Introduction
Isle Royale National Park, the largest island in Lake Superior, provides biologists with a fairly unique system for studying the interactions between different trophic levels. Isle Royale has a rather simple food chain consisting of producers and a single large herbivore that in turn has only a single predator, the gray wolf (Canis lupus). The island had a rather large abundance of balsam fir (Abies balsamea) until the park was colonized by moose (Alces alces) that swam to the island in the early 1900s. After the establishment of this large herbivore, the balsam fir declined from 46% of the overstory in the 19th century to about 5% today. Nearby islands that are inaccessible to moose continue to have a large fir component in their forests; thus the decline of the fir on Isle Royale has been attributed to moose herbivory. Balsam fir is not considered optimal forage for moose but it can comprise up to 59% of their winter diet.

Over the last several decades, significant temporal fluctuations have been observed in the densities of the wolf and moose populations and the growth rates of balsam firs. Two hypotheses have been suggested to account for these fluctuations. The primary productivity or "bottom up" hypothesis suggests that plant growth is limited by the energy available to plants which is determined in turn by temperature and precipitation. Additional plant growth means more forage is available--thus herbivores, and ultimately carnivores, should increase in abundance. Alternatively, the trophic cascade or "top down" model predicts that changes in one trophic level are caused by opposite changes in the trophic level immediately above it. For example, a decrease in moose abundance should produce increased plant growth if moose herbivory limits plant growth. Changes in primary productivity would only have a discernible effect on vegetation if higher level interactions had been removed.

The Isle Royale ecosystem provides us with a good opportunity to test the predictions of these alternative hypotheses. Longitudinal data are available for each of the key variables, including annual plant growth, herbivore density, and carnivore density. The historical growth rates of balsam fir have been determined through tree-ring analysis. When herbivores remove large quantities of the foliar biomass, annual wood accrual decreases and ring widths are reduced. Thus tree ring data allow us to estimate the intensity of herbivory over time. Moose and wolf populations have been censused for decades on Isle Royale, providing us with annual estimates of herbivore and carnivore densities. Long-term records are available for each trophic level in the Isle Royale ecosystem, providing the necessary data to evaluate both hypotheses.

GO TO Handout #2: The moose, the wolf and the fir tree
GO TO Handout #3: The moose, the wolf and the fir tree

Image Credits:  Wolf graphic courtesy of NAWA (North American Wolf Association).  Moose graphic adapted from Adobe® Image Gallery's "Animal Life #ALI_025"

Last Updated:  10/21/02 nas
This file is also available in Adobe Portable Document Format (PDF).