A density estimate of sika deer using distance sampling techniques in forested habitat(In English)

摘要

Six night spotlight counts were conducted in September and October 1999 in an area of high density of sika deer Cervus nippon in Nikko National Park, Japan. Using a vehicle and laser range-finder, perpendicular distances of deer cluster were measured, and cluster sizes were recorded. Using distance sampling technique (DST), deer density was estimated to be 125.8 deer/km² (95% CI 100.7 - 157.3), and detection probability of sika deer in the 80 m wide line transect on both sides was 52.5%. By comparing the density estimate result of DST to that of the past naive methods including King's, Hayne's, Leopold's, Gates's I, Gates's Ⅱ , Gates's Ⅲ , and Frye's methods, it was found these latter methods over-estimated the deer density in our study area as high as 1.76 (Frye’s method) to 6.10 (Hayne's method) times the actual density. Thus, we suggest that DST should be used to estimate animal density and that the naive methods should be avoided; we also suggest that comparison of counted deer numbers obtained with the line transect method between different areas or seasons might be biased, and that the DST should be adopted for estimating animal density or abundance [ Acta Zoologica Sinica 50 (1): 27 - 31 , 2004].