Growth and chemical responses of trembling aspen to simulated browsing and ungulate saliva

Aims Woody plant-browser systems represent an understudied facet of herbivory. We subjected four genotypes of trembling aspen to artificial browsing, similar to that of a large mammalian herbivore, and applied deer saliva to clipped and unclipped trees to assess: (i) the effects of artificial browsing on aspen growth and phytochemistry of leaves and stems, (ii) genotypic variation in responses and (iii) potential alterations of responses by mammalian saliva. Methods Potted aspen trees were grown outdoors on the University of Wisconsin-Madison campus. The experiment consisted of a fullycrossed, 2 × 2 × 4 randomized complete block design, with two levels of artificial browsing (unclipped and clipped), two levels of saliva application (no saliva and saliva) and four aspen genotypes. To simulate ungulate browsing damage, we removed the upper 50% of the stem of half of the trees by pinching the stem with needle-nosed pliers and then separating it by tearing. For half of the damaged trees, we immediately swabbed the wound with deer saliva. Trees in the unclipped plus saliva treatment were swabbed with saliva at the 50% height mark. To assess the effects of clipping and saliva application, we harvested all trees after 2 months and measured various growth and chemical properties. Growth measurements included height, vertical growth, mass of leaves, stems and roots, leaf number and area and bud set. Chemical parameters included defensive, nutritional and structural components of both foliage and stems. Important Findings Clipping affected most of the growth parameters measured, decreasing tree height, leaf, stem, root and total tree mass and leaf area. Clipped trees had greater vertical growth, more leaves and higher specific leaf area (SLA) than unclipped trees. Deer saliva had little to no effect on plant growth response to the clipping treatment. Terminal budset was delayed by clipping and varied among genotypes but not in response to saliva application. Clipping also affected most of the phytochemical variables measured, reducing defensive compounds (phenolic glycosides and condensed tannins (CTs)) and nutrients (N), but increasing structural components (cellulose and lignin) in both leaves and stems. Saliva had very little effect on tree chemistry, causing only a slight decrease in the amount of CTs in leaves. In general, leaves contained more defensive compounds and nutrients, but much less cellulose, compared with stems. Genotypes differed for all physical and chemical indices, and in tolerance to damage as measured by vertical growth. In addition, for most of the physical and chemical variables measured, genotype interacted with the clipping treatment, suggesting that in natural stands some genotypes will resist or tolerate browsing better than others, affecting forest genetic composition and ultimately forest dynamics.