Singer et al. (1998) propose that the decline in populations of beaver (Castor canadensis) in Yellowstone National Park (YNP) has caused willow to be more vulnerable to browsing by clk (Alces alces). I do not believe that their scenario correctly characterizes the relationship between elk and willow in YNP

The authors developed their hypothesis based on 2 sets of observations. One was an experiment that compared willow growth in YNP to that in Rocky Mountain National Park (RMNP). Current annual growth was clipped from plants at 0%, 50%, and 100% levels in each of 4 years (1992–1995). From RMNP plants in 1 exclosure were used (Beaver Meadows); in YNP plants in 3 exclosures were treatcd (Junction Butte, Lamar East, and Lamar West; L.C. Zeigenfuss, personal communication). A second set of observations, which included additional sites in RMNP measured the growth and stature of browsed and unbrowsed plants.

Singer et al. (1998) reported response to the clipping experiment in their Table 5. Even under the most severe clipping treatments, willow height and annual production were maintained in RMNP willows but declincd in YNP willows. Willows in RMNP responded to the 50% clipping treatment by increasing the level of chemical defenses (tannins and phenolics), whereas the chemical defenses of YNP willows remained relatively constant. The authors surmised that 1) enhanced vigor may enable a plant's terminal leader to grow out of ungulates' reach and 2) increased production of chemical defenses may deter herbivory.

Singer et al. (1998) concluded that the betweenpark differences were directly related to better growing conditions in RMNP compared to YNP The better growing conditions in RMNP were attributed to: 1) higher effective precipitation, 2) more beaver activity, 3) more beaver dams in drainages, and 4) higher water tables near streamsides. There are several reasons the experiment conducted by Singer et al. (1998) does not support these conclusions.