Comparison of population estimators and indices for monitoring house mice in sorghum crops.

House mice (Mus musculus domesticus Schwarz & Schwarz, 1943) are monitored in Australia and China to track changes in mouse population densities and forecast their potential damage to cereal crops. The present study compared population indices based on the number of different mice caught and overall trap success from live-trapping with an oil card index (OC) and a tracking index (T) for monitoring mice in sorghum crops immediately before crop maturation. T was measured as the percentage of track board covered with mouse footprints night(-1), and OC as the percentage of card removed by mice night(-1). The reliability of these abundance indices was quantified by Pearson correlation coefficient with the trappable population size (Ñ), which was estimated by capture-recapture over eight consecutive nights on 175 × 5 trapping grids, in sorghum crops on two properties on the Darling Downs, Queensland. Because of differences among individual mice in capture probability, Model M(h) of program MARK was used to account for such heterogeneity and to estimate the size of each mouse population. The number of individual animals caught was more strongly correlated with Ñ than trap success and, therefore, might be a more reliable index; the data suggest that three trapping occasions provide optimal precision for this index. T correlated significantly with Ñ only at sites where the canopy of sorghum plants was closed, and its use should, therefore, be restricted to this habitat. OC did not correlate with Ñ because none or very little of the cards was eaten at low to moderate mouse densities. T and the number of animals caught over three trapping nights are recommended for monitoring mice in sorghum crops immediately prior to crop maturation.