Successful coexistence between large carnivores and humans is conditional upon effective mitigation of the impact of these species on humans, such as through livestock depredation. It is therefore essential for conservation practitioners, carnivore managing authorities, or livestock owners to know the effectiveness of interventions intended to reduce livestock predation by large carnivores. We reviewed the scientific literature (1990–2016), searching for evidence of the effectiveness of interventions. We found experimental and quasi-experimental studies were rare within the field, and only 21 studies applied a case-control study design (3.7% of reviewed publications). We used a relative risk ratio to evaluate the studied interventions: changing livestock type, keeping livestock in enclosures, guarding or livestock guarding dogs, predator removal, using shock collars on carnivores, sterilizing carnivores, and using visual or auditory deterrents to frighten carnivores. Although there was a general lack of scientific evidence of the effectiveness of any of these interventions, some interventions reduced the risk of depredation whereas other interventions did not result in reduced depredation. We urge managers and stakeholders to move towards an evidence-based large carnivore management practice and researchers to conduct studies of intervention effectiveness with a randomized case-control design combined with systematic reviewing to evaluate the evidence.

Carnivore predation on livestock often leads people to retaliate. Persecution by humans has contributed strongly to global endangerment of carnivores. Preventing livestock losses would help to achieve three goals common to many human societies: preserve nature, protect animal welfare, and safeguard human livelihoods. Between 2016 and 2018, four independent reviews evaluated >40 years of research on lethal and nonlethal interventions for reducing predation on livestock. From 114 studies, we find a striking conclusion: scarce quantitative comparisons of interventions and scarce comparisons against experimental controls preclude strong inference about the effectiveness of methods. For wise investment of public resources in protecting livestock and carnivores, evidence of effectiveness should be a prerequisite to policy making or large-scale funding of any method or, at a minimum, should be measured during implementation. An appropriate evidence base is needed, and we recommend a coalition of scientists and managers be formed to establish and encourage use of consistent standards in future experimental evaluations.

The ongoing recovery of terrestrial large carnivores in North America and Europe is accompanied by intense controversy. On the one hand, reestablishment of large carnivores entails a recovery of their most important ecological role, predation. On the other hand, societies are struggling to relearn how to live with apex predators that kill livestock, compete for game species, and occasionally injure or kill people. Those responsible for managing these species and mitigating conflict often lack fundamental information due to a long-standing challenge in ecology: How do we draw robust population-level inferences for elusive animals spread over immense areas? Here we showcase the application of an effective tool for spatially explicit tracking and forecasting of wildlife population dynamics at scales that are relevant to management and conservation. We analyzed the world’s largest dataset on carnivores comprising more than 35,000 noninvasively obtained DNA samples from over 6,000 individual brown bears (Ursus arctos), gray wolves (Canis lupus), and wolverines (Gulo gulo). Our analyses took into account that not all individuals are detected and, even if detected, their fates are not always known. We show unequivocal quantitative evidence of large carnivore recovery in northern Europe, juxtaposed with the finding that humans are the single-most important factor driving the dynamics of these apex predators. We present maps and forecasts of the spatiotemporal dynamics of large carnivore populations, transcending national boundaries and management regimes.

Population monitoring data may originate from multiple methods and are often sparse and fraught with incomplete information due to practical and economic constraints. Models that can integrate multiple survey methods and are able to cope with incomplete data may help investigators exploit available information more thoroughly. Here, we developed an integrated spatial capture–recapture (SCR) model to incorporate multiple data sources with imperfect individual identification. We contrast inferences drawn from this model with alternate models incorporating only subsets of the data available. Using extensive simulations and an empirical example of multi‐method brown bear (Ursus arctos) monitoring data from northern Pakistan, we quantified the benefits of including multiple sources of information in SCR models in terms of parameter precision and bias. Our multiple observation processes SCR model (MOP) yielded a more complete picture of the underlying processes, reduced bias, and led to more precise parameter estimates. Our results suggest that the greatest gains from integrated SCR models can be expected in situations where detection probability is low, a large proportion of detections is not attributable to individuals, and the degree of overlap between individual home ranges is low.