Influence of host genotype and temperature on bacterial colonization.

<p>(A) Experimental design, 2 genotypes for each geographic location were kept in 3 replicates at 3 different temperatures for 3 months; (B) PcoA (based on Jaccard metric, sampling depth = 15,800) illustrating similarity of bacterial communities based on ambient temperature; (C) PCoA (based on Jaccard metric, sampling depth = 15,800) illustrating similarity of bacterial communities based on host genotype; (D) beta-diversity distance box plots between different genotypes (Jaccard metric, sampling depth = 15,800), differences were tested through Kruskal–Wallis test (H = 38.91, <i>p</i> = < 0.001); for clarity the Dunn’s post hoc comparisons are reported in <a href="http://www.plosbiology.org/article/info:doi/10.1371/journal.pbio.3001726#pbio.3001726.s002" target="_blank">S2 Table</a>. (E) Reaction norms plotting average principal component 2 eigenvalues for each of the 12 genotypes at each temperature. NS (Nova Scotia), ME (Maine), NH (New Hampshire), MA (Massachusetts), MD (Maryland), NC (North Carolina), numbers near the location abbreviations indicate the different genotypes. Underlying data can be found in <a href="http://www.plosbiology.org/article/info:doi/10.1371/journal.pbio.3001726#pbio.3001726.s011" target="_blank">S1 Data</a>.</p>