SPECIES-LEVEL EFFECTS ON LONG-TERM SOIL PHOSPHORUS LOSS IN BOTTOMLAND FORESTS

Quantifying the interactions between plants and soil P, an important rock-derived nutrient, remains a challenge. This is because it is difficult to disentangle the different mechanisms involved in P limitation and loss. We address this challenge by exploring forest-soil interactions at 12 modern alluvial terrace sites in the Clarks River basin of western KY: six soils underlying forests dominated by more drought-tolerant post oak (Quercus stellata) and six soils underlying forests dominated by cherry bark oak (Quercus pagoda). The whole-soil geochemistry was examined using mass-balance. Two of the 12 sites were chosen for more in-depth measurements of clay mineralogy, exchangeable chemistry and soil moisture, along with tree and leaf litter chemistry. Results show 4x to 6x more soil P loss under post oak when compared to cherry bark oak forests. The P stock in post oak wood is >4x that of cherry bark oak; yet, leaf litter stocks of P do not differ by forest type. High-resolution soil moisture data suggests that the post oak forest has more efficient or rapid water uptake (or more generally higher ET), which may drive the larger aboveground storage of P in the wood. However, our mass balance suggests that plant uptake of P does not account for all soil P loss.

Interestingly, micromorphology of subsoil under a post oak forest shows significantly more soil fabric with high birefringence lining the exterior surface of Fe concretions/nodules (i.e., pressure faces) than in the subsoil under a cherry bark oak forest. Because Fe-Mn oxides can serve as an important sink for soil P, we argue that soil underlying post oak forests has lost more P than cherry bark oak forests because higher ET drives more shrink-swell and the "encasing" of Fe-Mn oxides with clay. This "encasing effect" prevents further storage of P. Ultimately, this P is either taken up by the plant or lost from the profile through drainage. Our work shows that long-term soil P loss is sensitive to changes in forest type at the species level and involves a complex plant-soil feedback. Furthermore, a deeper understanding of modern soil P dynamics will benefit the deep-time paleosol community, who often rely on bulk P as a proxy for soil biology.