ESTIMATION OF DENSITY OF MONGOOSES WITH CAPTURE-RECAPTURE AND DISTANCE SAMPLING

We captured mongooses (Herpestes javanicus) in live traps arranged in trapping webs in Antigua, West Indies, and used capture-recapture and distance sampling to estimate density. Distance estimation and program DISTANCE were used to provide estimates of density from the trapping- web data. Mean density based on trapping webs was 9.5 mongooses/ha (ra nge, 5.9-10.2/ha); estimates had coefficients of variation ranging fro m 29.82-31.58% ((X) over bar = 30.46%). Mark-recapture models were use d to estimate abundance, which was converted to density using estimate s of effective trap area. Tests of model assumptions provided by CAPTU RE indicated pronounced heterogeneity in capture probabilities and som e indication of behavioral response and variation over time. Mean esti mated density was 1.80 mongooses/ha (range, 1.37-2.15/ha) with estimat ed coefficients of variation of 4.68-11.92% ((X) over bar = 7.46%). Es timates of density based on mark-recapture data depended heavily on as sumptions about animal home ranges; variances of densities also may be underestimated, leading to unrealistically narrow confidence interval s. Estimates based on trap webs require fewer assumptions, and estimat ed variances may be a more realistic representation of sampling variat ion. Because trap webs are established easily and provide adequate dat a for estimation in a few sample occasions, the method should be effic ient and reliable for estimating densities of mongooses.