Artificial refuges are human-made structures that aim to create safe places for animals to breed, hibernate, or take shelter in lieu of natural refuges. Artificial refuges are used across the globe to mitigate the impacts of a variety of threats on wildlife, such as habitat loss and degradation. However, there is little understanding of the science underpinning artificial refuges, and what comprises best practice for artificial refuge design and implementation for wildlife conservation. We address this gap by undertaking a systematic review of the current state of artificial refuge research for the conservation of wildlife. We identified 224 studies of artificial refuges being implemented in the field to conserve wildlife species. The current literature on artificial refuges is dominated by studies of arboreal species, primarily birds and bats. Threatening processes addressed by artificial refuges were biological resource use (26%), invasive or problematic species (20%), and agriculture (15%), yet few studies examined artificial refuges specifically for threatened (Vulnerable, Endangered, or Critically Endangered) species (7%). Studies often reported the characteristics of artificial refuges (i.e. refuge size, construction materials; 87%) and surrounding vegetation (35%), but fewer studies measured the thermal properties of artificial refuges (18%), predator activity (17%), or food availability (3%). Almost all studies measured occupancy of the artificial refuges by target species (98%), and over half measured breeding activity (54%), whereas fewer included more detailed measures of fitness, such as breeding productivity (34%) or animal body condition (4%). Evaluating the benefits and impacts of artificial refuges requires sound experimental design, but only 39% of studies compared artificial refuges to experimental controls, and only 10% of studies used a before-after-control-impact (BACI) design. As a consequence, few studies of artificial refuges can determine their overall effect on individuals or populations. We outline a series of key steps in the design, implementation, and monitoring of artificial refuges that are required to avoid perverse outcomes and maximise the chances of achieving conservation objectives. This review highlights a clear need for increased rigour in studies of artificial refuges if they are to play an important role in wildlife conservation.

Both fire and predators have strong influences on the population dynamics and behaviour of animals, and the effects of predators may either be strengthened or weakened by fire. However, knowledge of how fire drives or mediates predatorprey interactions is fragmented and has not been synthesised. Here, we review and synthesise knowledge of how fire influences predator and prey behaviour and interactions. We develop a conceptual model based on predator-prey theory and empirical examples to address four key questions: (i) how and why do predators respond to fire; (ii) how and why does prey vulnerability change post-fire; (iii) what mechanisms do prey use to reduce predation risk post-fire; and (iv) what are the outcomes of predator-fire interactions for prey populations? We then discuss these findings in the context of wildlife conservation and ecosystem management before outlining priorities for future research. Fire-induced changes in vegetation structure, resource availability, and animal behaviour influence predator-prey encounter rates, the amount of time prey are vulnerable during an encounter, and the conditional probability of prey death given an encounter. How a predator responds to fire depends on fire characteristics (e.g. season, severity), their hunting behaviour (ambush or pursuit predator), movement behaviour, territoriality, and intra-guild dynamics. Prey species that rely on habitat structure for avoiding predation often experience increased predation rates and lower survival in recently burnt areas. By contrast, some prey species benefit from the opening up of habitat after fire because it makes it easier to detect predators and to modify their behaviour appropriately. Reduced prey body condition after fire can increase predation risk either through impaired ability to escape predators, or increased need to forage in risky areas due to being energetically stressed. To reduce risk of predation in the post-fire environment, prey may change their habitat use, increase sheltering behaviour, change their movement behaviour, or use camouflage through cryptic colouring and background matching. Field

Habitat loss and degradation, and their interaction with other threats, are driving declines in animal populations worldwide. One potential approach for mitigating these threats is to create artificial habitat structures as substitutes for lost or degraded natural structures. Here, we provide – to the best of our knowledge – the first general definition of artificial habitat structures and synthesize important considerations for their effective use. We show that such structures represent a versatile conservation tool that has been trialed in a variety of contexts globally, albeit with varying degrees of success. The design of these structures must be well informed by the drivers of natural habitat selection, and their use should be part of an experimental framework to enable evaluation and refinement. We highlight possible ecological risks associated with the use of artificial habitat structures and urge that they not be exploited as inappropriate biodiversity offsets or for greenwashing. Looking forward, cross‐disciplinary collaborations will facilitate the development of sophisticated and effective structures to assist animal conservation in this era of rapid global change.

Understanding how fire influences animal behaviour, such as movement and resource selection, is important for ecosystem management because it can improve our capacity to predict how species will respond. We assessed microhabitat selection by two small mammals, the bush rat (Rattus fuscipes) and agile antechinus (Antechinus agilis), in response to a low intensity prescribed fire. We used spool and line tracking and touch pole vegetation surveys to quantify microhabitat selection along 21 trails for bush rats and 22 for antechinuses before and after fire. In unburnt areas, bush rats showed positive selection for sedges, logs, and habitat complexity, with selection further increasing in burnt areas for sedges, ferns, shrubs, habitat complexity and unburnt patches. Agile antechinuses showed no significant microhabitat selection in unburnt or burnt areas and no change in response to fire. Their lack of response to ground fires may be due, partially, to their scansorial behaviour and use of tree hollows as refuge sites. Strong selection by bush rats for small unburnt patches suggests that even low intensity, patchy fires such as planned burns can impact bush rats and that high burn patchiness may help bush rats persist in recently burnt areas. Future fire planning should consider both behavioural and population responses of animals to fire.

In many countries, more resources are required for the conservation of threatened species than are made available. Australia, the home to many unique mammalian species, has been identified as one of the top seven countries worldwide with the highest biodiversity loss and lowest amount of funding dedicated to conservation. Conservation action is generally determined by what we perceive to be the most significant conservation problems; therefore, identifying potential biases in research effort is crucial for prioritising research actions for effective conservation outcomes. We reviewed the published literature on the koala Phascolarctos cinereus, a vulnerable arboreal marsupial. Our aim was to assess and quantify research efforts (number of publications) and research coverage (diversity of subject areas) of koala‐related research. We then focused on the research effort of threats to identify primary threats and fields requiring further research. Using a combination of key words, journal name, study title, and methods, we reviewed and categorised 714 publications to identify the focus of research efforts. We investigated subject area, geographic and temporal publication patterns, and the extent to which primary threats have been addressed. We found a trend of increasing number of publications with a diversification of threat‐based studies, and identified both geographic and subject bias in research effort. Geographic bias in the volume of research indicates that the level of koala research between regions may be a response to the species’ distribution, status and impact of threats. Research areas ‘anatomy and physiology’ and ‘threats’ received comparatively more research attention than other areas. We found a research bias between threats, with ‘disease’ representing a high percentage of all threat‐based publications. Areas that have been identified but are receiving little research attention include habitat selection and use, along with the effects of habitat loss, fragmentation, and degradation.