Possibility for pathogen escape
Using a subset of seeds collected and processed in the same manner and from the same source populations as our study, Mackin et al. (2021) found that epiphytic (seed surface) pathogen community diversity is significantly higher on chasmogamous than cleistogamous seeds ofD. californica. Although they found higher in vitro germination with chasmogamous seeds, it is likely that the negative effects of antagonistic pathogens on seed germination are diminished in a laboratory setting (Mackin et al. 2021). Seed pathogens can be a significant barrier to in situ germination, although they are rarely accounted for in demography studies (Nelson 2017, Mackin et al. 2021). In general, outcrossed progeny have been shown to have higher infection scores when compared to self-fertilized progeny, suggesting that outcrossing can disrupt pathogen resistance (Busch et al. 2004, Koslow & Clay 2007). There are several theories to explain the maintenance of both cleistogamous and chasmogamous mating systems (Oakley et al. 2007), to which seed-type dependent pathogens could contribute.
It is important to note that not all seed-associated fungi are pathogenic. For example, some species of the Danthonia genus have a symbiotic relationship with the endophytic clavicipitaceous fungiAtkinsonella hypoxylon which improves host plant vigor (Clay 1984) and exclusively uses cleistogamous seeds for transmission (Clay 1994). The fungus also causes infected plants to abort chasmogamous inflorescences, further improving the adaptive potential for cleistogamous seeds (Clay 1984). However, Mackin et al. (2021) did not find any evidence of clavicipitaceous fungi inside any D. californica seeds sourced from the populations used in this study. However, they did find that cleistogamous seeds had fewer fungi negatively associated with germination than chasmogamous seeds, which may partially explain the observed higher germination rates of cleistogamous seeds in our common garden study.
Pathogen escape could explain the higher germination rates of cleistogamous seeds sourced from populations located an intermediate distance from the HD common garden. This fitness advantage quickly dissipated as distance between source and common garden increased, however, suggesting that some degree of local adaptation may still play a role in germination. It appears that populations from farther away are still more maladapted to the common garden site, even if they do benefit from pathogen escape. Only the seeds from intermediately distanced populations likely stood to benefit from pathogen escape in our common garden experiment. Although most transplant experiments occur across much larger spatial scales (Joshi et al. 2001, Gaillart et al. 2018, van Boheemen et al. 2019, Zhang et al. 2019), nonlocal advantage at small scales could be common, especially if environmental heterogeneity is high (Hereford 2009).