Land Use History Shifts In Situ Fungal and Bacterial Successions following Wheat Straw Input into the Soil

Abstract

Soil microbial communities undergo rapid shifts following modifications in environmental conditions. Although microbial diversity changes may alter soil functioning, the in situ temporal dynamics of microbial diversity is poorly documented. Here, we investigated the response of fungal and bacterial diversity to wheat straw input in a 12-months field experiment and explored whether this response depended on the soil management history (grassland vs. cropland). Seasonal climatic fluctuations had no effect on the diversity of soil communities. Contrastingly fungi and bacteria responded strongly to wheat regardless of the soil history. After straw incorporation, diversity decreased due to the temporary dominance of a subset of copiotrophic populations. While fungi responded as quickly as bacteria, the resilience of fungal diversity lasted much longer, indicating that the relative involvement of each community might change as decomposition progressed. Soil history did not affect the response patterns, but determined the identity of some of the populations stimulated. Most strikingly, the bacteria Burkholderia, Lysobacter and fungi Rhizopus, Fusarium were selectively stimulated. Given the ecological importance of these microbial groups as decomposers and/or plant pathogens, such regulation of the composition of microbial successions by soil history may have important consequences in terms of soil carbon turnover and crop health.

Fig 1. Dynamics of bacterial (A and C) and fungal (B and D) richness and evenness during the 12-month experiment in control (full line) and wheat-amended plots (dashed line) for each soil history: grassland (grey) and cropland (black).

Error bars denote standard deviation of biological replicates (n = 3). Symbols in superscript indicate significant differences according to Fisher test (P<0.05) between control and corresponding wheat amended sub-plots for each grassland (stars) and cropland (circles) history.

Fig 2. Non-metric multi-dimensional scaling (NMDS) ordination plot derived from weighted pairwise Unifrac distances for (A) bacterial and (B) fungal communities over time according to soil management history: cropland (squares) and grassland (circles).

White labels represent the control treatment and grey labels represent the wheat-amended treatment. Numbers represent sampling dates and lines correspond to the three replicate samples obtained at each sampling date. Stress values for the two ordination plots were < 0.2 which indicates that these data were well-represented by the two dimensional representation. inset represents magnitude changes in Unifrac distance between the control and amended sub-plots over the year.

Fig 3. Heat map comparison of bacterial and fungal phyla detected in soils according to land management history (cropland and grassland) and treatments (control or amended) over time.

In the amended treatment, the white dashed line represents the time of wheat straw input. The captions show the Z-scores (relative abundances are expressed as median centered Z-scores between all samples, and the colors are scaled to the standard deviations. For each sampling date, the average relative abundance was obtained from the biological replicates (n = 3).

Fig 4. Dynamics of the relative abundance of some bacterial genera over time, according to soil management history (cropland and grassland) and treatment: control (dark grey) and wheat-amended (light grey).

Error bars denote standard deviation of biological replicates (n = 3). Asterisks indicate significant difference between control and amended treatments at each sampling date, according to Fisher test (P<0.05).

Fig 5. Dynamics of the relative abundance of some fungal genera over time, according to soil management history (cropland and grassland) and treatment: control (dark grey) and wheat-amended (light grey).

Error bars denote standard deviation of biological replicates (n = 3). Asterisks indicate a significant difference between control and amended sub-plots at different sampling dates, according to Fisher test (P<0.05).