Drivers of bacterial β-diversity depend on spatial scale

Abstract

The factors driving β-diversity (variation in community composition) yield insights into the maintenance of biodiversity on the planet. Here we tested whether the mechanisms that underlie bacterial β-diversity vary over centimeters to continental spatial scales by comparing the composition of ammonia-oxidizing bacteria communities in salt marsh sediments. As observed in studies of macroorganisms, the drivers of salt marsh bacterial β-diversity depend on spatial scale. In contrast to macroorganism studies, however, we found no evidence of evolutionary diversification of ammonia-oxidizing bacteria taxa at the continental scale, despite an overall relationship between geographic distance and community similarity. Our data are consistent with the idea that dispersal limitation at local scales can contribute to β-diversity, even though the 16S rRNA genes of the relatively common taxa are globally distributed. These results highlight the importance of considering multiple spatial scales for understanding microbial biogeography.iodiversity supports the ecosystem processes upon which society depends (1). Understanding the mechanisms that generate and maintain biodiversity is thus key to predicting ecosystem responses to future environmental changes. The decrease in community similarity with geographic distance is a universal biogeographic pattern observed in communities from all domains of life (as in refs. 2-4). Pinpointing the underlying causes of this "distance-decay" pattern continues to be an area of intense research (5-9), as such studies of β-diversity (variation in community composition) yield insights into the maintenance of biodiversity. These studies are still relatively rare for microorganisms, however, and thus our understanding of the mechanisms underlying microbial diversity-most of the tree of liferemains limited.β-Diversity, and therefore distance-decay patterns, could be driven solely by differences in environmental conditions across space, a hypothesis summed up by microbiologists as, "everything is everywhere-the environmental selects" (10). Under this model, a distance-decay curve is observed because environmental variables tend to be spatially autocorrelated, and organisms with differing niche preferences are selected from the available pool of taxa as the environment changes with distance.Dispersal limitation can also give rise to β-diversity, as it permits historical contingencies to influence present-day biogeographic patterns. For example, neutral niche models, in which an organism's abundance is not influenced by its environmental preferences, predict a distance-decay curve (8, 11). On relatively short time scales, stochastic births and deaths contribute to a heterogeneous distribution of taxa (ecological drift). On longer time scales, stochastic genetic processes allow for taxon diversification across the landscape (evolutionary drift). If dispersal is limiting, then current environmental or biotic conditions will not fully explain the distance-decay curve, and thus geographic distance will b...

Turnover of soil bacterial diversity driven by wide-scale environmental heterogeneity

“…The average similarities decreased significantly with increasing spatial scale and were always lower than the corresponding initial similarity (0.65 and 0.84 for w 16 and E d16 , respectively). This is in agreement with the hypothesis that neighbouring sites are more similar than distant sites 2,15 . The significant decrease of z bacteria with increasing spatial scales might be explained by the higher average abundance per OTU with increasing scales 21,22 .…”

“…This conclusion for the soil bacterial community as a whole is in agreement with phylogeographic studies focusing on particular soil bacterial groups 15,16,39 and with studies showing a distance effect on soil bacterial community structure 2 .…”

“…0.02), even in large areas of contiguous habitats, suggesting that communities of both bacteria and macroorganisms are structured by analogous processes 31 . Recent microbial biogeography studies have revealed that the main environmental filters shaping spatial microbial diversity distribution are soil physico-chemical characteristics, land use and plant cover, whereas climatic and geomorphologic filters are less important 9,11,15,37 . Similar filters were also reported to explain soil bacterial community distribution in contrasted ecological regions at the scale of France 27,34 .…”

“…Nowadays, TAR is commonly used in this form in most microbial biogeographical studies 3,9,[14][15][16] to assess microbial diversification and its potential relative dependency on ''dispersal'' and ''habitat heterogeneity'' (including habitat diversity and landscape configuration) 4 . Nevertheless, the TAR is still highly debated, particularly regarding its form, underlying hypotheses and factors affecting the relationship [17][18][19] .…”

“…Rosindell et al 19 highlighted that the TAR follows a power-law relationship under the hypothesis of infinite community size and for a steady z. As the soil microbial community is very large and diverse 20 , the relationship for these organisms exhibits power-law behaviour because the average abundance per microbial taxonomic unit is large according to Harte et al 21 On the other hand, the constant z hypothesis across spatial scales, which assumes self-similarity as a probability rule 13,18 for the spatial distribution of taxa abundance, may not hold for soil microbes 15,21,22 . Therefore, z may be considered constant only at a given scale of investigation, not across spatial scales 21 , and such spatial variations must be specified for microbial communities.…”