Genomic evidence for long-term ecological interdependence of a plant-hummingbird mutualism

For over 100 years, biologists have been fascinated with coevolution between mutualistic partners such as plants and their pollinators.^1,2,3,4,5 Due to the dissipating effects of species being entangled into networks of interacting species,^6,7,8 however, demonstrating reciprocal selective forces can be difficult except for strongly interdependent pairs of species.^9,10 Furthermore, temporal fluctuations in environmental disturbance or simple ecological shifts may result in partner swapping,¹¹ so that seemingly matching and coadapted species pairs may in fact have interacted only over short time frames. Here, we adopted a genomic approach^12,13 to scrutinize the history of the highly sex-specific and specialized mutualistic association between the purple-throated carib (Eulampis jugularis) hummingbird and its two Heliconia nectar-food plants on Dominica.^{14,15,16,17,18,19} Notably, we investigate whether long-term coadaptation is reflected in aligned genomic signatures as linked plant and hummingbird population trajectories over evolutionary time frames. As hypothesized, our analyses indicate synchronized population size fluctuations across the three mutualistic partners, which is consistent with long-term ecological interdependence and coadaptation over millions of years. Additionally, we find evidence for the hypothesis that genetic diversity and level of inbreeding of the two plant species mirror their ecological histories with hummingbird partners: territorial purple-throated carib males promote inbreeding and low genetic diversity in H. caribaea, whereas wide-ranging, traplining females pollinate H. bihai, resulting in less inbred, more genetically diverse plants. This genomic evidence demonstrates that specialized mutualisms can leave aligned evolutionary signatures, thus highlighting the power of integrating comparative genomics with ecological context to assess the long-term persistence and consequences of species interactions across evolutionary timescales.