Soil Microbes Trade-Off Biogeochemical Cycling for Stress Tolerance Traits in Response to Year-Round Climate Change

Abstract

Winter air temperatures are rising faster than summer air temperatures in high-latitude forests, increasing the frequency of soil freeze/thaw events in winter. To determine how climate warming and soil freeze/thaw cycles affect soil microbial communities and the ecosystem processes they drive, we leveraged the Climate Change across Seasons Experiment (CCASE) at the Hubbard Brook Experimental Forest in the northeastern United States, where replicate field plots receive one of three climate treatments: warming (+5°C above ambient in the growing season), warming in the growing season + winter freeze/thaw cycles (+5°C above ambient +4 freeze/thaw cycles during winter), and no treatment. Soil samples were taken from plots at six time points throughout the growing season and subjected to amplicon (rDNA) and metagenome sequencing. We found that soil fungal and bacterial community composition were affected by changes in soil temperature, where the taxonomic composition of microbial communities shifted more with the combination of growing-season warming and increased frequency of soil freeze/thaw cycles in winter than with warming alone. Warming increased the relative abundance of brown rot fungi and plant pathogens but decreased that of arbuscular mycorrhizal fungi, all of which recovered under combined growing-season warming and soil freeze/thaw cycles in winter. The abundance of animal parasites increased significantly under combined warming and freeze/thaw cycles. We also found that warming and soil freeze/thaw cycles suppressed bacterial taxa with the genetic potential for carbon (i.e., cellulose) decomposition and soil nitrogen cycling, such as N fixation and the final steps of denitrification. These new soil communities had higher genetic capacity for stress tolerance and lower genetic capacity to grow or reproduce, relative to the communities exposed to warming in the growing season alone. Our observations suggest that initial suppression of biogeochemical cycling with year-round climate change may be linked to the emergence of taxa that trade-off growth for stress tolerance traits.

Keywords: climate change; forest ecology; microbial communities; soil freezing; warming; winter.

FIGURE 1

Shifts in soil (A) fungal and (B) bacterial community composition with climate change treatments at CCASE. Samples are colored by climate treatments. Statistics are derived from multiple regression (MRM) analysis (Supplementary Table 2b). Points represent individual soil samples sequenced from each quadrant in each plot. Ellipsoids represent the 95% confidence interval around the centroid for each climate treatment group.

FIGURE 2

Dynamics of fungal functional guilds in soil at CCASE. Functional guilds include arbuscular mycorrhizal fungi (AMF) (A,E), brown rot fungi (B,F), plant pathogens (C,G), and animal parasites (D,H).

FIGURE 3

Dynamics of bacterial C and N cycling functional guilds in soil at CCASE. N-fixing bacteria were categorized as those with the genetic capacity to produce nitrogenase (A,E), denitrifying bacteria (B,F) were categorized based on the presence of genes involved in complete oxidation/reduction of N from NO₃^– to N₂ [i.e., nitrate reductase (nar G, H, or I), nitrite reductase (nir K, S), nitric oxide reductase (nor B, C), nitrous oxide reductase (nos Z)], and cellulolytic bacteria (C,G) were categorized as those with the genetic capacity to generate either extracellular β-glucosidase (GH 1, 3) and/or cellulases (GH 5, 6, 7, 8, 9, 12, 44, 45, 48). Copiotrophic bacteria (D,H) were categorized based on ecological classification of phyla from the literature.

FIGURE 4

Relationship between microbial C and N cycling functional guilds and biogeochemical processes in soil at CCASE. Change in relative abundance of arbuscular mycorrhizal fungi (AMF) (A), plant pathogenic fungi (B), brown rot fungi (C), N-fixing bacteria (D), cellulolytic bacteria (E) and copiotrophic bacteria (F) each correlated with at least one metric of C and N cycling processes in soil. Changes in functional group relative abundance and biogeochemistry processes were calculated relative to pre-treatment (2013) values averaged at the plot-level. Changes in process rates were correlated to changes in functional groups measured at that same time point.

FIGURE 5

Correlation between NADH/NAD⁺ metabolism genes and genes associated with growth and reproduction in the soil metagenome at full leaf-out.