. Econ. Entomol. 32: 872Ð874. December 2000 SAMPSON AND CANE:POLLINATION EFFICIENCIES OF THREE BEE SPECIES 1731

Smith, B. J. 1998. Botrytis blossom blight of southern blue- Torchio, P. F. 1990. Osmia ribifloris, a native bee species berries: Cultivar susceptibility and effect of chemical developed as a commercially managed pollinator of high- treatments. Plant Dis. 82: 924Ð927. bush blueberry (Hymenoptera: Megachilidae). J. Kans. Spiers, J. M. 1990. Rabbiteye blueberry. Fruit Var. J. 44: Entomol. Soc. 63: 427Ð436. 68Ð72. Whatley, B. T., and J. J. Lackett. 1978. Effects of honey bee Stubbs, C. S., and F. A. Drummond. 1999. Pollination of pollination on fruit set, yield and quality of rabbiteye lowbush blueberry by Anthophora pilipes villosula and blueberries Vaccinium ashei Reade. Md. Agric. Exp. Stn. Bombus impatiens (Hymenoptera: Anthophoridae and Spec. Misc. 1: 143Ð148. Apidae). J. Kans. Entomol. Soc. 72: 330Ð333. Winston, M. L. 1987. The biology of the honey bee. Harvard Stubbs, C. S., F. A. Drummond, and E. A. Osgood. 1994. University Press, Cambridge, MA. Osmia ribifloris biedermannii and Megachile rotundata (Hymenoptera: Megachilidae) introduced into lowbush blueberry agroecosystems in Maine. J. Kans. Entomol. Received for publication 6 December 1999; accepted 15 July Soc. 67: 173Ð185. 2000.