4.2 | Historical range contraction and expansion
Analyses of the Volga River only dataset revealed a spatial distribution of genetic diversity that does not fit expectations under the stream hierarchy model. Local populations exhibited low diversity (Table 1B) and lack any discernable pattern of population substructure that reflects the distribution (Figures 3B, 4B). In contrast to the combined river dataset, the Volga only dataset revealed a greater number of genetic clusters with the most upstream (VO1) and downstream (VO5) localities representing unique clusters with no admixed individuals (Figure 3B). However, there was no evidence of isolation by distance (Figure 5A). The FST values (Table S3A) indicated locality VO1 was the most genetically distinct; it contained an unusually high number of private alleles (Table 1B) as compared to the other localities. In addition, localities VO2 and VO3 are more similar to the most downstream locality, VO5, than the intervening locality VO4. The greatest sources of variation are differences among individuals within populations (Table 2B).
There are two possible explanations for the genetic signature observed in the Volga River: fragmentation of local populations due to isolation or historical range contractions and expansion. It is possible that the local populations are fragmented. The Volga River flows through a highly impacted landscape that has been dramatically altered by intensive agricultural activities and urbanization. Rivers and streams have been heavily impacted, causing extensive modification to habitat structure, water quality, and flow regime (Menzel, 1983; Knox, 2007). However, the spatial pattern of genetic diversity is random and does not fit expectations for fragmentation and isolation. In addition, the pattern does not make sense given the life history of the fish and the distribution of localities within the system.
The best explanation is this pattern results from the influence of historical range contractions and expansions during periodic glacial cycles. There is low overall genetic diversity in the Volga River only dataset, which is reflected in the lower number of variant loci and private alleles as compared to the population in the South (Table 1B). In addition, the observed heterozygosities are relatively high with low FIS values, which are consistent across sampled localities within the system. These observations suggest that the population in the Volga River is not in equilibrium; the observed genetic signature is not reflective of the contemporary demographic and evolutionary processes (Whitlock & McCauley,1990; Epps & Keyghobodi, 2015). Rather, the pattern observed is a function of the founding events that established populations in the North. The bottleneck caused by the founder effect reduces variation, but can have little effect on the observed heterozygosity (Allendorf & Luikart, 2007). This is especially true when a population returns to a large size relatively quickly, which is consistent with the life history of E. caeruleum . Even though populations can be isolated, the rainbow darter tends to be very abundant and one the most common fish in a stream (Page, 2000).
The disconnect between the contemporary landscape and the observed genetic pattern is the result of a time-lag (Epps & Keyghobadi, 2015). Even though it has been 10,500 years since the retreat of the last glacial maximum, a genetic signature of expansion persists. Following a major population disturbance, it takes time for genetic variation to reach equilibrium, which is determined by a number of factors including generation time, dispersal rates, effective population size and population growth rates (Epps & Keyghobadi, 2015). Based on the complex interactions of these factors, it can take tens of thousands of generations for a population to reach equilibrium (Varvio et al., 1986; Zellmer & Knowles, 2009). This conclusion is further supported by the results of the analyses of the Meramec River only dataset; see discussion below.
The interpretation of the observed genetic signature in the Volga River only dataset is consistent with other landscape genetic studies ofE. caeruleum . A recent study utilized microsatellite data to understand the spatial distribution of genetic diversity of the rainbow darter in tributaries of the upper Mississippi River basin in northeast Iowa (Davis et al., 2015). The analyses revealed a single genetic population and no evidence of significant population subdivision despite the large geographic distance separating local populations. The population of rainbow darters in this region were genetically diverse, but the diversity was evenly distributed across the landscape suggesting extensive gene flow and connectivity among population across the landscape. After considering the life history and habitat requirements of the rainbow darter and the glacial history of the region, they concluded that the best explanation was historical events overwhelmed the observed genetic signature and the data were unable to detect the influence of contemporary processes (Davis et al., 2015). Although this study came to a similar conclusion, the use of microsatellite DNA markers presumably limited the resolution of the genetic data. The genome-wide SNP data utilized in this study were able to reveal fine-scale population substructure that was not observed in Davis et al. (2015). Haponski et al. (2009) observed similar patterns of genetic variation in E. caeruleum populations in glaciated regions of the Lake Erie and the Ohio River basins east of the Mississippi River.