Are wolves in Wisconsin affecting the biodiversity of understory plant communities via a trophic cascade?

• Main Author: Callan, Ramana
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: uga 2010
• Subjects: Odocoileus viginianus; Canis lupus; Thuja occidentalis; wolves; white-tailed deer; trophic cascade; deer browsing intensity; wolf recovery; seedling recruitment; Great Lakes Region; northern white cedar; Wisconsin; multivariate models; local variables; re
• Online Access: http://hdl.handle.net/10724/26862; http://purl.galileo.usg.edu/uga_etd/callan_ramana_201012_phd

Wolf recovery in the Great Lakes region is anticipated to generate a top-down trophic cascade by altering white-tailed deer density, habitat selection, and/or foraging behavior. Through these direct impacts on deer, wolves are predicted to trigger additional indirect impacts on chronically browsed plant communities. To detect the signal of top-down effects, we performed vegetation surveys in northern white cedar wetlands to measure species richness (S) of understory plants across a gradient of wolf impact. We fit species area-curves of vascular plants grouped by vegetation growth form and duration of wolf occupancy. Sampling at multiple scales revealed that areas consistently occupied by wolf packs were characterized by higher S of forbs at local scales (1-10 m2), and higher S of shrubs at broader scales (10 - 400 m2). Once we detected the signal of a trophic cascade, we further refined our approach by calculating a weighted wolf occupancy duration index (WWODI) based on historic and current wolf pack territory data. We found strong positive correlations between WWODI and the density and diversity of woody stems 50-100 cm tall. Unpalatable species and size classes above the browse line showed no relationship with WWODI. The characteristic gap in the size structure of woody stems, indicative of over-browsed understories, was less apparent in cedar wetlands with high wolf impact. Finally, we tested three possible hypotheses for observed relationships between S and WWODI: (1) a top-down trophic cascade, (2) a bottom-up trophic cascade, and (3) non-trophic associations. Using environmental variables, we created multivariate models of S of forbs, shrubs, tree seedlings and ferns. We used an information theoretic approach to select the best fit models and found that inclusion of WWODI was supported for models of understory plant species known to respond positively to release from herbivory: forbs, shrubs and tree seedlings. When landscape variables associated with wolf habitat selection were used to generate models of plant species richness, these models performed poorly. Evaluated collectively, our results provide little support for either the bottom-up or non-trophic hypotheses. Instead, our results are consistent with wolves triggering a release from browsing pressure by white-tailed deer (a top-down trophic cascade). PhD Daniel B. Warnell School of Forestry and Natural Resources Ecology Nathan Nibbelink Nathan Nibbelink Robert Warren Chris J. Peterson Robert J. Cooper