| Summary: | Thesis (M.S.)--Humboldt State University, Wildlife Management, 1995 Various devices have been used to detect fishers (Martes pennanti) and martens (Martes americana) in California; however standardized methodologies that allow comparison between surveys or the potential to monitor changes in populations over time are lacking. I developed methods to detect fishers and martens and established a standardized survey protocol. At two locations in the field, martens were detected and the accuracy of the methodology for martens was determined. Smoked track-plates and triggered-cameras were selected for field tests based on their performance with captive fishers. Significantly (< 0.01) more marten detections occurred at the track-plates than at the triggered-cameras. Chicken bait was more effective (< 0.05) than tuna cat-food in eliciting responses by martens at detection devices. The minimum length of the survey period to adequately determine presence of martens in an area was 22 days. The accuracy of the survey protocol was examined by placing detection devices in known marten home ranges and determining the probability of detection. The probability was determined by examining whether a device would detect a marten when the device was placed in the marten's home range. There was a marked difference in the probabilities of detection both between individual martens and between study sites. On the Tahoe National Forest site during the winter, the probabilities of detection ranged from 0.4 to 1 (Ye = 0.67, SE = 0.10). On the Lassen National Forest site during the late summer, individual probabilities ranged from 0 to 0.5 (Ye = 0.19, SE = 0.12). Variation in the probabilities of detection may be attributed to multiple factors such as individual variation, gender, home range size and quality, and/or season.
|
|---|