Into the Deep: Variability in Soil Microbial Communities and Carbon Turnover Along a Tropical Forest Soil Depth Profile

Abstract:

Tropical forest soils store more carbon (C) than any other terrestrial ecosystem and exchange vast amounts of CO₂, water, and energy with the atmosphere. Much of this C is leached and stored within deeper soil layers, but we know exceedingly little about the fate of this C or the microbial communities that drive deep soil biogeochemistry. From the data that do exist, most organic matter (OM) in tropical soils appears associated with mineral particles, suggesting deep soils may provide greater C stabilization due to organo-metal co-precipitation and mineral-surface interactions. However, few studies have evaluated sub-surface soils in tropical ecosystems, the turnover times of deep soil C, and sensitivity of this C to global environmental change.

To address this critical research need, we quantified C pools, microbial communities and soil radiocarbon turnover times in bulk soils and soil fractions [free light (unprotected), dense (mineral-associated)] from 0-140 cm in replicate soil pits in the Luquillo Experimental Forest, Puerto Rico. Unsurprisingly, we found soil C, nitrogen, and root and microbial biomass all declined exponentially with depth; total C stocks dropped from 5.5 % at the surface to <0.5% at 140cm depth. Soil OM ¹⁴C and mean turnover times were variable across replicate horizons, ranging from 3-1500 years at the surface (0-20 cm), to 5000-40,000 years at 140 cm depth. Soil C in the mineral associated fraction was much older than the free light fraction C, which reflected modern ¹⁴C at all depths.

In comparison to temperate deciduous forests, these ¹⁴C values reflect far older soil C, and OM decomposition that highly favors free light C pools, even at depth. While previous work suggests these low C tropical subsoils contain small but metabolically active microbial communities at depths of ~100cm, these organisms appear highly OM limited, and preferentially degrade recent inputs.

In the coming half century, tropical forests are predicted to see a 2 – 5 ° C temperature increase and substantial differences in rainfall amount and timing. The data described here represent baseline data for a site now undergoing a 4°C warming experiment; upcoming research will examine soil C storage and mean residence times during and post-warming to improve numerical models of ecosystem warming effects in tropical forests.