• Version
• Download 5
• File Size 358.51 KB
• Create Date 1 January 1998
• Download

The average age of first reproduction appears to be a useful index of habitat quality for bear populations. However, the conventional method for calculating this age, using only bears whose first litters were observed, gives a low-biased estimate. This bias is a consequence of losing bears approaching their cub-producing years due to death, lost radio contact, or removal of radiocollar. Late-maturing bears are more likely than early-maturing bears to be lost before they produce a litter; thus, if they are not included in the sample, the estimate becomes biased low. We propose a method that generates virtually unbiased estimates of the mean and median ages of primiparity using data from bears that did not produce cubs as well as from those that did. First, we calculate the proportion of females that produced a first litter among all nulliparous females monitored at each age; these proportions are weighted by the proportion of females in the population that are available to have a first litter at each age. This procedure is analogous to the estimation of survivorship based on radio-days of monitoring. We tested our method using data from radiocollared black bears (Ursus americanus) in Minnesota and Massachusetts and found that the conventional method underestimated both the mean and median ages of first reproduction by 0.2-0.5 years. Our approach produced an estimate of the expected age of first reproduction free of the effects of study-related shortcomings and independent of mortality. This estimate is more appropriate for cross-population comparisons, as well as for population modeling.