A Two-Species Occupancy Model with a Continuous-Time Detection Process Reveals Spatial and Temporal Interactions

Spatiotemporal interactions between ocelots and agoutis in Neotropical protected forests

Understanding how predators and their prey coexist in space and time is a core interest in ecology. Vast amounts of photographic detection data are now available from a growing number of camera-trap studies worldwide. These data boost our ability to study an elusive yet important topic in ecology: species interactions in space and time. Here, we investigated the spatiotemporal configuration of the activity of a typically nocturnal and crepuscular predator and a diurnal prey in protected areas. We explored whether agoutis (Dasyprocta leporina) respond to predation risks by adjusting the timing of activity to the occurrence and timing of activity of its potential predator, the ocelot (Leopardus pardalis) and whether the ocelot's occurrence responds to that of the prey. Using a custom Bayesian occupancy model for two species, we analysed camera-trap data collected over three protected tropical forests from Brazil and Suriname (2 to 9 years). Our model distinguishes between occupancy (i.e. spatial distribution) and detection (i.e. activity in space and time). We detected a positive spatial association between predator and prey, suggesting that ocelots seek places where agoutis are present. Instead of segregating in space, agoutis appeared to respond to increased predation risk by reduced activity. The most pronounced reduction in agouti activity (i.e. probability of detection) was during twilight in sites where ocelots occurred. Our results illustrate how jointly modelling interactions in both space and time informs our understanding of predator-prey coexistence.

When and where? Day-night alterations in wild boar space use captured by a generalized additive mixed model

Wild boar (Sus scrofa), an abundant species across Europe, is often subjected to management in agro-ecosystems in order to control population size, or to scare them away from agricultural fields to safeguard crop yields. Wild boar management can benefit from a better understanding on changes in its space use across the diel cycle (i.e., diel space use) in relation to variable hunting pressures or other factors. Here, we estimate wild boar diel space use in an agro-ecosystem in central Belgium during four consecutive "growing seasons" (i.e., April-September). To achieve this, we fit generalized additive mixed models (GAMMs) to camera trap data of wild boar aggregated over 1-h periods. Our results reveal that wild boar are predominantly nocturnal in all of the hunting management zones in Meerdaal, with activity peaks around sunrise and sunset. Hunting events in our study area tend to take place around sunrise and sunset, while non-lethal human activities occur during sunlight hours. Our GAMM reveals that wild boar use different areas throughout the diel cycle. During the day, wild boar utilized areas in the centre of the forest, possibly to avoid human activities during daytime. During the night, they foraged near (or in) agricultural fields. A post hoc comparison of space use maps of wild boar in Meerdaal revealed that their diurnal and nocturnal space use were uncorrelated. We did not find sufficient evidence to prove that wild boar spatiotemporally avoid hunters. Finally, our work reveals the potential of GAMMs to model variation in space across 24-h periods from camera trap data, an application that will be useful to address a range of ecological questions. However, to test the robustness of this approach we advise that it should be compared against telemetry-based methods to derive diel space use.

Effects of resource availability and interspecific interactions on Arctic and red foxes' winter use of ungulate carrion in the Fennoscandian low-Arctic tundra

In the Arctic tundra, predators face recurrent periods of food scarcity and often turn to ungulate carcasses as an alternative food source. As important and localized resource patches, carrion promotes co-occurrence of different individuals, and its use by predators is likely to be affected by interspecific competition. We studied how interspecific competition and resource availability impact winter use of carrion by Arctic and red foxes in low Arctic Fennoscandia. We predicted that the presence of red foxes limits Arctic foxes' use of carrion, and that competition depends on the availability of other resources. We monitored Arctic and red fox presence at supp lied carrion using camera traps. From 2006 to 2021, between 16 and 20 cameras were active for 2 months in late winter (288 camera-winters). Using a multi-species dynamic occupancy model at a week-to-week scale, we evaluated the use of carrion by foxes while accounting for the presence of competitors, rodent availability, and supplemental feeding provided to Arctic foxes. Competition affected carrion use by increasing both species' probability to leave occupied carcasses between consecutive weeks. This increase was similar for the two species, suggesting symmetrical avoidance. Increased rodent abundance was associated with a higher probability of colonizing carrion sites for both species. For Arctic foxes, however, this increase was only observed at carcasses unoccupied by red foxes, showing greater avoidance when alternative preys are available. Supplementary feeding increased Arctic foxes' carrion use, regardless of red fox presence. Contrary to expectations, we did not find strong signs of asymmetric competition for carrion in winter, which suggests that interactions for resources at a short time scale are not necessarily aligned with interactions at the scale of the population. In addition, we found that competition for carcasses depends on the availability of other resources, suggesting that interactions between predators depend on the ecological context.

Using the multivariate Hawkes process to study interactions between multiple species from camera trap data

Interspecific interactions can influence species' activity and movement patterns. In particular, species may avoid or attract each other through reactive responses in space and/or time. However, data and methods to study such reactive interactions have remained scarce and were generally limited to two interacting species. At this time, the deployment of camera traps opens new opportunities but adapted statistical techniques are still required to analyze interaction patterns with such data. We present the multivariate Hawkes process (MHP) and show how it can be used to analyze interactions between several species using camera trap data. Hawkes processes use flexible pairwise interaction functions, allowing us to consider asymmetries and variations over time when depicting reactive temporal interactions. After describing the theoretical foundations of the MHP, we outline how its framework can be used to study interspecific interactions with camera trap data. We design a simulation study to evaluate the performance of the MHP and of another existing method to infer interactions from camera trap-like data. We also use the MHP to infer reactive interactions from real camera trap data for five species from South African savannas (impala Aepyceros melampus, greater kudu Tragelaphus strepsiceros, lion Panthera leo, blue wildebeest Connochaetes taurinus and Burchell's zebra Equus quagga burchelli). The simulation study shows that the MHP can be used as a tool to benchmark other methods of interspecific interaction inference and that this model can reliably infer interactions when enough data are considered. The analysis of real data highlights evidence of predator avoidance by prey and herbivore-herbivore attraction. Lastly, we present the advantages and limits of the MHP and discuss how it can be improved to infer attraction/avoidance patterns more reliably. As camera traps are increasingly used, the multivariate Hawkes process provides a promising framework to decipher the complexity of interactions structuring ecological communities.

Camera trapping expands the view into global biodiversity and its change

Growing threats to biodiversity demand timely, detailed information on species occurrence, diversity and abundance at large scales. Camera traps (CTs), combined with computer vision models, provide an efficient method to survey species of certain taxa with high spatio-temporal resolution. We test the potential of CTs to close biodiversity knowledge gaps by comparing CT records of terrestrial mammals and birds from the recently released Wildlife Insights platform to publicly available occurrences from many observation types in the Global Biodiversity Information Facility. In locations with CTs, we found they sampled a greater number of days (mean = 133 versus 57 days) and documented additional species (mean increase of 1% of expected mammals). For species with CT data, we found CTs provided novel documentation of their ranges (93% of mammals and 48% of birds). Countries with the largest boost in data coverage were in the historically underrepresented southern hemisphere. Although embargoes increase data providers' willingness to share data, they cause a lag in data availability. Our work shows that the continued collection and mobilization of CT data, especially when combined with data sharing that supports attribution and privacy, has the potential to offer a critical lens into biodiversity. This article is part of the theme issue 'Detecting and attributing the causes of biodiversity change: needs, gaps and solutions'.

The influence of human activity on predator-prey spatiotemporal overlap

Despite growing evidence of widespread impacts of humans on animal behaviour, our understanding of how humans reshape species interactions remains limited. Here, we present a framework that draws on key concepts from behavioural and community ecology to outline four primary pathways by which humans can alter predator-prey spatiotemporal overlap. We suggest that predator-prey dyads can exhibit similar or opposite responses to human activity with distinct outcomes for predator diet, predation rates, population demography and trophic cascades. We demonstrate how to assess these behavioural response pathways with hypothesis testing, using temporal activity data for 178 predator-prey dyads from published camera trap studies on terrestrial mammals. We found evidence for each of the proposed pathways, revealing multiple patterns of human influence on predator-prey activity and overlap. Our framework and case study highlight current challenges, gaps, and advances in linking human activity to animal behaviour change and predator-prey dynamics. By using a hypothesis-driven approach to estimate the potential for altered species interactions, researchers can anticipate the ecological consequences of human activities on whole communities.

Estimating parasite infrapopulation size given imperfect detection: Proof-of-concept with ectoparasitic fleas on prairie dogs

Parasite infrapopulation size - the population of parasites affecting a single host - is a central metric in parasitology. However, parasites are small and elusive such that imperfect detection is expected. Repeated sampling of parasites during primary sampling occasions (e.g., each host capture) informs the detection process. Here, we estimate flea (Siphonaptera) infrapopulation size on black-tailed prairie dogs (Cynomys ludovicianus, BTPDs) as a proof-of-concept for estimating parasite infrapopulations given imperfect detection. From Jun-Aug 2011, we live-trapped 299 BTPDs for a total of 573 captures on 20 plots distributed among 13 colonies at the Vermejo Park Ranch, New Mexico, USA. During each capture, an anesthetized BTPD was combed 3 times consecutively, 15 s each, to remove and count fleas. Each flea (n = 4846) was linked to the BTPD from which it was collected and assigned an encounter history ('100', '010', '001'). We analyzed the encounter histories using Huggins closed captures models, setting recapture probabilities to 0, thereby accounting for flea removal from hosts. The probability of detecting an individual flea (p) increased with Julian date; field personnel may have become more efficient at combing fleas as the field season progressed. Combined p across 3 combings equaled 0.99. Estimates of flea infrapopulation size were reasonable and followed the negative binomial distribution. Our general approach may be broadly applicable to estimating infrapopulation sizes for parasites. The utility of this approach increases as p declines but, if p is very low, inference is likely limited.

Human density modulates spatial associations among tropical forest terrestrial mammal species

The spatial aggregation of species pairs often increases with the ecological similarity of the species involved. However, the way in which environmental conditions and anthropogenic activity affect the relationship between spatial aggregation and ecological similarity remains unknown despite the potential for spatial associations to affect species interactions, ecosystem function, and extinction risk. Given that human disturbance has been shown to both increase and decrease spatial associations among species pairs, ecological similarity may have a role in mediating these patterns. Here, we test the influences of habitat diversity, primary productivity, human population density, and species' ecological similarity based on functional traits (i.e., functional trait similarity) on spatial associations among tropical forest mammals. Large mammals are highly sensitive to anthropogenic change and therefore susceptible to changes in interspecific spatial associations. Using two-species occupancy models and camera trap data, we quantified the spatial overlap of 1216 species pairs from 13 tropical forest protected areas around the world. We found that the association between ecological similarity and interspecific species associations depended on surrounding human density. Specifically, aggregation of ecologically similar species was more than an order of magnitude stronger in landscapes with the highest human density compared to those with the lowest human density, even though all populations occurred within protected areas. Human-induced changes in interspecific spatial associations have been shown to alter top-down control by predators, increase disease transmission and increase local extinction rates. Our results indicate that anthropogenic effects on the distribution of wildlife within protected areas are already occurring and that impacts on species interactions, ecosystem functions, and extinction risk warrant further investigation.

Intensive hunting changes human-wildlife relationships

Wildlife alter their behaviors in a trade-off between consuming food and fear of becoming food themselves. The risk allocation hypothesis posits that variation in the scale, intensity and longevity of predation threats can influence the magnitude of antipredator behavioral responses. Hunting by humans represents a threat thought to be perceived by wildlife similar to how they perceive a top predator, although hunting intensity and duration varys widely around the world. Here we evaluate the effects of hunting pressure on wildlife by comparing how two communities of mammals under different management schemes differ in their relative abundance and response to humans. Using camera traps to survey wildlife across disturbance levels (yards, farms, forests) in similar landscapes in southern Germany and southeastern USA, we tested the prediction of the risk allocation hypothesis: that the higher intensity and longevity of hunting in Germany (year round vs 3 months, 4x higher harvest/km²/year) would reduce relative abundance of hunted species and result in a larger fear-based response to humans (i.e., more spatial and temporal avoidance). We further evaluated how changes in animal abundance and behavior would result in potential changes to ecological impacts (i.e., herbivory and predation). We found that hunted species were relatively less abundant in Germany and less associated with humans on the landscape (i.e., yards and urban areas), but did not avoid humans temporally in hunted areas while hunted species in the USA showed the opposite pattern. These results are consistent with the risk allocation hypothesis where we would expect more spatial avoidance in response to threats of longer duration (i.e., year-round hunting in Germany vs. 3-month duration in USA) and less spatial avoidance but more temporal avoidance for threats of shorter duration. The expected ecological impacts of mammals in all three habitats were quite different between countries, most strikingly due to the decreases in the relative abundance of hunted species in Germany, particularly deer, with no proportional increase in unhunted species, resulting in American yards facing the potential for 25x more herbivory than German yards. Our results suggest that the duration and intensity of managed hunting can have strong and predictable effects on animal abundance and behavior, with the potential for corresponding changes in the ecological impacts of wildlife. Hunting can be an effective tool for reducing wildlife conflict due to overabundance but may require more intensive harvest than is seen in much of North America.