Canopy Herbivory and Soil Processes in a Temperate and Tropical Forest
Principal investigators:
Margaret D. Lowman, Selby Botanical Gardens
Mark D. Hunter, University of Georgia
Timothy D. Schowalter, Oregon State University
Graduate students:
Chris Frost, University of Georgia
Bruce Rinker, Selby Botanical Gardens
student research
Insect herbivores in forest canopies have the potential to influence soil processes (decomposition, respiration, nutrient availability) by introducing materials from the canopy to the forest floor. Laboratory work has shown that the products of defoliation can influence soil respiration and nutrient cycling in incubator experiments. We believe that it is important to assess whether natural defoliation under field conditions can also modify soil processes significantly. We are exploring this by monitoring natural levels of herbivory in one temperate and one tropical forest (LTER sites: Coweeta Hydrological Laboratory, North Carolina, and Luquillo Experimental Forest, Puerto Rico), and experimental manipulation of the products of herbivory.
Insect herbivory in forest canopies introduces materials into the soil community through two major pathways. First, defoliators introduce frass (insect feces), greenfall (fragmented leaf tissue dropped during herbivory) and leaves abscised prematurely onto the forest floor. Second, throughfall, which is rainwater modified by its passage through the forest canopy, is altered by the combination of dissolved frass and modified leachate from damaged leaves. These pathways combine to introduce increased amounts of carbon, nitrogen and phosphorus into the soil microbial community. We predict that increased activity in the microbial community following these herbivore derived inputs will result in increased rates of leaf litter decomposition and forest floor respiration.
We are measuring the accumulation of herbivores and herbivory, and the release of materials by herbivores into forest floor communities at Coweeta and Luquillo. Soil respiration, leaf litter decomposition, and C, N, and P dynamics during decomposition are also being measured in conjunction with estimates of the seasonal abundances of nematodes and soil arthropods. In replicated field experiments, we are excluding or augmenting frass-fall, green-fall, throughfall, and premature leaf abscission to test the hypothesis that these herbivore-derived inputs can influence soil processes.
Finally, we will compare the relative importance of herbivore-derived inputs from the canopy and other biotic/abiotic forces on soil processes. It is in this area that the opportunities for quantitative analyses and modeling are greatest. We wish to incorporate our results into models of soil nutrient dynamics and respiration to understand the roles of herbivores in ecosystem processes, and to compare among systems. A quantitatively-oriented graduate student would help to generate such models and to compare their outputs with herbivore-free models.
Student Research:
Canopy Herbivory and Soil Processes in a Temperate and Tropical Forest
Chris Frost
My work focuses on effects of herbivory on ecosystem processes in a temperate forest. In 1998, an herbivore outbreak on one site at the Coweeta Hydrologic Laboratory removed upwards of 50% of the canopy leaf area and deposited a large amount of frass (insect feces) on the forest floor. One month later, there was a "pulse" in nitrate concentrations in the soil underneath the outbreak area and in the stream draining the outbreak area. As with other somewhat unexpected results, we scratched our heads and wondered if small insect larvae could be responsible for an ecosystem-wide disturbance, and then designed experiments to test the hypothesis that they just might.
I am working with northern red oak (Quercus rubra) as the model tree, which is a dominant species at Coweeta. The white marked tussock moth (Orgyia leucostigma: Lepidoptera; Lymantriidae), reared in the laboratory, is the model herbivore. Using a series of outdoor
plant-soil replicated mini-ecosystems, we will attempt to reproduce the conditions under which the soil and stream pulses occurred to investigate the mechanisms underlying the phenomenon.
While short-term, factorial experiments will help explain potential mechanisms, we are also interested in searching for patterns in nutrient cycling as a result of herbivory. Quantitative modeling will help this search. We are using data collected from Coweeta and the manipulative experiments to parameterize a model focusing on inputs of canopy herbivory to forest soil systems. There are two driving questions, both of which are difficult to answer empirically. First, do herbivore inputs really matter at the watershed level over a period of decades? Second, do herbivore-mediated changes in nutrient availability affect the population dynamics of the herbivore at the watershed level over the course of decades or longer?
top
Overview