Cheadle Center for Biodiversity and Ecological Restoration

Modeling insect heat exchange and predicting thermal responses depends on accurate representationof body size and shape. Still, most biophysical models approximate these complex forms usingsimplified geometric solids, whose relationships to real body forms have not been rigorously tested.Advances in surface modeling of small objects allow us to interrogate these assumptions by capturingthe real 3D complexity of insect body forms. We used photogrammetry to construct 3D models ofhoney bee specimens and empirically measured body volume and surface area. Compared to empiricalmeasurements, we found that traditional, geometric size estimation methods systematicallyunderestimate body surface area and volume. We incorporated these error estimates into publishedheat budget data and found that these errors propagated non-linearly through the model, shiftingthe relative dominance of convective and radiative heat loss as temperature increases. These resultssuggest that body size and surface area assumptions can distort modeled heat transfer, particularlyunder low temperatures, demonstrating that morphological simplifications can bias physiologicalinference. This work underscores the utility of empirical 3D morphology for refining biophysical modelsof insect thermoregulation.

Wild bees are widely believed to be in decline, yet most species remain unassessed for IUCN extinction risk. Geographic range size is used in risk assessments under the assumption that species with smaller ranges are more vulnerable to anthropogenic impacts. Narrow diet breadth can also increase vulnerability but is not currently incorporated into assessments.Niche breadth theory predicts a positive association between range size and diet breadth, which could concentrate risk among dietary specialists, but this relationship is not well established for bees and may differ among taxa.Here, we combined pollen-use data from natural history collections with global occurrence records to test the relationship between diet breadth and range size across bees and among bee families. We assigned diet breadth using three metrics (categorical, numerical and phylogenetic) and estimated range size as the extent of occurrence for 633 species from six families.Across bees, range size increased with diet breadth, and diet specialists tended to occupy smaller ranges. These results suggest that range size and diet breadth jointly contribute to increased vulnerability in bees, indicating that some specialist species may merit conservation prioritisation.Our findings support the integration of trait-based approaches for assessing extinction risk and highlight the high value of natural history data in identifying patterns of vulnerability among pollinators.

Dynamic coloration is one of the most striking visual displays in the animal kingdom. While reversible colour changes are well characterized in animal communication, more passive effects of climate on baseline coloration remain poorly understood. Here, we present a novel experimental demonstration of reversible, humidity-induced colour change in bees. In controlled lab experiments, we show that relative humidity affects cuticle colour of the sweat bee Agapostemon subtilior, changing dramatically within 24 h from a deep blue-green at low humidity to a pale, coppery green at high humidity. Older specimens experienced greater magnitude colour shifts, suggesting that cuticular degradation may increase water permeability and amplify moisture effects. To understand whether these effects shape colour variation in the wild, we extracted colour data from a large dataset of crowd-sourced field images. We found that ambient humidity weakly predicts colour variation across A. subtilior’s western range, in a manner consistent with the direction of colour change established in lab experiments. While the structural basis for this colour change is still unknown, these shifts are directionally consistent with moisture-induced swelling of multilayer structures that causes reflection of longer wavelengths, a mechanism described in other insects and cephalopods. Together, these results demonstrate that climate modulates structural coloration in bees, emphasizing the role of abiotic conditions in shaping dynamic visual traits.

Global declines in bee populations are threatening the ecosystem services they provide, including pollination. Many bee–plant interactions are understudied, producing an incomplete understanding of resulting ecosystem-level vulnerabilities. The last decade has generated a wealth of opportunistic data originating from natural history collection records, published ecological datasets, and citizen/community science initiatives in online databases such as Global Biotic Interactions (GloBI). Here, we explore hypotheses related to bee–plant interactions and detection processes using the GloBI database, curated checklists of bee and flowering plant species, and an occupancy model. We hypothesized that larger, social bees would visit a larger number of plant species, while smaller, solitary bees would visit fewer. We also predicted that flowers with open, bowl-like shapes would attract a greater diversity of bee visitors compared to closed shapes. Further, we hypothesized that both floral and bee traits, such as bright colors and conspicuous patterns, would increase detectability, and that different data collection methods would vary in their ability to capture bee–plant interactions. Lastly, we hypothesized that the interaction network generated by the output of the occupancy model, which accounted for imperfect bee–plant detection, would yield more interactions, thereby increasing measures of evenness and decreasing nestedness and specialization, as compared to the network generated from recorded interaction data. We found that smaller bees exhibited higher probabilities of plant interactions than larger bees, but we did not find evidence that bee sociality influenced the probability of interacting with plants. We found that blue flowers and closed (not-bowl-shaped) flowers had higher probabilities of bee-plant interaction than other flower colors or bowl-shaped flowers, respectively. We also found that larger bee size, blue flowers, bowl shapes, and community science sources were associated with higher detection probabilities of bee–plant interactions. Lastly, the interaction network generated by the occupancy model output showed higher levels of evenness, nestedness, and connectance than the network generated by the GloBI data. Our study is among the first to utilize occupancy modeling to directly model species' interactions, leverage aggregated, open-source databases and expert checklists, and highlight the influence of detection and collection biases on our understanding of ecological interactions.

Bees are one of the most important pollinators in terrestrial ecosystems, supporting biodiversity and food production. However, global knowledge of their interactions with host plants remains limited. To address this, we describe and refine a subset of the Global Biotic Interactions (GloBI) database focused on bee-plant interactions. We updated taxonomy using current checklists and enhanced the dataset with metadata on geography, endemism, and human uses of plants. The resulting dataset includes 981,982 unique interaction records between 5,537 bee species and 12,699 plant taxa. Despite its scale, the dataset is affected by strong taxonomic and geographic biases. It covers only 26% of described bee species and 4% of flowering plant taxa—primarily those used by humans—and is heavily skewed toward North America and Western Europe. Nevertheless, GloBI represents a valuable resource for incorporating bee-plant interactions into biodiversity and conservation-oriented research and represents a considerable advance in our current knowledge.

Conservation and management of wild bees is hindered by the variety of ways wild bee occurrence data are recorded, managed, and shared. Here, we present solutions to address this issue and introduce The Wild Bee Data Standard, a standardized means of recording and reporting data associated with wild bee occurrences, including physical specimens and photo observations. This standard aligns with contemporary data management practices widely adopted by the broader biodiversity data community. We propose a set of terms for the standard that describe various features of bee occurrences, including collection method and location, taxonomic verification, and final record storage. We emphasize the importance of providing sampling protocol and effort information with wild bee occurrence data and offer guidance to make this a more common practice. We describe how to translate data not currently aligned with the standard to meet its conditions, and how to upload those data to an accessible online repository. We provide case studies, data entry templates, a glossary of terms, and additional resources to guide new users to implementing the standard. We also present a forum, established as a GitHub repository, to support continued development of the standard. Recognizing the significant change this represents for current data practices, we outline the benefits for the bee research and conservation community that will result from improved data standards. We advocate for making all historical, current, and future bee occurrence data openly available to facilitate more rigorous and comprehensive research, conservation, and management of wild bees. This contribution is part of a series developed in association with the U.S. National Native Bee Monitoring Network to standardize bee monitoring practices.

All bees depend on angiosperms for survival, while many angiosperms depend on bees for reproduction. However, bee and flowering plant species richness do not peak in the same geographical regions of the world, suggesting that the flora in regions where bees are not as diverse, such as the tropics, may be relatively less bee-dependent. We test this assumption by analysing whether local relative bee diversity can predict the proportion of angiosperm species that attract bees (i.e., “bee flowers”).

Community or volunteer participation in research has the potential to significantly help mobilize the wealth of biodiversity and functional ecological data housed in natural history collections. Many such projects recruit community scientists to transcribe specimen label data from images; a next step is to task community scientists with conducting straightforward morphological measurements (e.g., body size) from specimen images. We investigated whether community science could be an effective approach to generating significant body size datasets from specimen images generated by museum digitization initiatives. Using the community science platform Notes from Nature, we engaged community scientists in a specimen measurement task to estimate body size (i.e., intertegular distance) from images of bee specimens. Community scientists showed high engagement and completion of this task, with each user measuring 43.6 specimens on average and self-reporting successful measurement of 98.0% of the images. Community scientist measurements were significantly larger than measurements conducted by trained researchers, though the average measurement error was only 2.3%. These results suggest that community science participation could be an effective approach for bee body size measurement, for descriptive studies or for research questions where this degree of expected error is deemed acceptable. For larger-bodied organisms (e.g., vertebrates), where modest measurement errors represent a smaller proportion of body size, community science approaches may be particularly effective. Methods we present here may serve as a blueprint for future projects aimed at engaging the public in biodiversity and collections-based research efforts.

Based on findings of revisionary studies of Neotropical  Rinorea , we propose several taxonomic novelties and clarifications in this study, including the description of three new species, a new combination, the placement of three species in synonymy, and updated descriptions of two poorly known species from South America. One of the new species,  Rinorea idarragae , is described from Mesoamerica and belongs to  R. sect. Rinorea . The two other new species are from South America and are placed in  R. sect. Pubiflorae :  Rinorea cardenasii  occurs in the region of the western edge of the Guiana Shield in Colombia, and  Rinorea pabongonzaleziorum  is from the Amazon Basin in Colombia and Venezuela.  Rinorea lindeniana var. fernandeziana  is elevated to the rank of species, and  Rinorea cordata ,  R. antioquiensis , and  R. hymenosepala  are synonymized under  R. laurifolia ,  R. squamata , and  R. ulmifolia , respectively. Finally, we provide updated descriptions of  Rinorea deflexa  from Ecuador and Peru and  Rinorea bicornuta  from Colombia and Brazil. For the three new species, we provide descriptions, illustrations, distribution maps, and preliminary conservation assessments based on IUCN Red List categories and criteria.

Insects are declining in abundance and species richness, globally. This has broad implications for the ecology of our planet, many of which we are only beginning to understand. Comprehensive, large-scale efforts are urgently needed to quantify and mitigate insect biodiversity loss. Because there is broad interest in this topic from a range of scientists, policymakers, and the general public, we posit that such endeavors will be most effective with precise and standardized terms. The Entomological Society of America is the world’s largest association of professional entomologists and is ideally positioned to lead the way on this front. We provide here a glossary of definitions for biodiversity loss terminology. This can be used to enhance and clarify communication among entomologists and others with an interest in addressing the multiple overlapping research, policy, and outreach challenges surrounding this urgent issue.