Faunal composition and species richness differences of bees (Hymenoptera: Apiformes) from two north American regions Robert Minckley To cite this version: Robert Minckley. Faunal composition and species richness differences of bees (Hymenoptera: Api- formes) from two north American regions. Apidologie, Springer Verlag, 2008, 39 (1), pp.176-188. hal-00891939 HAL Id: hal-00891939 https://hal.archives-ouvertes.fr/hal-00891939 Submitted on 1 Jan 2008 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Apidologie 39 (2008) 176–188 Available online at: c INRA/DIB-AGIB/ EDP Sciences, 2008 www.apidologie.org DOI: 10.1051/apido:2007062 Original article Faunal composition and species richness differences of bees (Hymenoptera: Apiformes) from two north American regions* Robert Minckley Department of Biology, University of Rochester, Rochester, 14627 NY, USA Received 11 July 2007 – Revised 9 October 2007 – Accepted 10 October 2007 Abstract – Host breadth and global bee species diversity are thought to be linked. Areas where bee species richness is greatest have a greater proportion of oligolectic species and fewer social species. I compared the bee faunas of two North American regions (one mesic, one xeric) and two nearby habitats (riparian and desert scrub). Species richness is greater in the xeric than in the mesic North American region. Despite strongly bimodal bloom in the xeric region and continuous bloom in the mesic region, their bee faunas were similar in the proportion of solitary oligolectic and polylectic bees. Oligolectic species of both areas have short lifespans. Social and cleptoparasitic species made up a greater percentage of the fauna in the mesic North American region. Nearby mesic and xeric habitat both had social species but xeric habitats were richer in oligolectic species. Phylogeny and historical biogeography in combination with ecology of bees and plants will be needed to understand differences of bee faunas. biodiversity / host specialization / species gradient / pollinator / oligolecty 1. INTRODUCTION relationships among bees and sister groups in the Spheciformes (sensu Michener, 2007) will Recent changes in our understanding of be needed to support or refute this latter hy- the evolutionary relations among bees indicate pothesis. Nevertheless, this points out how a that pollen host specialization (i.e., oligolecty) solid understanding of apoid phylogeny pro- occurred early in the evolutionary history of vides insights into our understanding of host bees, that shifts among pollen host general- breadth ecology and evolution among bees. ization (i.e., polylecty) and specialization have Pollen host breadth and worldwide bee di- occurred repeatedly (Danforth et al., 2006), versity are thought to be linked. Where bee and that the loss of specialization may be diversity reaches its maximum in xeric tem- as common as its gain (Müller, 1996; Larkin perate areas (Michener, 1954, 1979, 2000), et al., 2006). These findings indicate host the proportion of the solitary bee species that breadth in bees can respond quickly to eco- are pollen specialists is greatest (Moldenke, logical variability, and opens the tantalizing 1976). Furthermore, the proportion of the possibility that narrow host specificity origi- bee fauna that are social species is low. nated in the sphecid ancestors of bees and in Multiple, often interrelated, hypotheses have the context of hunting insects rather than in the been suggested to explain these patterns. context of harvesting pollen. Further work on Michener (1954, 1979) proposed that compe- tition among bee species for floral resources Corresponding author: R. Minckley, may generate the diversity gradient. Social bee [email protected] species are pollen generalists that require flo- * Manuscript editor: Bryan Danforth ral resources throughout the spring and sum- Online material is available at: mer as their colonies develop. Therefore, so- http://www.apidologie.org cial bees may be rare in xeric areas because Article published by EDP Sciences and available at http://www.apidologie.org or http://dx.doi.org/10.1051/apido:2007062 Bee faunal changes in mesic and xeric areas 177 floral resource availability can be seasonally species that are active either in the spring or scarce. This lack of social species in deserts late summer. Such low overlap of two fau- may result in less competition for available nas per year could further contribute to greater floral resources and, in turn, allow greater di- species richness. versification of solitary bees. In addition, the In this study, I compare the bee fau- high numbers of solitary bee species in xeric nas of a mesic region and a xeric region areas may allow more species to co-occur be- in North America to examine if these pre- cause available floral resources are divided dictions are supported. The mesic region is more finely. Carlinville, Illinois, a temperate area with con- Hypotheses unrelated to competition have tinuous flowering throughout the spring, sum- also been proposed to explain bee species di- mer and fall months. The xeric region is the versity gradients. First, specialist bee species San Bernardino Valley Sonora, Mexico and may be more common in xeric areas be- Arizona, USA (hereafter referred to as San cause they are able to predict more accu- Bernardino), an area with one spring bloom, rately when bloom on their host plant oc- an intervening dry period, and a second late- curs than can pollen generalist bee species summer bloom. Bee collections from Car- (Minckley et al., 2000). Deserts are charac- linville and San Bernardino were made from terized as having low overall annual precipita- limited areas that are not topographically com- tion and periods lasting months to years of lit- plex and have been continued for multiple tle or no precipitation. Minckley et al. (2000) years so are unusually well-studied for bees. showed that species richness and abundance of In addition to the comparison of two geo- pollen specialist bees that use Larrea triden- graphically distant regions in North America, tata was greatest where the host plant bloomed I compare bee samples made from a mesic ri- least frequently, and suggested that the cues parian habitat and from a nearby xeric desert pollen specialist species use that signal if they scrub habitat at San Bernardino, to evalu- should either remain in diapause or emerge en- ate if patterns found between Carlinville and ables these species to persist in deserts. For San Bernardino are repeated at local spatial bees that specialize on one or several closely- scales. Because these plots share a common related plant species, the likelihood of evolv- pool of bee species, differences in the bee fau- ing cues that seasonally synchronize emer- nas should be primarily due to local ecological gence activity with host plant flowering should conditions. The specific predictions I examine increase and selection for this match of emer- with these comparisons are as follows. One, gence/diapause cues and host plant bloom are social species less common in xeric areas? should be strongest where rainfall/bloom is Support for this hypothesis will be if the pro- least predictable, a pattern with what Minckley portion of social species is significantly greater et al. (2000) found. Generalist bees do not ex- in Carlinville than in San Bernardino and in the perience the same level of selection to syn- riparian habitat more than desert scrub habi- chronize with bloom because they use multiple tat at San Bernardino. Two, are short-lived hosts. Similarly, under these same conditions species favored in xeric areas with a short, of unpredictable bloom, social species tend to discrete bloom period? This hypothesis pre- be excluded because floral resources are not dicts that a greater proportion of bee species available throughout the entire season during are long-lived in the Carlinville fauna than which colonies develop. the San Bernardino bee fauna, and more long- A second hypothesis proposed by Rozen (in lived species occur in mesic plots than in xeric Michener, 2007) suggests that ground-nesting plots in the San Bernardino. Finally, are pollen solitary species are excluded from tropical ar- specialist bee species more diverse in xeric eas because fungi and other pathogens that habitats? This prediction will be supported if attack nest provisions are more common the proportion of specialist species is greater at in water-saturated tropical soils. Finally, San Bernardino than at Carlinville, and greater Michener (1979) noted that bee communities in xeric desert scrub habitat than in mesic ri- in xeric areas are dominated by short-lived parian habitat at San Bernardino. 178 R. Minckley 2. MATERIALS AND METHODS a few plant species per day and kept a detailed voucher collection. Plants that bloomed only briefly 2.1. Study sites and bee faunal data were transplanted to his residence and observed, so insect records include visits to plants in native and 2.1.1. Carlinville, Illinois, USA cultivated habitats. Charles Robertson collected floral visitors within a 16 km radius of Carlinville, Illinois be- 2.1.3. San Bernardino Valley, Sonora tween 1884 and 1916 and self-published his obser- vations (Robertson, 1929). Based on climate data Mexico and Arizona, USA taken at Carlinville between 1971 and 2000, the area is mesic temperate. Average total precipitation/ The San Bernardino Valley runs north-south at year is 98 cm. and when bees are active (April– the Mexico-United States border in northeastern October) rain averages nearly 10 cm/month, with Sonora, Mexico and southeastern Arizona, USA.