Discussion
The redistribution of species globally has ignited interest in and urgency for understanding eco-evolutionary dynamics of range shifts. Foundational theoretical work has provided a basis for understanding the interplay between stochastic and deterministic forces in facilitating or hindering range expansion. In many cases, range expansions are expected to lead to higher divergence and lower genetic diversity at the expanding edge compared to the core due to small population sizes, serial bottlenecks, reduced gene flow, and selection pressures . This mixture of neutral and adaptive processes across the expansion axis often leads to spatial structuring and patterns of isolation by distance . Termed a “pulled wave”, the founders at the range edge pull the expansion forward through increased dispersal and reproduction that stratifies demes . However, the opposite pattern of maintained/increased genetic variance at expansion fronts has also been reported in several empirical studies . Conceptual frameworks term these cases “pushed waves,” where genetic variation is maintained at the range edge due to gene flow from the range core and potentially positive density dependence, novel interspecific competition or environmental stress that result in less genetic sorting at the edge . These varying outcomes suggest that the interplay between neutral and selective evolutionary processes create variation in range expansion outcomes giving rise to more nuanced approaches to eco-evolutionary dynamics. . In the well-documented contemporary range expansion in C. anna , we show patterns largely consistent with expectations of a “pushed wave” expansion; gene flow is high throughout the entire range and we find no strong divergence in allele frequencies between the core of the range and expansion fronts. We do find reduced genetic diversity at expansion fronts, which are characteristic of pulled waves, but the magnitude of the reduction in genetic diversity is small. Together our evidence highlights both the complexity of rapid range shifts in natural populations and potential limitations of genomic data in investigating eco-evolutionary phenomena.
Evidenced by low species-wide genetic divergence and a lack of spatial structuring or signals of selection at the expansion edges or species-wide, we show that C. anna has few, if any, limits to gene flow. These results support a previous genetic study in C. anna that found no genetic structure among three California populations using mitochondrial DNA . The preservation of genetic diversity across the expanded ranges is consistent with recent, rapid range expansions characterized by a short time frame, growing population sizes, and multiple independent expansion fronts. Our results align with the well-documented rapid range expansion in C. anna over the past 80 years . While movement details remain enigmatic, C. anna has a broad diet, relatively large territories, and some seasonal migration that was documented based on abundance data , all of which could contribute to high gene flow in this system. Regardless of the mechanism, we observe no population structure and little evidence for increased differentiation at the expansion fronts. Long-distance dispersal, especially from the core, has been shown to preserve genetic diversity in other taxa . This result is often seen in highly mobile species and recent invasions. Examples of high gene flow within species in newly colonized territories include the following: invasive Indo-Pacific lionfish (Pterois volitans ) in the Caribbean , European starlings (Sturnus vulgaris ) in South Africa and North America , Pyrenean rocket (Sisymbriumaustriacum ) in the Meuse River Basin , and round gobies (Neogobius melanostomus ) in the Great Lake tributaries . The similarities with colonizing species expansions (e.g., propagule and dispersal pressure, novel biotic and abiotic interactions) underscore the emerging work viewing range shifts and expansions of native species, especially those caused by climate change, through the lens of invasion biology . By incorporating theory from invasion biology including assessing the potential impacts, positive and negative, of colonizing species on novel environments, we can gain a more holistic understanding of range expansion.
Range expansions often expose species to novel environments containing new combinations of biotic and abiotic interactions that can coincide with niche shifts, expansions, or unfilling . Previous modeling showed that the expanded regions fell within C. anna’s fundamental climatic niche prior to the range expansion, suggesting that previous range limits were defined by the presence of resources . Although it is possible that selection was overlooked due to the limitations of genome scans (see below), the lack of selective signatures between the core and expanded ranges identified here aligns with the previous hypothesis that the range expansion in C. anna could be the result of an ecological release facilitated by human-mediated landscapes. This hypothesis states that introduced plants and supplemental feeding have allowed C. anna to fill out its existing climate niche even in the expanded regions. While other ecological factors induced by urbanization and climate change could also be aiding the expansion, a similar pattern of spatial expansion and ecological release associated with supplemental feeding has been documented in Eurasian Blackcap warbler . Together, these studies provide evidence for the role of local anthropogenic alterations of the landscape shaping broadscale shifts in species’ ranges.
High gene flow can be maladaptive at the expansion edge and inhibited selection during pushed wave expansions often slows and/or prevents further range expansion , such as the southeastern invasion edge of cane toads in Australia that is thought to be limited by cold temperatures . Despite this paradigm, C. anna expansions do not appear to be hindered by gene flow from the range core or lack of selection at the expanding edge, adding to the growing literature that gene flow does not limit species ranges or their range expansions (Kottler, Dickman, Sexton, Emery, & Franks, 2021). This again supports the hypothesis that the C. anna range was potentially limited by resource availability and thus adaptation may not be required for expansion. Recently, a growing breeding population of C. anna has been found in Idaho , an area predicted to be suitable for C. anna . Southeast Washington state and sections of Utah are also predicted to have suitable habitat for C. anna , potentially suggesting that the expansion will continue in the coming decades.
Despite high gene flow and lack of genomic signatures of selection, we do find very subtle evidence of classic core-edge patterns of genetic diversity. While we did not detect structure in the PCA or admixture analysis, we observed lower nucleotide diversity at both expansion fronts. This result could indicate that mating is not completely random across the entire range despite high gene flow . Increased relatedness among individuals due to small populations sizes could drive the decrease, however this does not appear to be the cause in C. anna . Alternatively, lower genetic diversity could be a result of decreased observed individual heterozygosity, which is what we observed. The loss of heterozygosity at the expansion front could be caused by genetic drift, specifically in response to population bottlenecks or allele surfing or by selective sweeps in the expanded ranges, although we were not able to detect any with our genome scans. Relatedly, many of the northern expansion front samples were collected in earlier years (2000 vs late 2010s) which may represent a time point closer to a founding bottleneck before more birds dispersed from the core, a pattern previously suggested in the invasive Indo-Pacific lionfish . However, this decreased heterozygosity does not appear to be strong and consistent enough to result in signatures of divergence and selection between the range core and expansion fronts.
The seemingly contradicting observations of decreased genetic diversity in the absence of signals of selection or structure could have several biological or technical explanations. One possibility is that the expansion is too recent for the detection of significant divergence in range edge populations and more differentiation may develop over time. Further, our low-coverage approach and moderate sample size may not have the power to detect multiple small shifts in allele frequencies across loci that could lead to adaptive evolution, a common issue with genome-wide scans for polygenic traits (Kemper, Saxton, Bolormaa, Hayes, & Goddard, 2014; Pritchard, Pickrell, & Coop, 2010). Reduced diversity in the expanded regions could therefore reflect this weak genome-wide selection that was not detected at any single SNP. Alternatively, environment-mediated trait differences may be plastic - perhaps looking at gene expression or plasticity would provide a more timeframe-appropriate picture of how these birds are responding to the novel environments at the expansion edges. For example, there is widespread use of torpor in Trochilidae and C. anna is no exception. In fact, C. anna were found to increase their use of torpor in cold temperatures . Plastic behavioral changes at the expansion edges, such as increased use of torpor, could account for how these birds are surviving colder northern temperatures without seeing genetic changes, and warrant further study. Finally, our samples represent a single snapshot in time thus cannot rule out that selection is happening at the range edges, but not enough time has passed to shift allele frequencies that would be detected in the current study, especially in the face of high gene flow.
Anthropogenic influences are changing the genetic landscape through shifting species ranges. Much of the recent focus has been on the role of climate change in facilitating range shifts and the likely eco-evolutionary dynamics of these phenomena. However, this study demonstrates that not all expanding species respond in predicted ways, in fact, not all human-induced range expansions show obvious signatures of evolution. Further work is needed to confirm these results and test the stability of our conclusions over time. For example, using museum specimens to understand the genetic landscape in C. anna before the expansions could confirm past gene flow or illuminate if increased urbanization is decreasing genetic diversity, increasing homogenization, or favoring certain alleles. Additionally, while we focused on the northern and eastern expansions, sampling individuals from what might be the “trailing edge” in Mexico would further our understanding of whether climate or resources are defining species range limits inC. anna . This study contributes to the growing literature on the consequences of human-mediated range expansions by adding empirical evidence that eco-evolutionary dynamics are not one-size fits all.