Divergence, admixture and speciation
The role of climate fluctuations in causing secondary contact of recently diverged species is well known (Hewitt 1999), with the outcomes varying from free admixture, to directional gene flow, to complete isolation or even hybrid speciation. Several cases demonstrating the range of these outcomes are known from alpine taxa within the Sierra Nevada Mountains (Gompert, Fordyce, Forister, Shapiro, & Nice, 2006; Rovito 2010; Rubidge, Patton, & Moritz, 2014). The degree and geographical scale of admixture likely varies across taxa due to a number of factors, including the dispersal ability of species, habitat suitability and connectivity, as well as the reproductive compatibility of lineages and fitness of hybrids (Barton & Hewitt 1985; Dufresnes, Berroneau, Dubey, Litvinchuk, & Perrin, 2020).
In this study, we found that an intermediate morphotype in the N. ingens complex (Schoville et al. 2012) extends to populations throughout the Central Sierra Nevada, encompassing the entire San Joaquin drainage and extending south to contact zones in the Kings River watershed (sites 14 and 15; Figure 1 ). The distribution of the morphological phenotype is different from the observed genetic patterns, where mitochondrial haplotype variation suggests introgression fromN. riversi into nearly the entire southern range of N. ingens (Figure 2 & S2 ), and genome-wide nuclear markers suggest that the intermediate morphotype is an independent lineage (site 9, 10 and 12; Figure 3 ). For example, in the sNMF analysis (from K=3 to K=11), most individuals representing the intermediate morphotype do not show extensive ancestry from either N. riversior N. ingens (Figure S1 ). Additionally, TMRCA estimates between populations of the intermediate morphotype and either species mostly predate the last glacial maximum. Despite this genetic divergence, we found no evidence of premating isolation among the three lineages (Table 3 ), although our sample sizes were low.
Based on these results, we infer that 1) an early glacial cycle caused deep divergence to form between northern and southern populations, which formed N. ingens and N. riversi . In a subsequent glacial cycle, 2) a contact zone formed near the San Joaquin drainage and led to the formation of the intermediate morphotype. This lineage diversity was maintained during the last glacial maximim and 3) allowed for the formation of more complex population structure as each major lineage recolonized high elevation habitat. It is clear that limited gene flow occurs in parts of the alpine range, leading to the formation of a genetic cline in genetic divergence measures such asF ST. The general pattern we observe is consistent with models of diversification linked to the orbitally forced range dynamic (ORD) model (Dynesius & Jansson, 2000), and the more recently coined ’mix-isolation-mix’ model proposed by He et al. (2019).