Elevational Gradients Impose Dispersal Limitation on Streptomyces

Abstract

Dispersal governs microbial biogeography, but the rates and mechanisms of dispersal remain poorly characterized for most microbial taxa. Dispersal limitation is driven by limits on dissemination and establishment, respectively. Elevation gradients create striking patterns of biogeography because they produce steep environmental gradients at small spatial scales, and these gradients offer a powerful tool to examine mechanisms of dispersal limitation. We focus on Streptomyces, a bacterial genus common to soil, by using a taxon-specific phylogenetic marker, the RNA polymerase-encoding rpoB gene. By targeting Streptomyces, we assess dispersal limitation at finer phylogenetic resolution than is possible using whole community analyses. We characterized Streptomyces diversity at local spatial scales (100 to 3,000 m) in two temperate forest sites located in the Adirondacks region of New York State: Woods Lake (<100 m elevation change), and Whiteface Mountain (>1,000 m elevation change). Beta diversity varied considerably at both locations, indicative of dispersal limitation acting at local spatial scales, but beta diversity was significantly higher at Whiteface Mountain. Beta diversity varied across elevation at Whiteface Mountain, being lowest at the mountain's base. We show that Streptomyces taxa exhibit elevational preferences, and these preferences are phylogenetically conserved. These results indicate that habitat preferences influence Streptomyces biogeography and suggest that barriers to establishment structure Streptomyces communities at higher elevations. These data illustrate that Streptomyces biogeography is governed by dispersal limitation resulting from a complex mixture of stochastic and deterministic processes.

Keywords: assembly; bacterial; biogeography; community; diversity; microbial; soil.

FIGURE 1

The sites at Whiteface Mountain (WM; top) span more than 1,000 m elevation while those at Woods Lake (bottom) span less than 100 m elevation. The topographical profile is provided in the inset. Horizontal distances (X-axis) are measured as geodesic distance or the shortest distance between the GPS coordinates of each site while the Y-axis represents difference in elevation relevant to the base elevation at WM.

FIGURE 2

Alpha diversity is represented as number of observed OTUs in each sample. Streptomyces at Whiteface Mountain (left) exhibited maximal richness at mid elevations with decreasing alpha diversity at the highest and lowest sites. Streptomyces at Woods Lake (WL; right) exhibited little change in alpha diversity with respect to elevation change. Sample names are provided instead of elevations in the right panel because elevation does not vary across the WL watershed (largest elevation difference is 33 m, Table 1).

FIGURE 3

Partial Mantel tests indicate that beta diversity varies significantly with respect to elevation at Whiteface Mountain (A) but not at Woods Lake (B), and that beta diversity does not significantly correlate with horizontal distance at either site (C,D). Filled points indicate significant correlation (p < 0.01).

FIGURE 4

Relative contributions of selection, dispersal, and ecological drift to community assembly vary between sites and within elevation zones in Whiteface Mountain. For WM, High indicates elevations above 1,000 m, Low indicates elevations below 500 m, and Middle indicates 500–1,000 m. (A) Variable selection drives beta diversity at sites above 1,000 m in WM while lower elevations have higher levels of homogenizing dispersal resulting in reduced beta diversity. (B) Dispersal limitation is highest between sites above 1,000 m and the rest of the mountain, while dispersal plays an important role in homogenizing Streptomyces communities between lower elevations. (C) Community assembly at WL is dominated by ecological drift with some variable selection.

FIGURE 5

Scatter plots show the distribution of pairwise phylogenetic distances (A) between indicator OTUs below 500 m, and (B) between indicator OTUs above 1,000 m. The histograms represent a bootstrapped distribution of 1,000 random draws from pairwise phylogenetic distances within the WM community, and 95% confidence intervals are indicated by vertical dashed lines. Each dot represents the phylogenetic distance between a pair of OTUs in that respective category.

FIGURE 6

Abundance-weighted average elevation of each indicator OTU or its conspecific representative is represented as preferred elevation for (A) continental-scale distribution of Streptomyces and (B) the distribution of Streptomyces at WM. The dashed line indicates maximum elevation sampled at WM. Preferred elevations are significantly different between High and Low indicator OTUs at WM (Welch’s t-test, p < 0.001).

FIGURE 7

Phylogenetic reconstruction of indicator OTU lineages shows clade-level conservation of elevational preferences within Streptomyces found at WM. Some clades show mixed elevation preferences with branch lengths suggesting that indicator OTUs above 1,000 m have greater evolutionary divergence relative to those found at lower elevations. Colors are indicative of elevation and dataset for each OTU as indicated in the legend.