Mixed Disease Phenotype Results
Despite the significant trend of the number of R genes increasing with CMI, leaf rust severity, leaf rust response, and stem rust severity did not significantly vary in the common gardens. These three traits are different phenotypic metrics to assess a specialist fungal pathogen,Puccinia silphii , commonly known as Silphium rust. These results are consistent with results from common gardens that found that eastern populations of S. integrifolium were not, on average, more resistant to Silphium rust (Cassetta et al., 2023). One possible explanation for why resistance to Silphium rust was not associated with variable R gene counts may lie in the concept of the ‘disease triangle’ reviewed by Velásquez et al., 2019. The clones in this study were placed in environments where S. integrifoliumpathogens were present, and experienced conditions that would be ideal for pathogenic growth: warm and wet. Even with effective immunity represented by the R genes sequenced in the plant populations, the level of resistance conferred by the resistance genes might have been obscured by the pathogen spread. Disease resistance in these wild populations is highly likely to be quantitative, meaning that resistance is not an all-or-nothing response, but rather a distribution of potential responses ranging from minor to major (French et al., 2016). Further studies of the wild populations with sequenced R genes, with controlled inoculations of plants with pathogens, would elucidate associations between R genes that may confer various levels of response to pathogens.
Further, the evolutionary dynamics of R genes may be muddled by seemingly competing forces of selection, with “arms race” dynamics expected to generate gradients in total host resistance, while oscillatory “red-queen” dynamics may not (Bergelson et al., 2002). It is possible, for example, that local co-evolution of S. integrifolium and Silphium rust has resulted in adaptation of the rust to its local host population, rather than variation in overall resistance. This red-queen dynamic is supported by observations from common gardens distributed along the gradient that showed much greater virulence of Silphium rust on their populations from their local region than from other regions (Cassetta et al. 2023). In the case ofSilphium rust, R gene identity may be more important than R gene diversity.
We did observe that resistance allele diversity was associated with suppression of leaf blotch (Fig. 6d). Our result affirms observations of a previous common garden study that found that eastern populations ofS. integrifolium were more resistant to Silphium blotch, as well as to clear vein virus (Cassetta et al. 2023). These results are consistent with expectations of greater diversity of resistance alleles contributing to greater resistance to pathogens, and are qualitatively consistent with arms-race-type dynamics. In S. integrifolium , leaf blotch is caused by the generalist fungal pathogenColletotrichum dematium (Cybernome; Farr 1989), as well as its more specialized congener, Colletotrichum silphii Davis (Horst 2008). It is possible that arms race dynamics are more likely between hosts and generalist pathogens such as leaf blotch, while specialist pathogens such as the Silphium rust are more likely to generate red-queen dynamics.
We note that the observed relationship between resistance allele diversity and resistance to blotch may be spurious, as other heritable factors that covary with rainfall might contribute to disease resistance, such as plant secondary chemicals. We note, however, that insect herbivory may be an indicator of such overall resistance, and we did not observe any relationship between R gene diversity and insect leaf herbivory. This result is consistent with R genes predominantly acting as pathogen effector receptors, and not targeted toward resistance for insects (Chovelon et al., 2021).
A BROADER PATTERN OF R GENE EVOLUTION