Modelling the sustainable harvest of wild populations for the conservation of a threatened amphibian

Harvesting wild populations is increasingly used to support conservation initiatives, including translocations, captive assurance colonies, and breeding programs. Best-practice guidelines recommend that population viability analyses should be developed to assess the potential impact of harvest on source populations, allowing quantification of the risk incurred by the conservation programme. We developed a stochastic population model to assess the relative impact of harvesting southern bell frogs (Litoria raniformis) from three well-studied populations. We compared the probability of local extinction when harvesting from three life stages (spawn, tadpoles, and adults) and at 21 harvest proportions (from 0 to 1, in 0.05 increments). Source populations with an estimated average of 100 adult frogs and with three or fewer breeding opportunities every 5 years showed increased risk of local extinction and lower minimum adult populations from all harvest strategies. Sources with an estimated average population >600 adult frogs and breeding opportunities in at least eight out of every 10 years were more resilient to harvest. For these populations, harvesting up to 25% of spawning masses or late-stage tadpoles had a negligible impact on the probability of local extinction and caused a slight decrease in the minimum adult population. Harvesting adults impacted source populations more strongly than harvesting spawn or tadpoles. Our model highlights the importance of frequent breeding opportunities to offset the effects of harvesting L. raniformis. Populations sustained by environmental water delivery should be prioritized for the two consecutive breeding seasons following harvest in the absence of natural wetland inundation. We demonstrate a tractable approach to assess and compare the relative risk of wild harvest for threatened amphibians, with implications for the conservation of our focal species, L. raniformis. We encourage conservation practitioners to implement population viability analyses prior to wild harvest, whenever sufficient demographic information is available to build the requisite models.