Estimation of wildlife population ratios incorporating survey design and visibility bias

Age and sex ratio statistics are often a key component of the evaluation and management of wildlife populations. These statistics are determined from counts of animals that are commonly plagued by errors associated with either survey design or visibility bias. We present age and sex ratio estimators that incorporate both these sources of error and include the typical situation that animals are sampled in groups. Aerial surveys of elk (Cervus elaphus) in northcentral Idaho illustrate that differential visibility of age or sex classes can produce biased ratio estimates. Visibility models may be used to provide corrected estimates of ratios and their variability that incorporates errors due to sampling, visibility bias, and visibility estimation.