Genetic variation in filtering force of mutualistic partnerships
Despite symbiosis filtering to favor specific genotypes, genetic diversity of Frankia was high in the rhizosphere soil under a host plant. Consistently, previous studies have found that genetic variation in mutualistic partners is sustained in nature (Barrett et al, 2015; 2016; Clawson et al., 1998; 1999; Huguet et al., 2001; Põlme et al., 2014). However, stabilizing mechanisms alone cannot explain the persistence of the variation because the evolutionary fixation of the most beneficial partner genotypes is predictable (Heath & Stinchcombe, 2014; Law & Koptur, 1986; Parker, 1999; Weyl et al., 2010). Furthermore, the fixation of the beneficial genotypes may lead to the loss of discrimination ability by hosts if it needs a cost (Foster & Kokko, 2006; Frederickson, 2013; Heath & Tiffin, 2009; McNamara & Leimar, 2010). One solution to this paradox is mutation-selection balance (Heath & Stinchcombe, 2014): the hypothesis that variation in partner quality is maintained by the balance in mutation, which introduces genotypes of lower quality, and selection, which acts to favor the partner genotypes of higher quality. Genetically diverse pool and strong filtering for symbiosis found in this study supports the mutation-selection balance hypothesis. This study suggests mutation-selection balance in the field. Future studies should pay attention to quantifying the strength of selection that stabilizing mechanisms impose on partner quality in nature. This would provide great insights for an understanding of the evolution of partner quality and the maintenance of genetic diversity in symbiosis (Heath & Stinchcombe, 2014).
This study quantified the symbiosis filtering force using a community dissimilarity approach and found the possibility of genetic variation of hosts in the filtering forces in the field (Fig. 3). The reason why similar bacterial compositions were sustained among almost all alders in spite of the variation in the filtering force among host individuals is likely to be explained by the spatial heterogeneity in genetic diversity of Frankia . Frankia genotypes are heterogeneously dispersed in forests (Kagiya and Utsumi 2020). Additionally, NMDS revealed the high variation of Frankia composition in rhizosphere soils among host individuals (Fig. 2a). The necessity of filtering force could depend on genetic diversity and/or the relative abundance ofFrankia genotypes in the surrounding local soil. In fact, the genetic community structure of bacterial symbionts in the host’s rhizosphere soil differed in line with soil nutrients, whereas no effects of genetic variation in a host plant were detected to shape rhizospheric bacterial communities (Table 1). If the ability of filtering requires costs and a common optimal symbiont composition is consistent among host individuals, the balance between the necessity and costs can maintain the homogeneity of symbiont composition in a host plant along with genetic variation in filtering force.
The variation in filtering force might also contribute to the maintenance of genetic variation in microsymbionts. When variation in stabilizing mechanisms alters selection pressure for microsymbionts, genetic variation in microsymbionts, even those involving lower quality genotypes, may be sustained (Heath & Stinchcombe, 2014). For example, some studies have reported that sanctions are not perfectly effective because they cannot respond to non-zero benefit levels or cannot punish individual partner genotypes (Charlotte et al., 2012; Kiers et al., 2006). If genetic variation in a host plant alters the strength of symbiosis filtering, genetic variation in the host will be the key to revealing coevolution in mutualistic interactions under natural ecosystems with a high genetic diversity of microsymbionts.
The present study clearly demonstrates that host plants interact with specific symbiont genotypes via symbiosis filtering in the field. Our finding that symbiosis filtering indices correlated with the genetic distance of hosts is likely to be evidence of linkage between host genetics and symbiont assembly in nature, even if symbiosis filtering indices depend on the dissimilarity of Frankia composition in rhizosphere soils among rivers and/or non-infective strains that make upFrankia communities in rhizosphere soils. Toward an integrative understanding of mutualistic coevolution and community assembly in natural ecosystems, future studies should pay attention to differences in partner quality and infectivity between infecting and non-infecting members of communities. Moreover, revealing the genetic basis underlying stabilizing mechanisms is fruitful toward an understanding of the interplay between coevolutionary dynamics in mutualism and community assembly in symbiosis.