DOIAbstract
Nearest neighbour methods traditionally used to estimate density of a sessile biological population treat individuals as points. The present paper suggests distance-based density estimators which treat individuals as circles with variable areas. Distribution of distance between a sample point and the k-th (k = 1, 2, 3,...) nearest circle is derived. Maximum likelihood estimator of density is obtained from a random sample of point to k-th order distances. Assuming a skewed distribution for the circle radius, moment estimators of density and mean circle area are derived
Add to ORKG