Responses of trembling aspen to simulated browsing and application of ungulate saliva

Background/Question/Methods

We tested the separate and combined effects of simulated large mammalian herbivore browsing and saliva application on small aspen (Populus tremuloides) trees. The experiment consisted of a fully-crossed 2 × 2 × 4 design with two levels of artificial browsing (control and clipped), two levels of saliva application (no saliva and saliva), and four aspen genotypes grown in pots in an outdoor common garden with ten replicates per treatment. In June 2008, we simulated large mammal browsing by removing the top 50% of the stem of half of all trees by removal with needle-nosed pliers. For half of the clipped trees we immediately applied deer saliva directly to the wound site. In late August of the same year, we harvested all trees and measured total leaf area, and total leaf, stem, and root biomass. Foliar and stem tissues were analyzed for condensed tannins (CTs), phenolic glycosides (PG), fiber and lignin, and C:N ratio.

Results/Conclusions

We found large differences among genotypes for all physical and chemical variables, except total leaf mass, reflecting the inherent genetic variability of this species. We observed differences among clipped and unclipped trees for all physical measurements, with clipped trees having lower total leaf area, and total leaf, stem, and root biomass. Clipping produced mixed results for plant chemistry. Leaves and stems responded similarly for the PG tremulacin (lower in clipped), and fiber and lignin (higher in clipped), but differently for the PG salicortin (no difference in leaves, less in clipped stems), CTs (less in clipped leaves, no difference in stems), and C:N ratio (no difference in leaves, greater in clipped stems). Deer saliva application, either on wounded or unwounded trees, did not influence any measures of tree biomass. Saliva affected only one of the chemical variables, with the leaves of saliva-treated trees producing slightly higher amounts of CTs. These results demonstrate that the physical and chemical responses of aspen to browsing are due primarily to damage per se, and not to chemical elicitors contained in ungulate saliva. In addition, the differential reaction of genotypes to damage indicates that browsing can be genetically context dependent in aspen.