Abstract
Wildlife agencies are tasked with sustaining healthy wildlife populations. Advances in understanding health in wildlife, are, however, suppressed by issues with surveillance. Cost and logistical complexity are leading reasons why ideal surveillance implementation is often infeasible. A particularly important issue in surveillance is the sample size necessary to declare a local wildlife population (or social group) disease free. More precisely, how many animals must we sample to conclude with high confidence that prevalence of an infectious disease is below a specified threshold? Here we show that the answer to this question hinges on the ease of transmission between animals, a factor not considered in standard sample-size formulas. When disease statuses of animals in a local population are positively correlated, such as when a species forms social groups, the sample size requirement needed to declare freedom from disease is substantially lower relative to sample sizes suggested by existing hypergeometric and binomial models. Local wildlife populations or social groups must satisfy key properties for scientists to leverage this saving, but a reduction in cost arising in such a scenario is a welcome win in surveillance implementation.