The effects of flow regulation on the population biology and ecology of beavers in northwestern Colorado

Abstract

There are approximately 75,000 dams on the streams and rivers of the United States, with nearly all the large rivers having been severely altered. The downstream effect of a dam often includes the elimination of the flood pulse, which plays an important role in the creation and maintenance of riparian ecosystems, some of the most ecologically valuable ecosystems particularly in the western U.S. Such large-scale manipulations of a key disturbance are responsible for numerous ecological changes. My research assessed the effects of flow regulation on beaver (Castor canadensis) populations and on the interaction between beavers and two important plant species, sandbar willow (Salix exigua) and Fremont cottonwood (Populus deltoides spp. wislizenii). I gathered comparative data on two 6th-order alluvial rivers, the free-flowing Yampa River and flow-regulated Green River in Northwest Colorado from 1997-2000. In these systems beavers are bank dwelling and cottonwood and willow are the dominant native woody species. I assessed the effects of flow regulation on the demography of beavers by comparing the density, survival, home range and body size of bank-dwelling beavers. Flow regulation on the Green River has altered fluvial geomorphic processes, influencing availability of willow and cottonwood that has, in turn, influenced the demography of beavers. Beaver density was higher on the Green River (0.5-0.6 colonies/km of river) than the Yampa River (0.35 colonies/km of river). Adult and sub-adult beavers were in better condition on the Green River as indicated by their larger body mass and tail size. Evidence suggested mortality due to predation was higher for adult beavers on the Yampa River. There was no detectable difference in home range size though there were areas on the Yampa River that no beavers used. I attribute the improved habitat quality on the Green River to a greater availability of willow. I suggest that the sandy flats and bars that form during base flows on the Yampa River increase the energy expended to attain food and increase the predation risk and thus lower the availability of woody forage. I quantified the abundance and density of sandbar willow and the amount of willow taken by beavers on alluvial sections of the flow-regulated Green River and free-flowing Yampa Rivers in northwest Colorado. I used beaver exclosures to determine differences in growth rate of willow between rivers. On the Green River, willow has shifted its distribution from a primarily bank-oriented to a primarily island-oriented species. The total area of willow was similar on each river, though on the Green River patches were smaller, more numerous, contained a lower density of stems, and stems grew faster. Beavers selected stems that were larger in diameter and closer to the river and cut twice as much willow on the Green River as on the Yampa River. This resulted in an 8-fold difference in the percent of stems removed on the Green River because of the lower density of willow stems. The difference in rate of herbivory between rivers can be attributed to a higher availability of stems, a higher density of beavers, and less young cottonwood (an alternative food source) on the Green River. Flow regulation increased the availability of willow by altering its distribution so that patches were adjacent to water year around and had a greater amount of willow perimeter interfacing with water on the Green River than on the Yampa River. This provides safer foraging and involves less energy expenditure. I studied the impact of beaver herbivory on young cottonwood tree survival by measuring the abundance and density of cottonwood on a 10-km stretch of the Green River and a 7.5-km stretch of the Yampa River. I followed the fates of individually marked young saplings in 3 patches of cottonwood on the Yampa and 2 on the Green River. I measured the diameter and distance from water of each monitored tree. From May 1998 through November 1999 two natural floods occurred on the Yampa River and one controlled flood was released on the Green River, which allowed us to determine the effect of flooding on beaver herbivory. Flow regulation has caused a decrease in number of cottonwood patches per km of river, area of patches per km, and average stem density within cottonwood patches on the Green River. The number of stems cut per beaver was 3 times lower on the Green River than on the Yampa River but cottonwood survival was still lower on the Green River because of the lower abundance of cottonwood there. Factors that influenced beaver herbivory included the abundance of cottonwood, the distance of cottonwood patches from the river and controlled flooding. Controlled flooding appeared to increase the rate of foraging on the Green River only during the flood in 1999 which was of greater magnitude and duration than the flood in 1998. I suggest regulation has magnified beaver's impact on cottonwood through interrelated effects on plant spatial distribution and cottonwood abundance. I demonstrated how interaction strength of beavers varies spatially depending on environmental context. I first qualitatively compared interaction strength of beavers in smaller streams where they are able to build dams, with beavers in larger rivers where they are bank dwelling and unable to build dams. Second, I quantified interaction strength of beavers on two comparable 6th-order rivers, the Green and Yampa in northwest Colorado, for two community traits (cottonwood survival and percent of willow removed). Interaction strength of beavers was context dependent, decreasing considerably from small ordered streams (2nd-4th) to large ordered rivers (5th-7lh), primarily because they are unable to build dams in larger-ordered rivers. Beaver interaction strength also varied within comparable 6th-order rivers, being much greater on the flow-regulated Green River than the tree-flowing Yampa River. I attribute variation in interaction strength to the amount and variability of hydrologic energy associated with the flow regime. Small ordered streams contain less energy enabling beavers to build dams, which increases their interaction strength, though variation in flows can reduce interaction strength by temporarily destroying dams. On the Yampa River variation in hydrologic energy lowered interaction strength of beavers by maintaining a high abundance of cottonwood trees and increasing the expenditure of energy and risk of predation for beavers. On the Green River variability in hydrologic energy has been eliminated through flow regulation, which has stabilized the system and allowed beavers to have greater interaction strength. Classifying beavers as keystone species is justified on 2nd-4th ordered rivers but not on larger ordered rivers.