Background/Question/Methods Woody plant expansion (WPE) into grasslands has been extensively documented across much of the globe and has large structural and functional consequences for grasslands. What is less certain is how WPE affects ecosystem-scale biogeochemical cycling, and in particular, proportions of above- and belowground carbon (C), nitrogen (N), and phosphorus (P) pools. One hypothesis (H1) is that WPE could induce a “sink-driven” depletion of soil N and P via uptake into biomass pools while simultaneously decreasing soil C via a positive priming effect. Conversely (H2), WPE could increase soil nutrients and soil C via deep-soil mining, mycorrhizal-driven nutrient acquisition, N-fixation, and slowed soil organic matter decomposition. A final hypothesis (H3) is a combination of outcomes, where WPE drives progressive nutrient limitation of N and P by reallocating soil N and P in aboveground biomass while increasing soil C pools. To explore these hypotheses, we examine WPE rates and the subsequent biogeochemical implications across a woody cover gradient in the Northern Great Plains (NGP). We use a combination of dendrochronology, aerial imagery, stoichiometry, and soil chemistry to 1) quantify long-term WPE rates 2) simulate biomass and soil C, N, and P (CNP) changes over a 60-year span, and 3) examine the possible effects of WPE on biomass and soil CNP pools. Results/Conclusions Here we compile the most comprehensive evidence of WPE in the NGP. We leverage a gradient of woody cover in 25 long-term monitoring plots in the heart of the NGP. While some trees have been part of the ecosystem for centuries, we find strong evidence for WPE in recent decades. Comparing aerial imagery from 1953 and 2013, we find that two-thirds of our study region increased in tree cover by ~38% relative to 1953. Furthermore, increasing tree cover was negatively associated with soil N and P, but positively associated with soil C. Together these findings support our third hypothesis (H3) that WPE is driving progressive N and P limitation. Comparing simulated soil N and P declines with biomass N and P uptake, we find that soil N losses are well explained by tree uptake but soil-biomass P changes are more challenging to account for. We speculate on why landscape patterns of P are more challenging to explain in this system undergoing extensive WPE. Nonetheless, without new N and P inputs, we expect the ecosystem will continue to become progressively nutrient-limited which may induce a bottleneck to future growth.
