A homeotic shift late in development drives mimetic color variation in a bumble bee.

Natural phenotypic radiations, with their high diversity and convergence, are well-suited for informing how genomic changes translate to natural phenotypic variation. New genomic tools enable discovery in such traditionally nonmodel systems. Here, we characterize the genomic basis of color pattern variation in bumble bees (Hymenoptera, Apidae, Bombus), a group that has undergone extensive convergence of setal color patterns as a result of Müllerian mimicry. In western North America, multiple species converge on local mimicry patterns through parallel shifts of midabdominal segments from red to black. Using genome-wide association, we establish that a cis-regulatory locus between the abdominal fate-determining Hox genes, abd-A and Abd-B, controls the red-black color switch in a western species, Bombus melanopygus Gene expression analysis reveals distinct shifts in Abd-B aligned with the duration of setal pigmentation at the pupal-adult transition. This results in atypical anterior Abd-B expression, a late developmental homeotic shift. Changing expression of Hox genes can have widespread effects, given their important role across segmental phenotypes; however, the late timing reduces this pleiotropy, making Hox genes suitable targets. Analysis of this locus across mimics and relatives reveals that other species follow independent genetic routes to obtain the same phenotypes.

The role of toxic nectar secondary compounds in driving differential bumble bee preferences for milkweed flowers.

While morphological differences such as tongue length are often featured as drivers of pollinator floral preferences, differences in chemical detection and tolerance to secondary compounds may also play a role. We sought to better understand the role of secondary compounds in floral preference by examining visitation of milkweed flowers, which can contain toxic cardenolides in their nectar, by bumble bees (Bombus spp.), some of their most abundant and important pollinators. We examine bumble bee species visitation of common milkweed (Asclepias syriaca) compared to other flowers in the field and test whether observed preferences may be influenced by avoidance and tolerance of cardenolides, as measured by the cardenolide ouabain, in the lab. We reveal that common milkweed is visited predominantly by one bumble bee species, Bombus griseocollis, in a ratio much higher than the abundance of this species in the community. We confirmed the presence and toxicity of cardenolides in A. syriaca nectar. Lab experiments revealed that B. griseocollis, compared to the common bumble bees B. impatiens and B. bimaculatus, exhibit greater avoidance of cardenolides, but only at levels that start to induce illness, whereas the other species exhibit either no or reduced avoidance of cardenolides, resulting in illness and mortality in these bees. Toxicity experiments reveal that B. griseocollis also has a substantially higher tolerance for cardenolides than B. impatiens. Together, these results support a potential evolutionary association between B. griseocollis and milkweed that may involve increased ability to both detect and tolerate milkweed cardenolides.

Unsupervised machine learning reveals mimicry complexes in bumblebees occur along a perceptual continuum.

Müllerian mimicry theory states that frequency-dependent selection should favour geographical convergence of harmful species onto a shared colour pattern. As such, mimetic patterns are commonly circumscribed into discrete mimicry complexes, each containing a predominant phenotype. Outside a few examples in butterflies, the location of transition zones between mimicry complexes and the factors driving mimicry zones has rarely been examined. To infer the patterns and processes of Müllerian mimicry, we integrate large-scale data on the geographical distribution of colour patterns of social bumblebees across the contiguous United States and use these to quantify colour pattern mimicry using an innovative, unsupervised machine-learning approach based on computer vision. Our data suggest that bumblebees exhibit geographically clustered, but sometimes imperfect colour patterns, and that mimicry patterns gradually transition spatially rather than exhibit discrete boundaries. Additionally, examination of colour pattern transition zones of three comimicking, polymorphic species, where active selection is driving phenotype frequencies, revealed that their transition zones differ in location within a broad region of poor mimicry. Potential factors influencing mimicry transition zone dynamics are discussed.