Multiscale analysis of habitat, vegetation change, and streamflow as ecological factors affecting population dynamics of Rana chiricahuensis

Amphibian declines are likely caused by multiple synergistic effects. Studies have recently focused on identifying population stressors including habitat modification, pollution, and drought. Determining how landscape change affects stressors provides a foundation to study, manage, and monitor species and communities. Chiricahua leopard frog (Rana chiricahuensis ) was used as a model species to identify habitat, quantify changes to habitat, and identify potential stressor thresholds on populations. A literature-derived deductive model, historical-survey inductive model, and a museum-record inductive model were created and compared. The historical survey model and literature model consistently performed better than the museum record model and a random dataset. The historical survey model had higher overall accuracy and specificity whereas the literature model had higher sensitivity, Kappa statistic values, and predicted fewest hectares of suitable habitat. Two new concepts were introduced including merging inductive and deductive modeling approaches to provide a weighted suitable habitat model and a distance to model metric to provide a fuzzy assessment using existing species occurrence records. Inductive-deductive models provide identification of potential primary habitat and potential marginal or corridor habitat. The distance to model metric provides a method to identify model confidence. Three vegetation maps were created using Landsat imagery from 1979, 1989, and 2000. Change detection identified declines in rangeland (-12%) and woodland (-10%) and increases in forest (24%) from 1979 to 2000. Two USGS daily streamflow datasets from different watersheds were analyzed for changes in base flow, peak flow, and instantaneous peak flow. Changes included increased average flow and more floods in both watersheds. Base and peak flow declined in the smaller watershed but remained within range of variability whereas the larger watershed base flow and peak flow increased. Sensitivity of a R. chiricahuensis population to effects from peak flows, base flow, and Chytrid fungus was determined. Negative Chytrid fungus effects larger than 10% caused declines in populations in conjunction with stochastic base and peak flows, implicating Chytrid fungus as the proximate cause of many amphibian declines, but the added effects of base and peak flow can hasten declines. Any combined chronic effects greater than 20% caused long term population declines.