Dejan Stojanovic

Implementing effective, affordable and ethical conservation management will be crucial for minimising future biodiversity losses. Such management requires reliable foundational evidence to help managers make informed choices about how to address the needs of their target species. Unfortunately, such evidence is still lacking for many species and management scenarios. One major global challenge is improving the reproductive success of threatened species in the context of predation. We conducted a systematic review and meta‐analysis of in situ experiments that used non‐lethal methods to protect bird nests against predators, with the aims of summarising global trends in nest protection efforts, comparing the effectiveness of different protection measures and informing future research and management. We considered peer‐reviewed studies in English. We detected a large geographic and taxonomic bias in the evidence base with 58% of articles conducted in North America and 76% on ground‐nesting birds. Less than 3% of articles involved taxa listed as Endangered or Critically Endangered and 51% of study units lasted just a single breeding season. Nests protected with exclosures, fences and guards were more likely to be successful than their unprotected controls. Interventions involving deterrents, conditioned taste aversion, chemical camouflage and diversionary feeding did not have a significant positive effect on nest success, but the interventions in these categories were less common and more diverse in nature. Synthesis and applications . To increase their conservation value, future non‐lethal nest protection experiments should whenever possible clearly state overall aims, take place over multiple seasons, use a comparable control and test non‐lethal protection methods independently of lethal predator control. Greater focus is required on under‐studied taxa such as cup‐nesting songbirds and birds in South America, Africa and Asia, and novel protection techniques such as deterrents and chemical camouflage. Practitioners should consider the evidence we synthesise here when deciding whether non‐lethal nest protection approaches are optimal for their study system, to increase conservation success and reduce ethical and financial costs.

Detection of animals is influenced by species traits, environment, season and the methods used. Analytical techniques can address imperfect detection, such as false absences, but data limitations hinder accurate density estimation. Identifying field survey methods that optimize detectability is therefore a high research priority, especially for species of conservation concern. We consider a model species, the sugar glider Petaurus notatus , to evaluate how optimizing detections can affect survey outcomes. A literature meta‐analysis found little innovation in baiting approaches for sugar gliders over four decades. Honey‐based baits prevailed, despite emerging evidence of opportunistic carnivory. We compared sugar glider detection probabilities using honey and fish‐based baits in a spatially explicit capture‐recapture study. Fish bait increased detections 33‐fold, and density estimation (0.12 ha −1 ) was only possible using the fish‐baited data – detections were too sparse using honey‐bait to facilitate analysis. Other factors that influence detectability include trap height and habitat connectivity, which were top moderators in our meta‐analysis, but these may be secondary to the bait or lure used. By using (i) species‐specific bait that accounts for the biology of the target species, and (ii) analytical tools to account for imperfect detection and heterogeneous movement of animals, we achieved good enough detection probabilities to undertake detailed analysis of the spatial ecology of sugar gliders. Compared to honey‐based baits (the prevailing approach of other studies on this species), the high detection probability attained with fish‐based baits provides an improvement on conventional practices. Our study is an important reminder that accounting for the biology of the target species during survey effort may mean the difference between collecting enough data for analysis or not. In the context of species conservation, optimizing detection likelihoods is critical for answering ecological questions about cryptic species, and our results demonstrate the need to reappraise generic approaches when better information is available.