Estimation of Population Age Structure, Detection Probability, and Density of Red Tree Voles in Oregon

Graduation date: 2017Red tree voles (Arborimus longicaudus) are a unique species of Arvicoline rodent found in western Oregon and northwest California. Their preferred habitat is mature and old-growth coniferous forest dominated by Douglas-fir (Pseudotsuga menziesii). Tree voles are an important prey item for northern spotted owls and many other predators, but are extremely difficult to sample. As a result, little is known about their abundance or population dynamics. We used several methods to describe tree vole detectability, density, and population age structure. We verified a relationship between molar tooth development and age in red tree voles, allowing us to estimate the age of voles recovered in pellets regurgitated by spotted owls in Oregon. We found that the age distribution of tree voles in spotted owl pellets varied from year to year but not from region to region within the species range. We found evidence for a positive relationship between annual precipitation and the proportion of juvenile tree voles in owl pellets on one of two study areas.\ud We conducted modified line-transect surveys at 31 locations representing three forest stand age classes (young, mature, old) in the Oregon Coast Range to visually detect potential red tree vole nests. We climbed trees where potential nests were detected from the ground in order to determine the species and the occupancy status of nests. We used the transect and climbing data and distance sampling techniques to estimate detection probabilities and density of active tree vole nest trees in relation to forest age class. Our\ud surveys suggested that density of tree vole nest trees increased with forest age, whereas detection probability of nest trees decreased with forest age. Regardless of forest age, mean detection probability of nest trees was low (< 30%). We used estimates of nest tree density on our study area to estimate the minimum density of adult tree voles (~1.91/ha) and the density of individual tree vole home ranges (~4.2/ha). These estimates of detection probability and density suggest that only a small proportion of nest trees are detected and buffered against disturbance using methods described in the survey protocol used by federal agencies in Oregon and California. We recommend methods to improve detection rates of nest trees, including application of our density estimates to occupied stands, modification of stand-specific data using our estimates of detection probability, performing full-tree nest searches in every tree climbed, performing a 100 m search radius around confirmed nest trees, and targeting larger-than-average trees (with respect to stand-specific data) for spot searches