If parasite populations adapt to local host communities, it follows that the abiotic environments which shape host communities also shape distributions of generalist and specialist parasites (Q3). In support of our ENM-based approach, we found that host-specific ENMs readily distinguished parasite populations along environmental axes associated with differences in host performance (Fig. 4A-D). These axes may correspond to a multivariate ‘optimum’ that differs for parasites on different host species, possibly due to host environmental tolerances (Reynolds et al. 2015). Parasites may perform best on hosts that are stressed enough to exude chemical signals for recruitment of mycorrhizae, signals which are also required for parasites to sense hosts and germinate (Yoneyama et al. 2010). However, parasite reproductive fitness may be low if hosts are too stressed to provide adequate nutrition.  Supporting an important role of abiotic environment in shaping distribution of specialist parasites, we found significant associations between ENM contrasts and relative parasite emergence on pearl millet and sorghum observed in common garden experiments and in pot trials (Fig. 3G-H; Table S6).