Assessing impacts of large herbivores on shrubs: tests of scaling factors for utilization rates from shoot-level measurements

1. Accurate methods for estimating the intensity of browsing by herbivores are fundamental to understanding the ecology of shrub communities. Quantifying browse utilization on shrubs at large scales is difficult because shrubs have complex, spatially variable growth forms. Most existing methods estimate browsing rate at the scale of linear current-year shoots or twigs. How such fine-scale estimates relate to the proportion of current-year growth consumed from whole plants or plots is often unknown. The relationship is likely to be complex because herbivores selectively browse more productive plants and plant parts. 2. Using a clipping experiment designed to mimic elk Cervus elaphus browsing, we quantified how utilization estimates at the scale of individual current-year shoots of two willow species, Salix bebbiana and Salix geyeriana, relate to actual mass removed at the scale of rooted stems. Three approaches to scaling were examined: (i) taking an average, (ii) multiplying by the proportion of shoots clipped and (iii) multiplying by a novel scaling factor that weights utilization by productivity. To address how to scale-up from stems to plots, we applied the most accurate stem-level method to elk-browsed willow and compared plot-level estimates by two scaling approaches. 3. In scaling from shoots to stems, the novel scaling factor was most successful and resulted in accurate estimates for up to c. 45% of current annual growth clipped. In scaling from the stem to the plot, elk preference for more productive stems caused a simple average of stem-level utilization to differ from a productivity-weighted average by 15%. 4. Synthesis and applications. In order to reflect accurately the proportion of biomass consumed at a whole-plant level, fine-scale estimates of utilization should be weighted by an estimate of pre-browse productivity, as this is mathematically equivalent to summing pre-browse and post-browse mass before calculating the proportion consumed. In developing methods to estimate utilization at plot scales, an important consideration is the choice of sampling unit, which should be both amenable to unbiased sampling and tractable in terms of measuring productivity.