A POPULATION ESTIMATOR BASED ON NETWORK SAMPLING OF TRACKS IN THE SNOW

We developed a technique to use stratified network sampling to sample animal tracks in the snow and to obtain to population estimates. This method requires sufficient snow conditions to allow animals to leave c ontinuous tracks and a recent snowstorm or windstorm for delineation o f fresh (poststorm) tracks. Additional requirements are that no fresh tracks in serially surveyed sample units are completely missed, that t hese tracks can ire followed to identify all sample units containing t hem, and size of the group that made these tracks can be correctly enu merated. Using this technique, we estimated gray wolf (Canis lupus) po pulation density to be 8.16 +/- 0.91 wolves/1,000 km(2) in a 31,373-km (2) game management unit in Interior Alaska. This sample design also a llowed us to obtain population estimates and confidence intervals for those portions of the Koyukuk and northern Innoko national wildlife re fuges (NWR) within the study area. Using concurrently collected radiot elemetry on 9 wolf packs, we did not detect any violations of assumpti ons.