Evolution of mimetic coloration in bumble bees

       The ~270 species of the cold-adapted bumble bees are exceptionally color diverse. One of the main factors driving such variation is Müllerian mimicry – taxonomically diverse species in the same geographic region have converged on similar color patterns as selection favors shared advertisement of their toxicity (their sting) to predators. Bumble bee color diversity provides extensive replicates for understanding the genomic targets of evolution (evolutionary genetics) (Hines and Rahman, 2019), how ecological factors drive adaptive phenotypes (evolutionary ecology), and how developmental gene modifications create phenotypic variation (evo-devo).

Building from foundational work on the development of coloration of bumble bees (Tian and Hines, 2018, Hines et al., 2017, Hines et al., 2022) and bumble bee genomics (Lozier et al., 2026), we are studying how changes in genomes and transcriptomes lead to diverse mimetic phenotypes in these bees.  We have discovered key segmentation genes (Hox genes) switch segmental location late in development to cause evolutionary shifts in the location of red segmental stripes in western North American bumble bees (Tian et al., 2019). Using genomics and transcriptomics we have found these genes are repeatedly employed with the convergent evolution of mimics and that they act by triggering a downstream gene network that ultimately targets genes in the melanin pathway (Hines et al., 2025, Rahman et al., 2021). We also have characterized mimicry complexes in the United States and examined the factors promoting their evolution (Ezray et al., 2019). 

Recently we have studied the genetics of white mimetic color form divergence in bumble bees from Anatolia, finding an upstream regulator BarH, which also drives white-yellow wing colors in butterflies (!), is responsible for this coloration (Dabak et al., 2026). We studied a bumble bee gynandromorph with a white and yellow mosaic to better understand how this gene operates. Our studies in these bees, as well as ongoing work on Bombus flavifrons, both reveal the role of regulatory region duplication in creating new phenotypes. We are currently exploring the role of the Hox gene hotspots using RNAi and CRISPR approaches in bumble bees.

Photo credits: J. Cnaani (upper right), Sam Droege, USGS Survey (upper middle)

Bombus niveatus gynandromorph which is a mosaic of colors and sexes, showing its white partly female left side, its yellow mostly male right side and a shot of its bilaterally split face.

Bumble bee conservation

Bumble bee diversity has declined in the last 30 years due to a combination of habitat loss, pesticides, pathogens, and climate change. We are studying several of the major factors threatening these bees. We are currently studying:

• NUTRITION – Importance of pollen nutrition for bumble bees. We are assaying how protein content of multifloral pollen impacts preference, development, and foraging patterns across multiple bumble bee species as part of a USDA NIFA grant. We are finding that high protein pollen is especially important for these bees and that bumble bee species vary in nutritional flexibility, which may impact their success in the face of land modification.

(Left) Ethan Dean weighing bees and pollen for nutrition research. (Middle) Reared Bombus impatiens colony. (Right) Reared Bombus penyslvanicus colony.

• CLIMATE – Thermal tolerance in bumble bees and the biological and environmental factors that impact it. We recently have assessed how biological factors, including species, caste, age, and reproductive state, impact bumble bee thermal tolerance (Feuerborn et al., 2023). We find all are impacted and that species that occupy different local habitats as well as different latitudinal ranges differ in heat tolerance. Our prior and current work has examined the impact of diet on thermal resilience in bumble bees, where we show that insufficient feeding (Quinlan et al., 2023) and protein diets will impact thermal tolerance. We also found the toxic secondary compounds in milkweed nectar can impact thermal tolerance (Shippee et al., 2025). We are also examining the impacts of thermal reprieves on heat outcomes in bees (e.g., Quinlan et al., 2023).

(Left) Cody Feuerborn collecting bees in the field. (Middle) Differences in thermal tolerance by bumble bee species in the Eastern U.S. from Feuerborn et al. 2023. (Right) Summary of results from Shippee et al., 2025 showing that milkweed nectar lowers thermal tolerance in bumble bees.

• PATHOGENS – The environmental factors driving pathogen loads in wild bees. Using landscape modeling, we have examined the landscape factors, such as habitat, climate, and disturbance, that most impact wild bee pathogen loads, showing that forests and open valleys differ in their pathogen loads and that development, heat, inadequate nutrition, and honey bee incidence all relate to increased viral loads in bumble bees (McNeil et al., 2020, Gratton et al., 2023, 2024, Wham et al., 2024)). We are now examining how pathogens fluctuate across bee communities using -omics based sequencing approaches to document pathogen communities across bee species by habitat. See some of our recommendations for land management for bumble bees here. 

(Left) Elena Gratton collecting bees in the field. (Middle) We performed studies in three different regions assessing impact of landscape on pathogens in bumble bees. (Right) The environmental factors we assessed in the landscape against pathogen loads.

• PREDICTING STRESSORS – Landscape transcriptomics towards deciphering stress in bumble bees. We are applying transcriptome response to understand the leading stressors in bumble bees (Keagy et al., 2023, Quinlan et al., 2024). We have shown transcriptomes respond in a predictable manner to different stressors and that we can detect these stressors in bees collected in the field. We are currently developing this further by understanding reaction norms, stressor interactions, and impacts of diet on other stressors, using transcriptome data.

From Quinlan et al., 2023 – (Left) Bumble bees show distinct clustering by stressor. (Middle) The transcriptomic response by stressor is predictable. (Right) Bees in the wild during a heat wave show elevated heat stress in their transcriptome, showing transcriptomics can inform the stressors encountered by bees.

• DOCUMENTING DECLINES – Documenting factors contributing to the decline of Bombus pensylvanicus in the NE U.S. Bombus pensylvanicus has seen precipitous declines particular in the northeast portion of its range. We are documenting geographic patterns of this declines and comparing genome of current and historic specimens of the species to understand allelic diversity and selection related to these declines.

• LAND MANAGEMENT – Documenting bumble bee foraging patterns towards optimal land management for bumble bees. We have worked with Ask a Bumble Bee in mass collecting visitation data for bumble bees and are now assessing best plants for bumble bees considering their nutritional needs. We are using these data as well as data on native plants of the northeastern U.S. to design optimal bumble bee plantings for the region. 

Mechanisms of gall induction by gall wasps

Gall wasps oviposit into developing plant tissues where larvae induce the formation of elaborate and predictable gall morphologies. The ways in which both adults and developing larvae induce this developmental change remain largely unknown. We are combining comparative phylogenomics, comparative gland morphology (Guiguet et al., 2023a), developmental data, and transcriptomic and metabolomic approaches to explore potential mechanisms of gall induction. These analyses are comparing across species, sexual/asexual generations, and larval/adult generations of gall wasps. We have also been contributing natural history data (acid galls: Guiguet et al., 2023b, cryptic galls: Davis et al., 2026) and describing the biodiversity and revising the taxonomy of these wasps (Nastasi et al., 2024, 2025; Montelongo et al., 2026).

Just a sampling of some of the incredible diversity of gall morphologies induced by gall wasp species. From Left: Wooly catkin gall wasp (Callyrhytis quercusoperator) including both asexual and sexual galling generations, Hedgehog gall (Acraspis erinacei), Honeycomb leaf gall (Callyrhytis favosa), Clustered midrib gall (Andricus dimorphus), Wooly oak gall (Callyrhytis lanata). Images by collaborator Andy Deans, available on iNaturalist.

Bumble bee systematics, evolution, and natural history

We are examining phylogenetics, population genetics/species delimitation (Ghisbain et al., 2020, Lhomme et al., 2021, Wham et al., 2021), biogeography, and the evolution of life history traits, such as tongue length, nest biology (Hines 2025), and setal diversity (Hines et al., 2022), in the bumble bees. The lab rears multiple bumble bee species for which we are making observations on their natural history and biology. We currently are completing a phylogeny of world bumble bees using phylogenomic data and using this to examine the correlated evolution of traits and reassessing biogeography and divergence times.