Environmental Controls on Soil Microbial Communities in a Seasonally Dry Tropical Forest

Abstract

Several studies have shown that rainfall seasonality, soil heterogeneity, and increased nitrogen (N) deposition may have important effects on tropical forest function. However, the effects of these environmental controls on soil microbial communities in seasonally dry tropical forests are poorly understood. In a seasonally dry tropical forest in the Yucatan Peninsula (Mexico), we investigated the influence of soil heterogeneity (which results in two different soil types, black and red soils), rainfall seasonality (in two successive seasons, wet and dry), and 3 years of repeated N enrichment on soil chemical and microbiological properties, including bacterial gene content and community structure. The soil properties varied with the soil type and the sampling season but did not respond to N enrichment. Greater organic matter content in the black soils was associated with higher microbial biomass, enzyme activities, and abundances of genes related to nitrification (amoA) and denitrification (nirK and nirS) than were observed in the red soils. Rainfall seasonality was also associated with changes in soil microbial biomass and activity levels and N gene abundances. Actinobacteria, Proteobacteria, Firmicutes, and Acidobacteria were the most abundant phyla. Differences in bacterial community composition were associated with soil type and season and were primarily detected at higher taxonomic resolution, where specific taxa drive the separation of communities between soils. We observed that soil heterogeneity and rainfall seasonality were the main correlates of soil bacterial community structure and function in this tropical forest, likely acting through their effects on soil attributes, especially those related to soil organic matter and moisture content.IMPORTANCE Understanding the response of soil microbial communities to environmental factors is important for predicting the contribution of forest ecosystems to global environmental change. Seasonally dry tropical forests are characterized by receiving less than 1,800 mm of rain per year in alternating wet and dry seasons and by high heterogeneity in plant diversity and soil chemistry. For these reasons, N deposition may affect their soils differently than those in humid tropical forests. This study documents the influence of rainfall seasonality, soil heterogeneity, and N deposition on soil chemical and microbiological properties in a seasonally dry tropical forest. Our findings suggest that soil heterogeneity and rainfall seasonality are likely the main factors controlling soil bacterial community structure and function in this tropical forest. Nitrogen enrichment was likely too low to induce significant short-term effects on soil properties, because this tropical forest is not N limited.

Keywords: N deposition; functional N genes; nitrogen deposition; rainfall seasonality; soil bacterial community structure; soil heterogeneity; tropical dry forest.

FIG 1

Abundances of indicated marker genes in the two soil types of the experimental plots during the two sampling seasons: 16S rRNA (a), bacterial amoA (b), archaeal amoA (c), nirK (d), and nirS (e). Bars represent average values (n = 3), and error bars show standard errors. Letters in each bar represent significant differences among mean values (Tukey's post hoc test, P < 0.05).

FIG 2

Principal component analysis (PCA) of the soil samples (symbols) and soil properties (black vectors for biochemical parameters and blue vectors for microbial gene abundances; P < 0.05) in the experimental plots. Symbol abbreviations of the samples are as follows: W, wet season; D, dry season; R, red soil; B, black soil; C, control plots; N, N-enriched plots. Symbol abbreviations of the soil properties are as follows: CEC, cation exchange capacity; TC, total C; TN, total N; WSC, water-soluble C; WSN, water-soluble N; TP; total P; Cbiom, microbial biomass C; Nbiom, microbial biomass N; Cm, C mineralization; DHa, dehydrogenase activity; GLa, β-glucosidase activity; URa, urease activity; 16S, 16S rRNA gene; AOA, amoA gene for ammonia oxidizing archaea; AOB, amoA gene for ammonia oxidizing bacteria.

FIG 3

Relative abundances of bacterial taxonomic groups from the experimental plots. Abbreviations below the bars are explained in the legend to Fig. 2.

FIG 4

Nonmetric multidimensional scaling (NMDS) plot of the soil samples based on Bray-Curtis distance metric of soil bacterial community composition from the experimental plots. Soil biochemical properties were fitted onto the NMDS ordination, and only significant variables are shown (P < 0.05). Stress value is 0.001. Abbreviations are explained in the legend to Fig. 2.

FIG 5

Venn diagrams showing the numbers of shared and unique OTUs among the experimental plots in the two sampling seasons. Total observed richness was 90,553 OTUs at 97% similarity. Percentages reflect OTUs unique to that treatment type. Abbreviations are explained in the legend to Fig. 2.

FIG 6

Heatmap of core OTUs present in 95% of the soil samples and with relative abundances of >0.5% of the total core microbiome. The value shown in the color key scale represents the number of sequences detected for each OTU from lower (white) to higher (dark blue) abundance. Rows indicate OTUs, and columns indicate the different soil types and treatments in the two sampling seasons. Abbreviations are explained in the legend to Fig. 2.