Remote Camera Research Articles

Deep learning to extract the meteorological by-catch of wildlife cameras.

Microclimate-proximal climatic variation at scales of metres and minutes-can exacerbate or mitigate the impacts of climate change on biodiversity. However, most microclimate studies are temperature centric, and do not consider meteorological factors such as sunshine, hail and snow. Meanwhile, remote cameras have become a primary tool to monitor wild plants and animals, even at micro-scales, and deep learning tools rapidly convert images into ecological data. However, deep learning applications for wildlife imagery have focused exclusively on living subjects. Here, we identify an overlooked opportunity to extract latent, ecologically relevant meteorological information. We produce an annotated image dataset of micrometeorological conditions across 49 wildlife cameras in South Africa's Maloti-Drakensberg and the Swiss Alps. We train ensemble deep learning models to classify conditions as overcast, sunshine, hail or snow. We achieve 91.7% accuracy on test cameras not seen during training. Furthermore, we show how effective accuracy is raised to 96% by disregarding 14.1% of classifications where ensemble member models did not reach a consensus. For two-class weather classification (overcast vs. sunshine) in a novel location in Svalbard, Norway, we achieve 79.3% accuracy (93.9% consensus accuracy), outperforming a benchmark model from the computer vision literature (75.5% accuracy). Our model rapidly classifies sunshine, snow and hail in almost 2 million unlabelled images. Resulting micrometeorological data illustrated common seasonal patterns of summer hailstorms and autumn snowfalls across mountains in the northern and southern hemispheres. However, daily patterns of sunshine and shade diverged between sites, impacting daily temperature cycles. Crucially, we leverage micrometeorological data to demonstrate that (1) experimental warming using open-top chambers shortens early snow events in autumn, and (2) image-derived sunshine marginally outperforms sensor-derived temperature when predicting bumblebee foraging. These methods generate novel micrometeorological variables in synchrony with biological recordings, enabling new insights from an increasingly global network of wildlife cameras.

Temperate rocky reef fish community patterns in a coastal marine protected area (MPA) from northern Chile, utilizing remote underwater video cameras (RUVs)

Shallow water seascapes are complex systems, important for the life history and connectivity of many marine fish species. Remote Underwater Video cameras (RUVs) are an increasingly used, non-lethal technique able to record the occurrence of reef fish species with little disturbance. In this study, RUVs were installed in a Marine Protected Area (Chañaral Island, northern Chile, 29°2'S, 71°34'W) at three different depth strata (5, 10, 15 m) in three different habitats (sand and boulder, bedrock, and kelp forest) during October 2018, July 2021 and October 2021. The RUV system recorded a total of 20 species belonging to 16 families over 73.5 h of videotaping. The most recurrent fishes were damselfish Chromis crusma, Chilean sandperch Pinguipes chilensis, and marblefish Aplodactylus punctatus. There was a significant difference in fish assemblage composition among depth strata and seasons, but not among habitats. In July 2021 (austral winter) higher densities of small cryptobenthic fish (Helcogrammoides spp.) and medium-sized carnivorous Chirodactylus variegatus were detected, whereas P. chilensis was associated to the deeper stratum, and the herbivorous blenny Scartichthys viridis was related to the shallower depths, which is likely due to differences in the habitat structure and composition. Winter 2021 showed a higher species richness than in spring 2018, while the evenness (Pielou’s J’) was significantly higher in spring 2018 than winter/spring 2021. There was no other strong or consistent pattern, which suggests that most fish species use the entire depth gradient throughout the year. Finally, the results suggest that the use of RUVs for the estimation of biodiversity of rocky reef fishes is a cost-effective way to improve our understanding of nearshore community shifts in relation to environmental forcing.

Comparison of in‐person and remote camera lek surveys for prairie grouse (Tympanuchus spp.)

AbstractIn‐person lek count surveys are commonly used for estimating population size and trends for prairie grouse. However, the emergence of noninvasive camera trap survey methods holds promise for more cost‐effective and precise estimates of lekking species. To evaluate the efficacy of using camera traps, we deployed a remote camera study at lekking grounds over 3 years in conjunction with in‐person surveys. Our objectives were to 1) develop an effective remote camera survey for greater prairie‐chickens (GRPC; Tympanuchus cupido) and sharp‐tailed grouse (STGR; T. phasianellus), 2) compare metrics of male detection, maximum male counts, and male abundance estimates derived from in‐person versus remote camera surveys, 3) assess lek activity over the survey season to inform survey timing, and 4) evaluate costs for each survey type. We found that in‐person surveys resulted in maximum male GRPC and STGR counts. The estimated number of male GRPC and STGR on leks were comparable between in‐person surveys and camera monitoring when accounting for detection probability with N‐mixture models. Camera traps constantly monitored leks over the season which provided daily and seasonal activity patterns of prairie grouse. Total cost of GRPC remote camera surveys was higher than in‐person surveys, but hourly cost was less ($0.77 vs. $160 per hour). Remote camera survey costs for GRPC were high because of time classifying photos which could be reduced by decreasing the amount of time remote cameras were operated or using automated classification software to remove blank photos. We believe the use of remote cameras could supplement in‐person surveys for future lek monitoring and aid future survey efforts by identifying yearly differences in activity and presence at leks inconsistently visited by birds.

A chimpanzee’s time to feast: seasonality of Macrotermes flight hole activity and alate dispersal flights detected by termite-fishing experiments and camera traps in the Issa Valley, Tanzania

IntroductionInvestigation into chimpanzee termite-fishing behavior has provided invaluable insights into the evolution of animal tool use and culture. However, research to date often overlooks how the ecology of termites may shape the nature of this predator–prey interaction. Though anecdotal links between meteorological factors, termite dispersal flights and the ability of chimpanzees to termite-fish have been made, these relationships were never empirically tested. In this study, we examined if and how the reproductive ecology of mound building termites (Macrotermes subhyalinus) affects the availability of this resource to termite-fishing chimpanzees in the Issa Valley, Tanzania.MethodsWe utilized in-person termite-fishing experiments, remote camera trap footage, as well as meteorological data to model both the availability of termite prey across seasons for chimpanzees as well as the timing of dispersal flights of the termite reproductive caste (alates). Additionally, we describe the predator diversity that characterizes alate swarming events during their seasonal dispersal flights.ResultsWe identify strong relationships between meteorological factors, Macrotermes alate dispersal flights and vulnerability of the colony to predation. As rainfall accumulates, Macrotermes become more abundant in their flight holes and are thus more vulnerable to researcher termite-fishing experiments. This increased accessibility to Macrotermes continues until alate dispersal flights which we also find are linked to cumulative rainfall as well as negatively correlated to wind speed.DiscussionThis type of baseline data on Macrotermes ecology is fundamental when forming hypotheses and designing studies into topics such as chimpanzee culture, tool-use, and foraging cognition.

Variation in spatiotemporal activity may reduce competitive interactions between invasive wild pigs (Sus scrofa) and native mammal species

Interactions between sympatric species influence the ecology and behaviour of individuals and species. Invasive species can often alter community dynamics by generating novel pressures that native species are less able to respond to. The scope of these pressures may depend on the ecology and life history of the invasive species. We used remote camera traps to investigate how native mammals’ spatiotemporal activity patterns were altered by the presence of invasive wild pig ( Sus scrofa Linnaeus, 1758). By analyzing patterns of daily activity and implementing Cox proportional hazard models, our results suggest that (a) temporal partitioning may play a role in ameliorating competition between native mammals and wild pigs and (b) native species may be avoiding wild pigs by prolonging their return times to particular sites. We provide an example for monitoring the ecological changes wild pigs may create within ecosystems as they become more abundant in non-traditional ranges.

Eat or be eaten: Implications of potential exploitative competition between wolves and humans across predator-savvy and predator-naive deer populations.

Recolonization of predators to their former ranges is becoming increasingly prevalent. Such recolonization places predators among their prey once again; the latter having lived without predation (from such predators) for a considerable time. This renewed coexistence creates opportunities to explore predation ecology at both fundamental and applied levels. We used a paired experimental design to investigate white-tailed deer risk allocation in the Upper and Lower Peninsulas (UP and LP) in Michigan, USA. Wolves are functionally absent in the LP, while deer in the UP coexist with a re-established wolf population. We treated 15 sites each in UP and LP with wolf olfactory cues and observed deer vigilance, activity, and visitation rates at the interface of habitat covariates using remote cameras. Such a paired design across wolf versus no-wolf areas allowed us to examine indirect predation effects while accounting for confounding parameters such as the presence of other predators and human activity. While wolf urine had no effect across most metrics in both UP and LP, we observed differences in deer activity in areas with versus without wolves. Sites treated with wolf urine in the UP showed a reduction in crepuscular deer activity, compared to control/novel-scent treated sites. Furthermore, we observed a strong positive effect of vegetation cover on deer vigilance in these sites. This indicates that simulated predator cues likely affect deer vigilance more acutely in denser habitats, which presumably facilitates predation success. Such responses were however absent among deer in the LP that are presumably naïve toward wolf predation. Where human and non-human predators hunt shared prey, such as in Michigan, predators may constrain human hunting success by increasing deer vigilance. Hunters may avoid such exploitative competition by choosing hunting/bait sites located in open areas. Our results pertaining to fundamental predation ecology have strong applied implications that can promote human-predator coexistence.

Bridge vibration measurements using different camera placements and techniques of computer vision and deep learning

In this paper, a new framework is proposed for monitoring the dynamic performance of bridges using three different camera placements and a few visual data processing techniques at low cost and high efficiency. A deep learning method validated by an optical flow approach for motion tracking is included in the framework. To verify it, videos taken by stationary cameras of two shaking table tests were processed at first. Then, the vibrations of six pedestrian bridges were measured using structure-mounted, remote, and drone-mounted cameras, respectively. Two techniques, displacement and frequency subtractions, are applied to remove systematic motions of cameras and to capture the natural frequencies of the tested structures. Measurements on these bridges were compared with the data from wireless accelerometers and structural analysis. Influences of critical parameters for camera setting and data processing, such as video frame rates, data window size, and data sampling rates, were also studied carefully. The research results show that the vibrations and frequencies of structures on the shaking tables and existing bridges can be captured accurately with the proposed framework. These camera placements and data processing techniques can be successfully used for monitoring their dynamic performance.

Spatial and temporal niche overlap of aardwolves and aardvarks in Serengeti National Park, Tanzania.

Species interactions can influence species distributions, but mechanisms mitigating competition or facilitating positive interactions between ecologically similar species are often poorly understood. Aardwolves (Proteles cristata) and aardvarks (Orycteropus afer) are nocturnal, insectivorous mammals that co-occur in eastern and southern Africa, and knowledge of these species is largely limited to their nutritional biology. We used aardwolf and aardvark detections from 105 remote cameras during 2016-2018 to assess their spatial and temporal niche overlap in the grasslands of Serengeti National Park, Tanzania. Using a multispecies occupancy model, we identified a positive interaction between occupancy probabilities for aardwolves and aardvarks. Slope, proportion of grassland and termite mound density did not affect the occupancy probabilities of either species. The probability of aardwolf, but not aardvark, occupancy increased with distance to permanent water sources, which may relate to predation risk avoidance. Diel activity overlap between aardwolves and aardvarks was high during wet and dry seasons, with both species being largely nocturnal. Aardwolves and aardvarks have an important ecological role as termite consumers, and aardvarks are suggested to be ecosystem engineers. Our results contribute to a better understanding of the spatial and temporal niche of insectivores like aardwolves and aardvarks, suggesting high spatial and temporal niche overlap in which commensalism occurs, whereby aardwolves benefit from aardvark presence through increased food accessibility.

Forest structure has stronger effects than cattle occurrence on the occupancy of a carnivore guild

Anthropogenic activities can profoundly affect ecological communities. This is true of the most ubiquitous type of anthropogenic land-use, livestock grazing. While livestock grazing is known to impact vegetation structure, soil, and hydrological features, the effects of livestock occurrence on animal communities are often more complex. Herein, we estimated the relative effects of cattle occurrence and vegetation structure on the occupancy of a carnivore guild in northern California and southern Oregon, USA. We used remote cameras to non-invasively collect detection and non-detection data of cattle and nine carnivore species. We incorporated detection data into a Bayesian hierarchical occupancy model to estimate the effects of livestock occurrence and vegetation structure on carnivore occupancy. We found varied effects of cattle occurrence on carnivore occupancy, with most species showing no clear response to cattle occurrence. Vegetation structure, including structural diversity and vegetation productivity, had stronger effects on carnivore occupancy than cattle occurrence. This work provides an exploration of the effects of cattle occurrence and forest structure on carnivore space-use in a grazed forest system, and suggests vegetation structure may have stronger effects than cattle occurrence on carnivore occupancy in grazed forest systems. Future work to clarify the direct and indirect effects of livestock occurrence can inform conservation and management strategies in forested ecosystems.

A review of field techniques to estimate lion presence and abundance

Abstract Since the 1990s, the number of lions (Panthera leo) has declined by about 43% across African range states. Reliable lion population estimates can contribute to effective management and inform local and international conservation policies. Though many survey methods have been used to estimate lion abundance, an assessment of their usefulness and effectiveness is lacking. We reviewed and identified available methods used to estimate lion populations, then investigated their feasibility, cost‐effectiveness and suitability to achieve primary objectives. We identified 12 methods in two categories: direct (minimum counts, call‐in surveys, capture–recapture, opportunistic sightings, remote cameras, distance‐based and aerial counts) and indirect (track counts, secondary information, prey‐based, genetic samples and roar counts). We identified 48 studies (57%) in which lion population estimation was not repeatable. Additionally, 38% did not provide any measure of precision. About 28% of the studies were conducted specifically for management purposes, while remaining studies focused on population estimates and dynamics. Because data from these studies have influenced local management and global lion conservation policies, we recommend basing future policies and management on repeatable, precise and accurate population estimates supported by appropriate and cost‐effective experimental designs. In most cases, these estimates are obtained through actual lion counts derived from direct observation methods including call‐in, capture–recapture and remote camera surveys or through reliable animal signs such as genetic samples.

Dining with a glutton: an intraguild interaction between scavenging wolverine (Gulo gulo) and lynx (Lynx canadensis)

AbstractIntraspecific interactions between mesocarnivores are not commonly observed and consequentially not well understood, particularly for species inhabiting wilderness regions. Using a remote camera, we captured interactions between a Canada lynx (Lynx canadensis) and wolverine (Gulo gulo) scavenging on the remains of a bison (Bison bison). Over a 23‐min period representing nine photographic sequences (1–3 photographs per sequence), we observed (1) a wolverine approach the remains that were being fed on by a lynx, (2) the two animals feeding on the remains at the same time for about 19 min, (3) the wolverine chasing the lynx off of the carcass and, thereafter, (4) the wolverine excluding the lynx from the carcass. Our observation is of potential scientific value because it sheds light on behavioral interactions between these two species, particularly at a rich food source. Intraguild interactions between these species appear to include tolerance toward one another, as well as interference competition, with the wolverine being the dominant species. Our observation highlights the new information on interactions between mesocarnivores that are becoming illuminated with the use of remote cameras, revealing that these interactions may be more complex than once thought.

Rural selection drives the evolution of an urban-rural cline in coat color in gray squirrels.

Phenotypic differences between urban and rural populations are well-documented, but the evolutionary processes driving trait variation along urbanization gradients are often unclear. We combined spatial data on abundance, trait variation, and measurements of fitness to understand cline structure and test for natural selection on heritable coat color morphs (melanic, gray) of eastern gray squirrels (Sciurus carolinensis) along an urbanization gradient. Population surveys using remote cameras and visual counts at 76 sites along the urbanization gradient revealed a significant cline in melanism, decreasing from 48% in the city center to <5% in rural woodlands. Among 76 squirrels translocated to test for phenotypic selection, survival was lower for the melanic than gray morph in rural woodlands, whereas there was no difference in survival between color morphs in the city. These results suggest the urban-rural cline in melanism is explained by natural selection favoring the gray morph in rural woodlands combined with relaxed selection in the city. Our study illustrates how trait variation between urban and rural populations can emerge from selection primarily in rural populations rather than adaptation to novel features of the urban environment.

DEEP LEARNING OBJECT DETECTION APPLICATION TO SURFING WAVE QUALITY

Quantitative monitoring is imperative to the sustainable management of coastal resources. Surfing resources have been both created and degenerated or destroyed by activities in the coastal zone. Effective surfing wave quality monitoring requires identification and tracking of the breaking part of the wave and the unbroken wave crest. Remote Camera Systems (RCS) have proven their utility in being able to monitor the coastal zone and provide almost continuous, high frequency data collection. RCSs lend themselves very well to the monitoring of surf breaks which are highly dynamic. The images captured from an RCS monitoring a surf break on the west coast of Aotearoa New Zealand are used to train a Convolutional Neural Network (CNN) to detect the break points (BP), associated crest orientation and relative Still Water Level (SWL) of each instance of breaking waves in each image. Model settings and image annotations were modified over a suite of training cases to improve model efficacy, which was evaluated each epoch of training with mean Average-Precision (mAP; max 1). A mAP of 0.794 was achieved for the BP and Crest Point (CP) CNN, and 8.634 for the SWL. The model was used to detect ~1.6 M objects across ~1 million images, with a mean confidence value of all BP-CP detections of 0.63 and more than 70percent of detections being greater than 0.5. This model enables the first automation of meaningful surfing wave quality monitoring.

Evaluating a tandem human-machine approach to labelling of wildlife in remote camera monitoring

Remote cameras (“trail cameras”) are a popular tool for non-invasive, continuous wildlife monitoring, and as they become more prevalent in wildlife research, machine learning (ML) is increasingly used to automate or accelerate the labor-intensive process of labelling (i.e., tagging) photos. Human-machine hybrid tagging approaches have been shown to greatly increase tagging efficiency (i.e., time to tag a single image). However, those potential increases hinge on the extent to which an ML model makes correct vs. incorrect predictions. We performed an experiment using a ML model that produces bounding boxes around animals, people, and vehicles in remote camera imagery (MegaDetector) to consider the impact of a ML model's performance on its ability to accelerate human labeling. Six participants tagged trail camera images collected from 12 sites in Vermont and Maine, USA (January–September 2022) using three tagging methods (one with ML bounding box assistance and two without assistance). We used a generalized linear mixed model to examine the influence of ML model performance and tagging method on tagging efficiency. We found that ML bounding boxes offer significant improvement in tagging efficiency when labelling data compared to unassisted tagging. Additionally, the time taken to label with bounding boxes was not statistically different from an unassisted tagging approach. However, we found that gains in efficiency are contingent on the ML algorithm's performance and that incorrect ML predictions, particularly the 4.2% false positive and 3.6% false negative predictions, can slow the tagging process compared to a non-hybrid approach. These findings indicate that although practitioners usually forgo the production of bounding boxes when selecting a data labelling process due to the increased effort, ML bounding box-assisted tagging can offer an efficient method for labeling. More broadly, ML-assisted data labelling offers an opportunity to accelerate the analysis of trail camera imagery, but an assessment of the ML model's performance can illuminate whether the hybrid-tagging approach is ultimately a help or hinderance.

Virtual snow stakes: a new method for snow depth measurement at remote camera stations

AbstractRemote cameras are used to study demographics, ecological processes, and behavior of wildlife populations. Cameras have also been used to measure snow depth with physical snow stakes. However, concerns that physical instruments at camera sites may influence animal behavior limit installation of instruments to facilitate collecting such data. Given that snow depth data are inherently contained within images, potential insights that could be made using these data are lost. To facilitate camera‐based snow depth observations without additional equipment installation, we developed a method implemented in an R package called edger to superimpose virtual measurement devices onto images. The virtual snow stakes can be used to derive snow depth measurements. We validated the method for snow depth estimation using camera data from Latah County, Idaho, USA in winter 2020–2021. Mean bias error between the virtual snow stake and a physical snow stake was 5.8 cm; the mean absolute bias error was 8.8 cm. The mean Nash Sutcliffe Efficiency score comparing the fit of the 2 sets of measurements within each camera was 0.748, indicating good agreement. The edger package provides researchers with a means to take critical measurements for ecological studies without the use of physical objects that could alter animal behavior, and snow data at finer scales can complement other snow data sources that have coarser spatial and temporal resolution.

An integrated spatial capture–recapture approach reveals the distribution of a cryptic carnivore in a protected area

AbstractQuantifying animal abundance, density, and distributions affords the opportunity to understand the effects of landscape structure and change on species of conservation interest, but estimating these parameters can be difficult for rare and cryptic species. Noninvasive sampling methods, such as remote cameras or scat DNA, can mitigate the challenges of studying rare and cryptic species while also minimizing effects on species of conservation interest or concern. Data derived from these methods can be integrated into robust, contemporary quantitative methods, including spatial capture–recapture (SCR) models, which provide a hierarchical framework for jointly estimating animal abundance, distribution, and, consequently, density. Herein, we developed an integrated SCR model to estimate the abundance, density, and distribution of a large carnivore of conservation interest, the cougar (Puma concolor), in a rugged and remote protected area, Yosemite National Park, California, USA. We combined spatial encounter data from DNA‐based individual identification of scats with detection count data derived from remote cameras to estimate cougar density and detection probability in Yosemite. We further estimated how cougar density and detection probability varied as a result of vegetation, topography, anthropogenic and natural linear features, and survey effort. Using data collected in 2019 and 2020, we estimated a median of 31 (SD = 3.96, 95% credible intervals = 24–39) cougars in Yosemite across the two years, with higher densities associated with productive, vegetated areas. We found detection probability by scent detection teams was higher for females than males and positively correlated with survey effort, proximity to trails, and distance farther from roads and streams. Our study illustrates the utility of noninvasive survey methods that yield individual identities in rugged and remote environments, where capture and handling of cryptic, low‐density animals is logistically challenging and cost prohibitive. Integrated modeling approaches, as used here, allow ecologists to leverage empirical data using a robust quantitative framework that can effectively address conservation objectives. Through this work, we demonstrate the importance of large, contiguous, and heterogeneous ecosystems to the ecology of wide‐ranging species that occur in dynamic landscapes.

Eating an elephant, one bite at a time: Predator interactions at carrion bonanzas

Resource specific competition between predators has typically been studied from their interactions at meso-herbivore carcasses, because such carcasses are abundant. Mega-carcasses like those of elephants are rare but unparalleled in the extent of carrion biomass they offer and the long durations they can persist. These rare resource bonanzas can thus provide unique opportunities to understand sympatric species interactions within likely relaxed competitive scenarios. Using remote cameras that were operational 24-h a day, we monitored two elephant carcasses in Tsavo, Kenya, from when they were discovered until they were completely consumed or became inaccessible. While we found high temporal overlaps in activity patterns between all predators, the terrestrial predator guild (lion/leopard/spotted hyena) was not observed to feed simultaneously, suggesting strong interference competition. Based on photo-analysis and video-evidence of exclusion from a carcass, interference competition within the terrestrial predator guild favored lions over hyenas, and hyenas over leopards. The carcass at the terrestrial-aquatic interface showed more simultaneous feeding bouts between predators (crocodile/spotted hyena), indicating either facilitation and/or higher coexistence between predators that typically occupy different niches. We also observed a hippopotamus scavenging from an elephant carcass, thereby documenting a rare instance of a megaherbivore feeding on a megaherbivore. Our results highlight the importance of monitoring such carcasses through remote cameras, which can significantly add to our existing understanding of food webs and carrion ecology.

Remote cameras capture dung burial by burrowing cricket

Animal dung holds an important role in the nutrient cycle of nearly every ecosystem on Earth. It also acts as an ephemeral resource for invertebrate communities and has facilitated the specialization of many life history strategies. Most ecosystem functions associated with dung are linked to burial - a process most often attributed to dung beetles. However, identifying other organisms that contribute to dung burial is important to understand how communities provide this ecosystem service. During a trail camera experiment intended to monitor dung beetle behavior, we observed an unidentified species of cricket burying dung pellets underground. We were unable to find previous documentation of dung-burying behavior for this animal group, raising questions of how common or extensive this behavior is. Particularly under the threat of global insect declines, understanding which species are responsible for key processes may be crucial to maintaining ecosystem function and resilience. Additionally, this experiment exemplifies how camera traps can be used in nontraditional applications to observe and monitor unique events, behaviors, and processes, such as the decay of ephemeral resource patches and insect behavior.

MAMMALNET – Citizen Science Data Collection from a One Health Perspective

Abstract The ambition of One Health (OH) is to focus on people, animals and the ecosystem equally (Tripartite and UNEP support OHHLEP’s definition of “One Health” (who.int)). This requires adequate data on wildlife. MAMMALNET is a European consortium set up to collect wildlife occurrence data, with the specific aim of improving our understanding and prediction of disease spread. MAMMALMET encourages citizens and professionals to report mammal sightings on an ad hoc basis (iMammalia app) or through surveys using remote camera traps (MammalWeb or Agouti). This combines data from different sources, increases our understanding of mammal distribution and aids in monitoring the spread of invasive species. MAMMALNET participants can see their records and maintain a list of species sightings. These data are vital to our understanding of the ecosystem and how this may change over time, providing background data for monitoring species. These data complement and contribute to reinforcing wildlife health reports, such as recording dead wild boar in outbreak areas of African Swine Fever. Such records are followed up for disease sampling to monitor the spread of disease. The data can also be used to predict the distribution and abundance of wild species, provide the denominator data for disease reports and predict the potential for disease spread and control. MAMMALNET is committed to open science since OH requires not only an interdisciplinary approach but practical collaboration and sharing of standardized data. These outputs can help predict the potential spread and control of zoonotic diseases, such as rabies, with benefits for human health. Information © The Authors 2023

Small mammals reduce activity during high moon illumination under risk of predation by introduced predators

Predation influences prey survival and drives evolution of anti-predator behaviour. Anti-predator strategies by prey are stimulated by direct encounters with predators, but also by exposure to indicators of risk such as moonlight illumination and vegetation cover. Many prey species will suffer increased risk on moonlit nights, but risk may be reduced by the presence of dense vegetation. Determining the role of vegetation in reducing perceived risk is important, especially given predictions of increased global wildfire, which consumes vegetation and increases predation. We used remote cameras in southeastern Australia to compare support for the predation risk and habitat-mediated predation risk hypotheses. We examined the influence of moonlight and understorey cover on seven 20–2500 g mammalian prey species and two introduced predators, red foxes and feral cats. Activity of all prey species reduced by 40–70% with increasing moonlight, while one species (bush rat) reduced activity in response to increasing moonlight more sharply in low compared to high understorey cover. Neither predator responded to moonlight. Our findings supported the predation risk hypothesis and provided limited support for the habitat-mediated predation risk hypothesis. For prey, perceived costs of increased predation risk on moonlit nights outweighed any benefits of a brighter foraging environment.