FrankiaAssemblages Associated withAlnus rubraandAlnus viridisAre Strongly Influenced by Host Species Identity

Abstract

This study examined the effect of host species identity on the structure of Alnus-associated Frankia bacterial assemblages in the Pacific Northwest, United States, using two approaches. First, Frankia in nodules were sampled from six stands of Alnus rubra or Alnus viridis. Second, a bioassay was conducted where A. rubra and A. viridis seedlings were grown in different soils collected from these two hosts. Frankia genotypes were characterized with nifH sequences and bacterial assemblages were compared using taxon-and divergencebased analyses. Strong host associations were evident in the field; the dominant Frankia genotypes showed significant associations with either A. rubra or A. viridis, and there were host-associated groupings at the assemblage level as well. In the bioassay, host associations among Frankia genotypes were evident but less pronounced, reflecting an interaction between host species and other factors. Although nodule abundance varied among bioassay treatments, seedling dry mass was not strongly correlated with either nodule quantity or soil chemistry. Collectively, our observation and experimental results indicate that host identity is a major factor influencing the genotype composition and abundance of Alnus-associated Frankia assemblages.

Symbiotic Frankia bacteria in Alnus forests in Mexico and the United States of America: is geographic location a good predictor of assemblage structure?

“…Lipus and Kennedy (2011) demonstrated that the observed host-association pattern in USA Alnus forests was reproducible in reciprocal soil transplant experiments, and therefore appears to reflect a preference on the part of either the host or the microsymbiont for certain partners over others. Lipus and Kennedy (2011) demonstrated that the observed host-association pattern in USA Alnus forests was reproducible in reciprocal soil transplant experiments, and therefore appears to reflect a preference on the part of either the host or the microsymbiont for certain partners over others.…”

“…In molecular diversity analyses of Frankia using nifH, the cutoff level of ≥97% sequence similarity results in taxonomic groups that reliably correlate to distinct, reproducible clades in phylogenetic analyses (Mirza et al 2009;Welsh et al 2009a;Kennedy et al 2010b;Lipus and Kennedy 2011). In addition, Lipus and Kennedy (2011) found evidence of host species-level associations among genotypes defined at this sequence similarity level.…”

“…We have previously documented that Frankia genotype richness in the northwestern USA forests is typically low (i.e., 2-5 genoytpes) (Kennedy et al 2010a, Kennedy et al 2010b, Lipus and Kennedy 2011, and a recent study of Frankia richness in root nodules of A. oblongifolia at three locations throughout the state of Arizona also yielded only three Frankia nifH genotypes at the ≥97% sequence similarity level (Welsh et al 2009b). We have previously documented that Frankia genotype richness in the northwestern USA forests is typically low (i.e., 2-5 genoytpes) (Kennedy et al 2010a, Kennedy et al 2010b, Lipus and Kennedy 2011, and a recent study of Frankia richness in root nodules of A. oblongifolia at three locations throughout the state of Arizona also yielded only three Frankia nifH genotypes at the ≥97% sequence similarity level (Welsh et al 2009b).…”

“…Clean sequences were trimmed to 574 bp and aligned using MUSCLE (Edgar 2004). The term KL is used to distinguish among the OTU groups recovered at ≥97% similarity, a threshold that has been commonly used in previous studies of Frankia ecology and results in readily reproducible clusters using various methods (Mirza et al 2009;Welsh et al 2009a;Kennedy et al 2010b;Lipus and Kennedy 2011). Hereinafter, the term genotype is used to refer to an OTU group at a given sequence similarity threshold.…”

“…The NMDS plot was generated from Bray-Curtis similarity values that were calculated from a square-root transformed site-by-genotype abundance matrix. in central Arizona (Welsh et al 2009b); (iii) 19 sequences representing Alnus-infective Frankia genotypes from the northwestern USA (Kennedy et al 2010b;Lipus and Kennedy 2011); (iv) 6 sequences representing the two novel Frankia genotypes recovered in Mexico (see below); and (v) 4 sequences representing the four Elaeagnus-infective Frankia groups (since Alnus species are known to occasionally host Frankia strains that typically infect members of the genus Elaeagnus (Bosco et al 1992)). Fast Uni-Frac uses phylogenetics-based analyses that differ from the OTU-based NMDS approach, which disregards the differences among sequences within any given OTU.…”