Response to Comment: Ungulate Herbivory on Willows on Yellowstone's Northern Winter Range

Abstract

We concur with Wagner et al (1995), that ungulate browsing has an import effect on growth form, productivity, and survival of willows in northern Yellowstone National Park (hereafter YNP). We concur with Wagner et al. (1995) that elevation, precipitation, willow height, aboveground biomass production, and concentrations of secondary compounds covary in willow shoots on Yellowstone's northern ungulate winter range. We do not imply cause and effect between the climate change in northern YNP this century and the willow decline, nor do we reject the overabundant elk hypothesis (Singer et al. 1994). We present the case for a more complex, multi-causal situation, where interactions between elk abundance, climate change, fire, mammalian predators, and beaver abundance, might all influence the status and recruitment of willows. Singer et al. (1994) was intended as a paper to suggest a set of hypotheses that might explain the willow decline in YNP. We use cautionary phrases like may, likely, proposed, apparently, speculate, and circumstantial 27 times in Singer et al. (1995). We disagree with Wagner et al. (1995) that elk densities in the area are inversely correlated with elevation (Coughenour and Singer 1995) and elk consume mostly a grass-sedge diet (86% of the elk diet) and few willows (<O.1I% of diet, Singer and Norland 1994). Furthermore, ungulate browsing intensity is not inversely correlated with either elevation or precipitation. Biomass removed by ungulates during winter from tall willow stands, mostly by moose, exceeds that removed from suppressed willow stands 9fold, and that removed from intermediate stands by nearly 7 foldduring the winter. Ungulate browsing consumption (kg/ha removed), is far greater in tall versus suppressed willows, the reverse of that contended by Wagner et al. (1995). Mean percent of current annual growth (CAG) removed 1988-93, averaged only slightly greater (31%) in suppressed stands, compared to tall willows (22%), and intermediate willows (26%). These data indicate that suppressed willows at low elevations are at a serious disadvantage, in terms of growth potential, to cope with ungulate herbivory. Wagner et al. (1995) state that woody plant species on the northern winter range have declined due to elk herbivory, and elk in northern YNP have been reduced to eating the least palatable forages, for example spruce (Picea engelmannii) and other conifers. The field evidence does not support their statement. No forages differed in elk diets compared between 1967-70 and 1986-88, a period when all northern range ungulates approximately tripled following cessation of artificial controls, and when the elk increase was 2.6 fold (Singer and Norland 1994). In fact, elk used aspen and conifer types significantly less during the recent period, and there was a tendency for less conifers in the elk diet during the recent period. Many upland woody shrubs have mostly increased, not decreased, over the past 5 decades (Singer and Renkin 1995). During this past century, until the large fires of 1988, there was a pattern of widespread conifer encroachment onto grasslands (Houston 1982), not a decline (Wagner et al. 1995). This appears to be largely due to fire suppression (Houston 1973, 1982). It is grass forage that regulates elk, not willows and aspen (Houston 1982, Coughenour and Singer 1995). An already marginally-distributed shrub type--suppressed willows--has become highly palatable to elk whose numbers are regulated by something else. Wagner et al. (1995) disregard the fact that the grasslands of northern YNP, which are very important to elk regulation, are in good to excellent condition (Coughenour 1991; Frank and McNaughton 1992, 1993; Wallace and Macko 1993; Merrill et al. 1994; Coughenour et al. 1995; Reardon 1995; Lane and Montagne 1995; Singer 1995; Singer and Harter 1995), and aboveground production of grasses was stimulated by ungulate herbivory (Frank and McNaughton 1992, 1993), suggesting elk have not depleted their preferred food resources. Wagner et al. (1995) question the significance of the climate change in YNP this century, and in doing so, they ignore a substantial volume of recent research and published literature. A climate change to increasing aridity in northern YNP this past century is well documented; summer temperatures increased 0.87?C, and January-June precipitation decreased 61 mm (Balling et al. 1992a, 1992b). Fewer large floods have occurred during the last century, with the last large flood in 1918, and the last period of frequent floods in the 1870's (Meyer et al. 1992; Bingham and Meyer 1994) also due, in part, to a drier climate. Aging of aspen in northern YNP also verified a wetter period from 1870-90. All large canopy aspen stems sampled were generated during the wetter period of 1870-90 (Romme et al. 1995). Considerable field evidence suggests drier conditions on sites occupied by suppressed willows. Willows are now co-dominant with dryland shrubs such as mountain big sagebrush (A. tridentata vaseyana) on many sites across the low and mid-elevations of the winter range. Almost no recruitment of willows is observed at