Soil Phosphorus Exchange as Affected by Drying-Rewetting of Three Soils From a Hawaiian Climatic Gradient

Abstract

Current understanding of phosphorus (P) dynamics is mostly based on experiments carried out under steady-state conditions. However, drying-rewetting is an inherent feature of soil behavior, and as such also impacts P cycling. While several studies have looked at net changes in P pool sizes with drying-rewetting, few studies have dynamically tracked P exchange using isotopes, which would give insights on P mean residence times in a given pool, and thus P availability. Here, we subjected three soils from a climatic gradient on the Kohala peninsula from Hawaii to 5-month drying-rewetting treatments. The hypotheses were that physico-chemical and biotic processes would be differently affected by repeated drying-rewetting cycles, and that response would depend on climatic history of the soils. Soils were labeled with 33P and 18O enriched water. At select time intervals, we carried out a sequential extraction and measured P concentration, 33P recovery (only first 3 months), and incorporation of 18O from water into phosphate. This allowed tracing P dynamics in sequentially extracted pools as well as O dynamics in phosphate, which are driven by biological processes. Results showed that P concentration and 33P recovery were predominantly driven by soil type. However, across all soils we observed faster dilution of 33P from resin-P into less mobile inorganic pools under drying-rewetting. On the other hand, O dynamics in phosphate were mostly governed by drying-rewetting treatment. Under drying-rewetting, considerably less O was incorporated from water into phosphate of resin-P, microbial-P and HCl-P, suggesting that drying-rewetting reduced biological P cycling. Hence, our results suggest that repeated drying-rewetting increases inorganic P exchange while reducing biological P cycling due to reduced microbial activity, independent of climatic history of the soils. This needs to be considered in P management in ecosystems as well as model representations of the terrestrial P cycle.