Prediction of litter size in American black bears

Publication Date


Journal or Book Title

URSUS, VOL 12 (2001)


: American black bear (Ursus americanus) litter production is likely dependent on maternal condition, which is in part dependent on the availability of fall foods. To date, no indices for black bear demographic parameters have been reported that particularly aid in population modeling. Thus, black bear population model parameterization is usually based on extensive field work with radiocollared animals. However, long-term, intensive field research can not be carried out indefinitely. We attempted to classify production of different litter sizes by black bears in Massachusetts using environmental and harvest-derived variables combined with individual black bear variables (weight, age, reproductive status). We used linear discriminant function analysis to classify litter events into categories (1 cub, 2 cubs, 3 cubs, or 4 cubs) and thus to identify variables that may index reproductive output. We also used 2 types of Bayesian analysis to estimate the probability distribution of litter sizes for Massachusetts black bears. During 1981-97, we observed 20 known first litters and 66 subsequent litters. We could not derive a predictable relationship among food abundance, bear traits, and litter size. This was due in part to black bears' propensity to use human-related food sources (primarily corn) in years of poor natural food abundance. Simple Bayesian estimates tended to overestimate the proportion of 2-cub first litters and 3-cub subsequent litters in Massachusetts. A different approach based on the multinomial distribution produced estimates of litter size distributions very close to that observed for subsequent litters. The observed distribution of first litters in Massachusetts was skewed much lower than other reported distributions, thus complicating our use of prior information in the Bayesian estimates. We suggest that litter size is relatively invariate locally and can be reliably estimated for modeling purposes using Bayesian techniques. Thus, researchers and managers can use the extensive data collected on black bear reproduction to help estimate sensitive parameters for their own specific populations in the absence of annual field data collection.





Book Series Title


This document is currently not available here.