Corinna Riginos
|
 |
Corinna Riginos
|
|
My dissertation research focuses on several aspects of tree-understory interactions in a savanna system. I am working at the Mpala Research Centre in the Laikipia District of central Kenya. This area, as with many parts of sub-Saharan Africa, is used by both wild and domestic herbivores. Wild herbivores inlcude elephants, girraffes, zebras, Grant's gazelles, elands, hartebeest, buffalos, steinbucks, and oryx. Domestic herbivores are primarily cattle.
Heavy or prolonged cattle grazing has been widely observed to cause 'bush encroachment,' or an increase in woody vegetation at the expense of the grass layer, in savanna systems. This is thought to occur because of a shift in the competitive balance between trees and grasses. However, although there have been numerous empirical and model-based studies of tree-grass interactions, several important aspects of these interactions have been overlooked. My research examines the effects of different guilds of herbivores, tree size, and tree density on tree-grass interactions.
My primary field assistants, Frederick Erii, Jackson Ekadeli, and John Lochikuya (from left to right).
Although cattle grazing has been blamed widely for altering the competitive balance between trees and grasses, this assertion has not previously been tested with direct experimentation. In African savannas, cattle have partially or wholly replaced a diverse set of wild herbivores, including both grazers and browsers of a wide range of body sizes. As a result, grazing by cattle may cause a release in trees growth via two mechanisms: (1) by releasing trees from browsing pressure (direct effect), and/or (2) by releasing trees from competition with grasses (indirect effect). If this is the case, tree-grass interactions should be different under a cattle grazing regime than under a wild herbivore regime.
Killing grass around a small Acacia drepanolobium tree. To test these hypotheses, I am conducting reciprocal tree and grass removals within Truman Young's Kenya Long-term Exclusion Experiment (KLEE) exclosures. I am comparing the growth and survival of manipulated and control grasses and trees of the dominant species, Acacia drepanolobium, within four herbivore regimes: cattle only, wildlife only, cattle and wildlife together, and no large herbivores. This design allows me to examine the relative strength and direction of tree-grass interactions under different herbivore regimes.
Most simulation models of tree-grass dynamics have assumed that all established trees have the same interaction with the grass layer. Interactions, however, may vary substantially with tree size. For example, small trees may be competitively suppressed by grasses, as has generally been assumed for tree seedlings, whereas larger trees may not be affected by grass competition. Meanwhile, large trees have often been demonstrated to provide shade, reduced evapotranspiration, and elevated soil nutrients — to the benefit of the understory layer. Small trees are less likely to have a canopy large enough or old enough to provide these benefits. On the other hand, smaller trees with branches low to the ground can protect grasses from herbivory, whereas it is not likely that taller trees do this as effectively.
Grass cleared around medium and small Acacia drepanolobium trees.
These are just some examples of possible ways that size could affect the strength or even the direction of tree-grass interactions. I am probing some of these possibilities by carrying out reciprocal tree and grass removals for trees of three size classes.
Up to now, I have been discussing the interactions between trees and grasses on the scale of the individual tree. However, the density of trees in a stand may be equally or more important to grass productivity than the properties of the nearest neighbor tree. Areas with a high density of trees have generally been shown to have lower grass cover or productivity. But many questions remain. What happens to tree growth as tree density increases? Does intraspecific competition become more limiting than interspecific competition? What limits grass growth in high density stands — direct competition with trees, or light-limitation, or some other mechanism? Finally, there may be interactions among tree density, herbivores, and tree-grass interactions. For example, herbivores may avoid dense stands and perfer more open stands. What repercussions might this have for tree growth/survival and tree-grass interactions?
My preliminary results from a large descriptive survey show that wild herbivores avoid areas with more trees. Moreover, I have found that certain grass species become more or less abundant as tree density increases.
Cutting trees to clear or thin stands of Acacia drepanolobium.
To address my study questions experimentally, I have manipulated tree density in replicate plots (60 x 60 m). The treatments include thinned, completely cleared, and control plots. I am examining tree-grass interactions and herbivore usage within those plots.
In collaboration with Jake Goheen, a Ph.D. student at the University of New Mexico, I am looking at the suite of possible factors that may limit seedling establishment in Acacia drepanolobium. These include: drought stress, herbivory by small mammals, herbivory by large mammals, and competition with grasses. These factors likely interact in complex ways. For instance, large herbivores depress grass cover as well as small mammal abundance. Does this enhance seedling survival? Or, does high grass cover provide essential shade or concealment from herbivores during the early stages of a seedling's life? We are addressing these questions with a series of seedling establishment experiments.
