A known population density was established for northern bobwhites (Colinus virginianus) on a lOO-ha pasture. Line transects were walked twice/day for 5 days to obtain a density estimate. Twelve density estimators using the linetransect data were compared to the known bobwhite density. On our study area, 6 estimators overestimated bobwhite density and 6 of them underestimated density. Because few coveys were flushed, individual transect replicates had to be pooled to provide estimates. This pooling of replicates did not enable us to calculate a mean density and specific variances for each estimator. Variances produced by a Jackknife method appeared to underestimate the true variances. During years of low bobwhite densities (which are common), use of line-transect methods may be inappropriate. Capture-recapture data, collected 1 week prior to running the line transects, underestimated quail density on the pasture.
