Bacterial Community Dynamics across a Floristic Gradient in a Temperate Upland Grassland Ecosystem

Abstract

Alterations in soil bacterial communities across a transect between a semi natural upland grassland and an agriculturally improved enclosure were assessed using culture-based methods and a nucleic-acid-based method, terminal restriction fragment length polymorphism (TRFLP). While plant diversity decreased across the transect towards the improved area, numbers of 16S rDNA terminal restriction fragments increased, indicating an increase in numbers of bacterial ribotypes. Bacterial numbers, microbial activity, and potential functional diversity also followed a similar trend, increasing with decreasing plant diversity. Alterations in bacterial community structure were coincident with changes in soil physicochemical properties which also changed across the transect. Increases in soil pH, nitrate, phosphorus, potassium, and calcium occurred toward the improved grassland, while organic matter and ammonium declined. The inverse relationship between floristic diversity and bacterial ribotype numbers suggests that soil physicochemical factors may be as influential in determining bacterial diversity in soils of upland grassland communities as floristic diversity.

The Influence of Tropical Plant Diversity and Composition on Soil Microbial Communities

“…It is possible, as suggested by Wardle [59] that plant diversity exerts a detectable effect on soil microbial communities only when plant diversity increases plant productivity. This finding is supported by Zak et al [61], where plant diversity was associated with increases in productivity, as well as Brodie et al [8], where microbial diversity was positively related to soil organic matter concentrations. We did not have productivity estimates for the secondary succession sites in our study, but we observed an increase in soil carbon concentrations with plant diversity, which may be a reasonable index of plant productivity [39]; thus, our study also supports the Wardle [59] hypothesis.…”

“…In fact, soil C/N, an indicator of the relative quality of soil organic matter [41], was the abiotic variable most related to differences in microbial community composition across the plant diversity gradient and among monocultures. We expected soil carbon concentrations to play an important role as well, given the importance of available energy in structuring communities [8,38], but we found no evidence of such a link. Although we did find a correlation between microbial community differences and soil C/N, it was always low (the correlation coefficient never exceeded 0.10).…”

“…Despite intensified interest in understanding the link between above-and belowground biodiversity and the implications that this link may have for ecosystem functioning [2,11,23,40], few studies have examined the relationship between plant diversity and soil microbial communities using culture-independent techniques; the results from the few that have are equivocal [8,9,58,61]. For example, Broughton and Gross [9] examined microbial fatty acid methyl esters (FAMEs) along a topographic gradient in a mid-successional temperate grassland that varied in plant diversity (from 2 to 16 species) and productivity.…”

“…They found no relationship between FAME composition and either factor, but they did find that microbial activity was positively correlated to productivity. Brodie et al [8] used analysis of 16S rDNA to demonstrate that microbial genetic diversity increased with decreasing plant diversity across a grassland transect in Ireland; the increase in microbial diversity was correlated with increases in soil organic matter and microbial biomass. Zak et al [61] found that the abundance of different PLFAs varied among plant diversity levels in a long-term experimental grassland system, and that these differences were linked to the plant productivity.…”

Shifts in microbial communities do not explain the response of grassland ecosystem function to plant functional composition and rainfall change

“…This resulted in a lower TRF number than is 266 reported in some other studies (Kennedy et al 2005; Brodie et al 2008). However, reducing 267 the number of TRFs to focus on the dominant types helps to eliminate artefactual results 268 (Singh et al 2006), and allowed us to use multivariate statistics.…”