top of page
_edited_edited.jpg)
Active Research Areas
Subsurface impacts of land-cover change in South African savannas and grasslands
Many savanna and grassland ecosystems in South Africa are experiencing increases in woody plant cover through encroachment by savanna trees / shrubs, expansion of native forests, or plantation (pine and eucalypt species) establishment. These changes represent major shifts in dominant vegetation cover, and can lead to substantial changes in soil structure, water fluxes and groundwater recharge, and carbon / nutrient cycling in the subsurface. In multiple projects centered in northeastern South Africa, we are assessing (1) how climate governs the hydrological impacts of bush encroachment in savannas, (2) the legacies of tree and shrub root systems after mechanical woody plant removal, and (3) the impacts of plantation establishment on water, carbon, and nutrient dynamics in higher-elevation ecosystems in the Northern Drakensbergs. Collaborators include Pam Sullivan (OSU), Tony Swemmer (SAEON, South Africa), and Jesse Nippert (KSU).
Shrub encroachment: Implications for tallgrass prairie water cycling in the central U.S.
Grasslands in the central U.S. and across the world are experiencing large-scale conversions from grassland to shrub- or woodland. This process is commonly referred to as 'woody encroachment', and is one of the primary threats (along with conversion to row-crop agriculture) to grassland ecosystems globally. Well known consequences of woody encroachment include reductions in plant biodiversity, loss of grassland-obligate birds and small mammals, and alterations to carbon storage and cycling. Transitions from grass- to woody- plant dominance on the landscape can also have profound impacts on water cycling in these ecosystems. In mesic grasslands, increasing woody cover can lead to increased vegetation water-use, increased canopy interception of rainfall, and the growth of deeper / coarser root systems that can alter soil structure and pathways of soil water infiltration. In a series of projects primarily located at Konza Prairie Biological Station in northeastern KS, we explore the consequences of woody encroachment for tallgrass prairie water cycling, including (1) landscape-scale evapotranspiration, (2) groundwater recharge and stream discharge, and (3) infiltration dynamics and seasonal soil wetting and drying trends.
In the media:
Associated publications:
Jarecke et al., 2025. Woody Encroachment Intensifies Deep Soil Drying at Daily, Seasonal, and Decadal Scales
Jarecke et al., 2025. Woody Encroachment Modifies Subsurface Structure and Hydrological Function
Anhold et al., 2025. Grassland woody encroachment alters subsurface mineral weathering and groundwater composition in a carbonate system
Keen et al., 2024. Unexpected hydrologic response to ecosystem state change in tallgrass prairie
Keen et al., 2024. Save or spend? Diverging water‐use strategies of grasses and encroaching clonal shrubs
Sadayappan et al., 2023. Drier streams despite a wetter climate in woody-encroached grasslands
Dodds et al., 2023. Trajectories and state changes of a grassland stream and riparian zone after a decade of woody vegetation removal
Keen et al., 2023. Impacts of riparian and non-riparian woody encroachment on tallgrass prairie ecohydrology
Drivers and consequences of preferential flow across sites and with soil depth
Preferential flow (PF) refers to the rapid movement of water through macropores, which bypasses the majority of the soil matrix. PF is exceedingly important for understanding soil moisture dynamics, vertical transport of solutes (and pollutants!) through soil profiles, and groundwater recharge. However, it is challenging to detect and quantify, and coordinated PF data across multiple sites is very rare. In a collaboration with a large group of researchers (AI4PF), we are using high resolution soil moisture data from sites across the United States to assess the probability of preferential flow occurrence and spatial uniformity based on climate, soil characteristics, and root distributions. In particular, we are interested in understanding how the drivers of preferential flow vary with soil depth and the consequences of preferential flow for recharge, water storage, and solute transport. Collaborators for this project include Matthias Sprenger (NC State), Pam Sullivan (OSU), Bonan Li (OSU), Sharon Billings (KU), and many others within the AI4PF working group.
Associated publications:
Li et al., In Review. Dominant controls on preferential flow and their implications for future soil water fluxes
Past Work
Early warning signals for drought-related forest mortality in the Sierra Nevadas
This project took place in the southern Sierra Nevada Mountains in central California, and centers around the 2012-2015 drought event. This severe, multi-year drought was punctuated by an epidemic-scale outbreak of western pine beetle (Dendroctonus brevicomis), and resulted in widespread mortality of ponderosa pines (Pinus ponderosa), among other dominant canopy tree species. In these studies, we assessed severity of drought stress in surviving trees vs. trees killed by bark beetles during the drought, and found that mortality during this drought event was preceded by decades of increasing sensitivity of tree growth and carbon isotope discrimination (∆13C) to Palmer Drought Severity Index (PDSI) and other hydroclimate variables. We propose that these shifts in sensitivity can serve as early warning signals for future drought-related mortality events in forests of the western United States.
Associated publications:
Keen et al., 2022. Changes in tree drought sensitivity provided early warning signals to the California drought and forest mortality event.
Keen et al., 2020. Stronger influence of growth rate than severity of drought stress on mortality of large ponderosa pines during the 2012–2015 California drought
bottom of page