A Test of Wildlife Habitat Relationship Models for Black Oak and Mixed Conifer Habitats in California

MASTERS THESIS: A survey of the terrestrial vertebrates and their habitat was made in black oak (Quercus kellogii) and mixed conifer-tanoak (Lithocarpus densiflora) forests of the northern Sierra Nevada in the spring and summer of 1980. The data were collected as a baseline assessment of the flora and fauna prior to the conversion of these forests to pure conifer stands in order to learn the importance of hardwood species to wildlife. Additionally, the data provided a basis for testing existing wildlife habitat relationship (WHR) models. Data on the flora and fauna were summarized and comparisons were made between the 2 forest types. The black oak forest was growing on a poor site with a high insolation index relative to the mixed conifer forest. The latter contained approximately twice the vegetative biomass and had a canopy height 2-3 times that of the black oak forest. Differences in vegetation physiognomy were also apparent in foliage height profile curves. Botanical composition was summarized with dominance-diversity curves. Curves for the black oak forest were similar to data from other deciduous forests containing many codominant plant species. More dominant and more rare species were recorded int he mixed conifer forest. Using indices of relative abundance, the importance of bird and mammal species were also displayed with dominance-diversity curves. The diversities of the bird communities in the 2 forest types were nearly equal, but the diversity of the mammal community in the mixed conifer habitat was higher. Based upon physiognomy and species composition, 80 survey plots were analyzed using cluster analysis and principal coordinate analysis to reveal plot similarities and to understand species distributions. A consensus of 2 resemblance function matrices provided the most realistic determination of plot relationships. A WHR model for the western Sierra Nevada was tested. First, a methodology is presented for the determination of habitat classification efficiency. An information theoretic optimality criterion was used to rate the efficencies of 4 habitat classifications for summarizing the recorded distribution of 17 mammal species. The habitat classification used in the WHR model was the most efficient alternative for predicting the distribution of mammals. The method permits an objective means for choosing the most cost effective habitat classification. Second, the wildlife species occurrence data were used to test the validity of the WHR matrix. The WHR model gives a better than chance prediction of habitat suitability for the mixed conifer and black oak habitats sampled in this study. Errors in the model were noted for 18 species recorded in habitats predicted as unsuitable. The applications of wildlife survey data collected by standardized techniques on many, relatively small sample plots are numerous. They include: 1. geographic mappping, 2. monitoring trends, 3. experimental analysis of impacts, 4. validation of existing WHR models, and 5. detection of wildlife habitat relationships useful in constructing second generation models. Theis study provides detailed illustrations of several of these applications.