The influence of environment and gene flow on fungal endophyte frequencies and vertical transmission rates in two native host grass species Elymus virginicus and E. canadensis

Background/Question/Methods

Vertically transmitted (VT) symbioses, passed from parent to offspring, are common and influential in nature. A widespread symbiosis occurs between cool-season grasses (sub-family Pooideae) and endophytic fungi (genera Neotyphodium and Epichloë), many of which are exclusively or predominantly transmitted vertically to seeds. VT couples host and symbiont fitness and is therefore hypothesized to select for tightly coevolved mutualisms, where benefits to both host and symbiont are maximized. Therefore, it is predicted that the frequency of VT mutualistic symbionts should reach 100% within host populations; however, endophyte frequencies are highly variable. Imperfect VT is emerging as a main driver of endophyte frequencies in host populations. Genetic mismatch between a symbiont and their host’s offspring due to host outcrossing is a proposed, yet underexplored mechanism for imperfect VT. Currently, little is known about how VT rates vary among outcrossing/hybridizing host populations and across environmental gradients. Here, we quantified endophyte infection frequencies and transmission rates in 23 populations across the overlapping ranges of two native host grass species that are known to hybridize, Virginia wild rye (Elymus virginicus) and Canada wild rye (E. canadensis), which are distributed across strong environmental gradients throughout the southern, mid-western, and inter-mountain regions of the US.

Results/Conclusions

Both host grass species identity and local climate (yearly average max temperatures) were the best predictors of endophyte infection frequencies. E. canadensis had significantly higher infection frequencies and transmission rates than E. virginicus throughout its range. Overall transmission rates were significantly higher in populations with high infection frequencies and lower in populations with overall low infection frequencies. These results support the hypothesis that infection frequencies and ultimately transmission may be traits of individuals or populations and perhaps contingent upon host and / or symbiont genotypes. To investigate the possible influence of host gene flow on VT rates, we also tested the significance of species co-occurrence as a predictor of endophyte infection and transmission frequencies. We did this because co-occurrence may influence endophyte frequencies if gene flow between host species is high (hybridization), potentially disrupting local adaptation of co-evolved host/symbiont genotypes. The predictor co-occurrence was within top ranked models, suggesting the possible influence of host gene flow on endophyte frequencies and transmission. This is the first study to quantify endophyte frequencies and transmission across populations and environmental gradients. Additional studies are needed to elucidate the influence of environment and host genetic background on symbiont prevalence and transmission efficiency.