Current research:

Tree recruitment limitation in Serengeti National Park, Tanzania

This NSF-funded project is being conducted in collaboration with Michael Anderson of Wake Forest University. The overarching goal of the project is to identify the stages (seed, seedling, sapling) and mechanisms (fire, herbivory, soil moisture limitation, tree-grass competition) driving tree recruitment limitation across a precipitation gradient. Exclosure with seedlingsWe are using a combination of observational methods (longitudinal tracking of individual trees within a network of 40 0.1-ha plots across the Serengeti rainfall gradient), targeted experiments (seed and seedling transplants), remote sensing and modeling to tease apart the drivers of tree cover limitation in this system.

Spatial processes and regime shifts in savannas

We are currently in the third year of an NSF-funded project in Serengeti that is 1) examining how spatial processes (primarily facilitation effects of tree canopies) drive tree cover dynamics, and 2) using remote sensing tools to reconstruct spatiotemporal patterns of tree cover change to identify regime shifts and their drivers.

Ecohydrological drivers of tree-grass dynamics

A lot of attention has been devoted to understanding the relative importance of bottom-up and top-down factors in determining the distribution of savannas and explaining tree-grass coexistence across the savanna biome. What is the ultimate factor limiting tree biomass in terrestrial ecosystems? How important are niche partitioning of soil moisture between trees and grasses versus ‘demographic bottlenecks’ in determining tree cover? At what spatial scale do trees and grasses interact? The lack of a general model that can address some of these basic questions partly arises from the fact that savannas are not monolithic, and how trees interact with grasses or respond to resources and disturbance can vary widely among savanna types. A typical tree in an African savanna ecosystem must contend with a wide range of challenges throughout its lifetime, including competition with other trees and grasses for water in highly stochastic environments and periodic biomass losses to fire and megaherbivores. We are currently pursuing two specific projects in this area:


– Developing a mechanistic understanding of water use and partitioning between trees and grasses and among tree species. We are using isotopic methods coupled with time-domain reflectometry, sap flow sensors, and infra-red thermometers to reconstruct tree rooting profiles and monitor and model soil moisture dynamics and tree water use at Wits Rural Facility, South Africa. By coupling observational methods with models and experimental manipulations of water availability and tree-tree and tree-grass interactions, we can evaluate the extent of niche overlap (in terms of water use) between functional groups and species in tropical savanna systems, and characterize species-specific water-use strategies. In the future, we will scale up this work to examine functional rooting profiles across rainfall and edaphic gradients to derive a general model of savanna rooting strategies. Collaborator: Jesse Nippert (Kansas State University).

Kruger dendro work

– Using modeling tools to link tree ecophysiology with demography. We are combining climate records with local soil conditions to model past patterns of soil moisture dynamics for individual trees. In conjunction with rooting profile data, this will lead to the reconstruction of historical patterns of water use, which will then be combined with dendrochronological data to derive mechanistic models of tree growth and resprouting. Such models are necessary for understanding the response of savanna systems to novel climate regimes (higher temperatures, fewer, more intense storms, and higher soil evaporation rates). The dendrochronological work is being conducted in Kruger National Park and Wits Rural Facility, South Africa.