Soil microbial functions are controlled by moisture, not vegetation history, in neotropical rainforests

Background/Question/Methods

Tropical rainforests are expected to experience increased seasonality over the next century, but potential soil feedbacks to climate change are poorly understood in tropical ecosystems. In addition, climate effects on soil microbes must be considered in the context of the plant community, which is often a primary driver of soil function and may be particularly important in hyperdiverse tropical forests. Here we present evidence that belowground microbial function responds to soil moisture regardless of historical vegetation, and these effects will likely be exacerbated by climate change. We collected soils from 27-year-old, monodominant, replicated plantations of four neotropical forest trees, along with secondary and primary forest control plots. To examine soil microbial functional responses, soils were maintained in microcosms at four levels of gravimetric soil moisture. The moisture treatments represented typical soil moistures in the wetter (79%) and drier (70%) seasons, as well as two standard deviations above (85%) or below (57%) average. After 6 weeks, we measured microbial biomass and the catabolic potential of 26 carbon substrates ranging from simple sugars and amino acids to the complex polymers of cellulose and chitin. Microcosms were complemented by seasonal field measurements of soil enzyme activities, microbial biomass and metagenomic profiles.

Results/Conclusions

The catabolic potential of soils was positively correlated (r² > 26%) with soil moisture regardless of chemical complexity or vegetation type. Similarly, in control soils that were not amended with carbon, soil respiration was a positive function of soil moisture (r² = 0.18) throughout the experiment. Microbial biomass was invariant with soil moisture in the microcosms, but microbial biomass collected from the same field plots over two years differed as a function of soil moisture, indicating that biomass responses may occur over longer time periods than our experiment. Enzyme activities in field soils confirm that functional potential of soils corresponds with soil moisture. Furthermore, metagenomics sequencing of soils from the field plots indicates that gene abundances related to carbon utilization may also be tracking moisture. The sensitivity of tropical soil carbon fluxes to soil moisture has been hypothesized to be due to the accumulation of labile carbon pools during dry periods. Here we show that moisture sensitivity of carbon fluxes occurs regardless of the recalcitrance of carbon inputs. Because the timing and magnitude of rainfall are expected to shift more than two fold over the next century, carbon storage in these soils may be especially vulnerable.