Estimating prey abundance and distribution from camera trap data using binomial mixture models

Visual line transect (VLT) surveys are central to the monitoring and study of marine mammals. However, for cryptic species such as deep diving cetaceans VLT surveys alone suffer from problems of low sample sizes and availability bias where animals below the surface are not available to be detected. The advent of passive acoustic monitoring (PAM) technology offers important opportunities to observe deep diving cetaceans but statistical challenges remain particularly when trying to integrate VLT and PAM data. Herein, we present a general framework to combine these data streams to estimate abundance when both surveys are conducted simultaneously. Secondarily, our approach can also be used to derive an estimate of availability bias. We outline three methods that vary in complexity and data requirements which are (1) a simple distance sampling (DS) method that treats the two datasets independently (DS-DS Method), (2) a fully integrated approach that applies a capture-mark recapture (CMR) analysis to the PAM data (CMR-DS Method) and (3) a hybrid approach that requires only a subset of the PAM CMR data (Hybrid Method). To evaluate their performance, we use simulations based on known diving and vocalizing behavior of sperm whales (Physeter macrocephalus). As a case study, we applied the Hybrid Method to data from a shipboard survey of sperm whales and compared estimates to a VLT only analysis. Simulation results demonstrated that the CMR-DS Method and Hybrid Method reduced bias by >90% for both abundance and availability bias in comparison to the simpler DS -DS Method. Overall, the CMR-DS Method was the least biased and most precise. For the case study, our application of the Hybrid Method to the sperm whale dataset produced estimates of abundance and availability bias that were comparable to estimates from the VLT only analysis but with considerably higher precision. Integrating multiple sources of data is an important goal with clear benefits. As a step towards that goal we have developed a novel framework. Results from this study are promising although challenges still remain. Future work may focus on applying this method to other deep-diving species and comparing the proposed method to other statistical approaches that aim to combine information from multiple data sources.

The Jianfengling region on southwestern Hainan Island is a sanctuary for a diverse range of wild animals. However, the exact extent of mammal species diversity and conservation status in Jianfengling remains largely unknown. Using camera-trapping data spanning from October 2020 to November 2021, this study focused on the diversity and distribution status of mammal species in Jianfengling, Hainan Tropical Rainforest National Park, China. The survey, with 41,571 camera days and 8091 independent detections, revealed 15 mammalian species belonging to 6 orders and 10 families. Among these detected species, one was categorized as Critically Endangered, and one as Near Threatened on the IUCN Red List. According to the Red List of China's Vertebrates, one was categorized as Critically Endangered, one as Endangered, three as Vulnerable, and three as Near Threatened. Additionally, one was designated China's national first-class key protected wildlife, and five were designated China's national second-class key protected wildlife. Notably, populations of the Chinese pangolin (Manis pentadactyla) were confirmed to persist in the wild of Jianfengling. Occupancy modeling results showed that the occupancy of Rhesus monkey (Macaca mulatta), Hainan muntjac (Muntiacus nigripes), common palm civet (Paradoxurus hermaphroditus), and Mainland leopard cat (Prionailurus bengalensis) increased with increasing elevation. Vegetation type would only affect the occupancy of the Rhesus monkey. Normalized Difference Vegetation Index and anthropogenic activities had no obvious effect on the occupancy of mammals. In terms of the relative abundance index (RAI) of mammals captured by camera traps, the most prevalent species identified was the Asiatic brush-tailed porcupine (Atherurus macrourus), followed by the wild boar (Sus scrofa), Pallas's squirrel (Callosciurus erythraeus), and Hainan muntjac. The monitoring also captured a significant number of images of domestic dogs and humans. These findings establish a baseline for species richness and occupancy of mammals in a typical island tropical rainforest in Hainan Tropical Rainforest National Park. We hope that our findings could be used to assess the effectiveness of conservation monitoring programs in the coming years, as species richness surveys provide an important baseline for conservation planning.

Wild boar impact on forest regeneration in the northern Apennines (Italy)

ABSTRACTAimThis study aims to describe the latitudinal pattern in plant functional groups' (PFGs') biomass and cover in Nunavik to test whether PFG attributes are better Normalized Difference Vegetation Index (NDVI) predictors than climate.LocationThe study spans a 700‐km latitudinal gradient from the lichen woodland to prostrate shrub tundra vegetation zones across Nunavik, Canada.TaxonOur analysis focuses on the following PFGs: erect and prostrate shrubs, herbaceous plants, bryophytes, and lichens.MethodsBiomass and cover data of the different PFGs were sampled in 40 sites distributed across the latitudinal gradient. NDVI data were obtained through remote sensing, while climatic, permafrost depth, and surficial deposits were derived from various databases. The PFG models were built to explore relationships between average NDVI (2016–2020) at the sampling site and ecological attributes such as PFG biomass or cover but also other variables such as surficial deposits and permafrost depth. A second series of models, the climatic models, were built using only climatic variables such as seasonal temperature and precipitation.ResultsThe most parsimonious PFG model was built with the biomass data of erect shrubs, herbaceous plants, bryophytes, and lichens and included surficial deposits and permafrost depth (R2 = 0.74). This biomass model performed better than the most parsimonious cover model (cover of erect shrubs and herbaceous, surficial deposits, permafrost depth; R2 = 0.63). However, the most parsimonious climatic model (fall temperature, annual, and winter precipitations) exhibited superior predictive power compared to the ecological ones (R2 = 0.87).ConclusionsPFG models built with PFGs aboveground biomass or cover are good predictors of NDVI of the plant formations sampled along the latitudinal gradient in Nunavik. Despite the intrinsic association between NDVI and vegetation attributes, our study emphasizes the importance of the regional climate in the control of primary productivity in Arctic and subarctic ecosystems. This study provides new insights into the interpretation of NDVI data and enhances our understanding of Arctic vegetation responses under rapid climate change. Furthermore, it underscores the balance between climatic drivers and ecological dynamics in shaping fragile Arctic ecosystems.

Conflicts between humans and wildlife, especially wild boar (Sus scrofa), have caused serious problems across the world in recent years. It is necessary to effectively control wild boar agricultural damage that may be influenced by many factors. In this study, we collected data on agricultural damage caused by wild boars from November 2009 to October 2010 using field surveys and social interviews in Taohongling National Nature Reserve of Jiangxi Province, China. We constructed models using binary logistic regression analysis to predict damage risks and to identify the factors influencing damage risks. About 8.1 % of croplands were damaged by wild boars, and the damage to rice, cotton, and other crops were not distributed based on their respective availability as shown by the result of a chi-square goodness-of-fit test. Five factors (Japanese silvergrass, soil conditions, terrain, distances to settlements, and water sources) were explained in a model based on damage area (area-based model) with the prediction accuracy being 72.1 %. In addition to these five factors, one additional factor (i.e. distance to forest edge) was retained in a model based on damage frequency (frequency-based model) with the prediction accuracy being 83.1 %. Caution is needed when we apply these two models to predict boar damage to crops, and it is recommended that both models be used in combination to predict the damage probabilities more accurately.

Mycobacterium avium subspecies paratuberculosis in fallow deer and wild boar in Spain

Modelling NDVI from decadal rainfall data in the North East Arid Zone of Nigeria

We compared the performances of the strip transect count method and the distance sampling method during colony surveys of large gulls to estimate the total number of nests. Ten colonies were surveyed by both methods. Nest detection probabilities varied from 0.519 ± 0.064 to 0.706 ± 0.049 and the average nest detection probability was 0.614 ± 0.015. Nest densities were highly variable, ranging from 77 nests/ha to 717 nests/ha. Estimates of the number of nests obtained by the strip transect count method averaged 9.3% lower than those obtained by distance sampling but by as much as 31% in some colonies. Underestimation by the strip transect counts increased at high nest densities (Kendall t = -0.556, P = 0.032). The strip transect method needed on average 6.5 observers per colony surveyed, whereas the distance sampling method required 1.4 observers per colony. In addition, the mean time spent per colony was 3 hours vs 1.7 hours for the strip transect and distance sampling methods respectively. Combining both these measures of effort, distance sampling required on average 87% less effort in the field than the strip transect method. We strongly advocate the use of distance sampling for surveys of large gull colonies.

Predation on livestock as an indicator of drastic prey decline? The indirect effects of an African swine fever epidemic on predator–prey relations in Poland

BackgroundThe wild boar (Sus scrofa) was extinct in Sweden when a few animals established in the 1970s. Over the past 35 years, the species has made a substantial comeback. In this paper, we analyse wild boar population growth using three indices of population size. We also map the legislative decisions and research prompted by the expanding population. We discuss to what extent, in the eyes of the state, the view of wild boar and the management focus has shifted over time, from a perceived pest (eradication) to scarce (conservation), overabundant (reduction/control) or somewhere in between (sustainable management).ResultsWild boar harvest started in the early 1990s with a few hundred animals annually and peaked at 161,000 in 2020/2021. The distribution now comprises most of southern Sweden. Analyses of harvest and traffic accidents involving wild boar showed that the population grew exponentially until 2010/2011, after which the increase levelled off. Thus, logistic growth models showed the best fit for the full study period. We recorded 38 legislative decisions or commissions to government agencies regarding wild boar. The first decision in 1981 was to eradicate the free-ranging population. In 1987 however, the parliament decided that wild boar is native to Sweden and should be allowed in restricted extent. Later decisions mainly concerned hunting regulations and hunting methods as direct means to increase harvest and regulate the population. Another topic, increasing in importance over time, was to facilitate the use of wild boar meat to indirectly stimulate harvest. A local outbreak of African swine fever in 2023 necessitated a stamping out strategy in the affected area. We found 44 scientific papers regarding the present free-ranging population. Topics include movements and feeding patterns, hunting, reproduction, and population development.ConclusionsThe state historically regarded wild boar as a pest to be eradicated. This changed with the decision that wild boar should be allowed in restricted extent, suggesting a conservation approach. In response to population growth, the focus shifted to means facilitating sustainable management and, lately, reducing growth. The story of wild boar in Sweden illustrates attempts to mitigate conflicts and balance interests in wildlife management.

The study aimed to determine the content of selenium (Se), zinc (Zn), copper (Cu) and cadmium (Cd) in the liver of predominantly plant-eating omnivore wild boar (Sus scrofa), predominantly meat-eating omnivore red fox (Vulpes vulpes) and herbivore red deer (Cervus elaphus), from North-Eastern Poland (Warmia and Mazury), in order to verify the distribution of these elements in the trophic pyramid. Furthermore, the study was used to assess the risk of eating venison. Samples were analyzed using atomic absorption spectrophotometry. The average concentration of Se was 3.9 (p⟨0.001) and 1.8-fold higher (p0.001) in the wild boar and red fox, respectively, in comparison to the red deer, and 2.1-fold higher in the wild boar comparing to the red fox (p⟨0.001). There was no difference in the average concentration of Zn. The average concentration of Cu was 9.3. Concentration of this element was 5.4-fold higher in red deer in comparison to red fox (p⟨0.001) and 9,34-fold higher than in wild boar (p⟨0.001). The average concentration of Cd was 1.9-fold higher in wild boar in comparison to the red fox (p⟨0.029). Correlation between Cu and Cd concentrations was also observed in the case of the red deer and red fox, while no such correlations were observed between the tested elements in the wild boar. In conclusion, the liver concentrations of these heavy metals in selected wild animas species from the hunting areas of Warmia and Mazury, do not exceed standard safe values for consumers. Moreover, the wild red deer population in North-Eastern Poland is significantly Se deficient.

Biological invasions are one of the main components of human-caused global change and their negative impact on invaded ecosystems have long been recognized. Invasive mammals, in particular, can threaten native biodiversity and cause economic impacts in the region where they are introduced, often through a wide range of conflicts with humans. Although the wild boar, Sus scrofa, is considered by the IUCN as one of the 100 invasive species most damaging to biodiversity in the world, in Argentina there have only been a few studies focused on its ecology with most of them conducted in protected areas. In this study, we evaluated the effect of several factors related with human disturbance, landscape composition, degree of fragmentation and the presence of a potential competitor and a predator on the habitat use of wild boar using data from camera traps and site-occupancy modeling. Additionally, we described the daily activity pattern of the species and we studied the level of overlap with both a potential competitor and a predator. The sampling effort totaled 7,054 camera trap days. Farm density, proportion of shrubland and proportion of grassland with bushes were the detection variables included in the most supported model whereas proportion of grassland and capture rate of the Pampas fox Lycalopex gymnocercus were the occupancy variables included in the most supported model. However, the proportion of grassland was the only variable that showed statistically significant support in the averaged model, indicating that habitat use of wild boar in this area was significantly negatively affected by the level of grass cover. Wild boars were mostly nocturnal, with more activity between 21:00 and 3:00 and a peak around midnight. Wild boars showed a high level of overlap with the activity pattern of the Pampas fox and a low overlap with the activity pattern of the puma Puma concolor. Despite wild boar being introduced in Argentina a few decades ago, this study is the first landscape-scale research carried out in an agricultural landscape in Argentina and the first one based on camera-trapping data. Our study contributes valuable information that could be used to design strategies to reduce wild boar population or to minimize the damage caused by this invasive species in Argentina.

Campylobacter infections are a public health concern and an increasingly common cause of food-borne zoonoses in the European Union. However, little is known about their spill-over from free-ranging livestock to sympatric wild ungulates, especially in regards to uncommon Campylobacter species. In this study, we aim to determine the prevalence of C. coli, C. jejuni and other C. spp. in game ungulates (wild boar Sus scrofa and Iberian ibex Capra pyrenaica) and free-ranging sympatric cattle in a National Game Reserve in NE Spain. Furthermore, we explore the extent to which Campylobacter species are shared among these co-habiting hosts. Faecal samples from Iberian ibex (n = 181) were negative for C. spp. By direct plating, two wild boars out of 150 were positive for C. coli (1.3%, 95% CI 0.16-4.73), and one was positive for C. jejuni (0.67%, 95% CI 0.02-3.66). The latter was predominant in cattle: 5.45% (n = 55, 95% CI 1.14-5.12), while C. coli was not isolated from this host. C. lanienae was the most frequent species in wild boar at 10% (95% CI 5.7-15.96), and one cow cohabiting with positive wild boars in the same canyon also carried C. lanienae. Four enrichment protocols (using Bolton or Preston broth combined with either mCCDA or CFA) were added for 172 samples (57 from wild boars, 55 cattle and 60 Iberian ibexes) to increase the number of isolates obtained allowing the detection of statistically significant differences. The prevalence of C. lanienae was statistically significantly higher in wild boar than in cattle (P < 0.01), but the prevalence of C. jejuni was higher in the latter (P = 0.045). These results suggest that wild boar and cattle carry their own predominant Campylobacter species, while Iberian ibex do not seem to play an important role in the epidemiology of Campylobacter. However, there is a potential spill-over of C. spp., and thus, further research is needed to elucidate the factors determining inter-species transmission.

The aims of the study were: 1) to detect antibodies against Toxoplasma gondii from wild boar meat; 1) establish seroprevalence of toxoplasmosis in the wild boar population; 3) establish risk factors concerned in higher possible seroprevalence; 4) to estimate the usefulness of meat juice for detection of T. gondii antibodies in wild boar. Diaphragm meat juice samples from 656 wild boar (Sus scrofa) were collected during the hunting seasons between September 2008 - October 2010 from 9 districts of the Czech Republic. The samples were stratified per age category into 2 groups: piglets (n = 279) and yearlings together with adults (n = 377). The in-house ELISA test was used for the detection of antibodies against T. gondii from the meat juice samples. Antibodies against T. gondii were detected by in-house ELISA in 260 of 656 wild boars (40%) with 26% prevalence in piglets (72/279) and 50% prevalence in yearlings and adults (188/377). The district total seroprevalences ranged between 32% - 59%, with a significantly higher prevalence in the district of Havlíčkův Brod (59%). Statistically significant differences (p-value < 0.05) were found between 2 age categories, and between 9 districts, with a significant variability in the district of Havlíčkův Brod. Seroprevalence correlated positively with farm density, but without any statistical significance. The obtained results indicate that consumption of raw or undercooked meat from wild boars can carry an important risk of toxoplasma infection. Post mortem detection of antibodies in meat juice samples using ELISA is a useful alternative to blood serum examination. In addition, a diaphragm sample has been well-proven as a matrix sample for the contemporaneous diagnostics of trichinellosis and toxoplasmosis.

The risk of transmission of pathogens from free-ranging wild boars (Sus scrofa scrofa) to outdoor domestic pigs (S. scrofa domesticus) is of increasing concern in many European countries. We assess this risk, using Switzerland as an example. We estimated 1) the prevalence of important pathogens in wild boars and 2) the risk of interactions between wild boars and outdoor pigs. First, we tested 252 wild boars from selected areas between 2008 and 2010 for infection with Brucella spp. Bacterial prevalence was estimated to 28.8% (confidence interval [CI] 23.0-34.0) when using bacterial culture (B. suis Biovar 2) and real-time polymerase chain reaction. Antibody prevalence was 35.8% (CI 30.0-42.0), which was significantly higher than in previous studies in Switzerland. We also tested 233 wild boars for porcine reproductive and respiratory syndrome virus (PRRSV). Antibody prevalence was 0.43% (CI 0.01-2.4) for EU-PRRSV and real-time reverse transcription polymerase chain reaction results were negative. These findings suggest that B. suis is increasingly widespread in wild boars and PRRSV is currently not of concern. Second, we documented the spatial overlap between free-ranging wild boars and outdoor piggeries by mapping data on their respective occurrence. Wild boars are most widespread in the mountain range along the western and northern Swiss borders, while most piggeries are located in central lowlands. A risk of interaction is mainly expected at the junction between these two bioregions. This risk may increase if wild boars expand eastward and southward beyond anthropogenic barriers believed to limit their range. Therefore, we evaluated the potential of expansion of the wild boar population. Population trends suggest a continuous increase of wild boars for the past 15 yr. Surveillance of selected wildlife passages using cameras on highways and main roads indicates that these barriers are permeable (average of up to 13 wild boar crossings per 100 days). Thus an increase of wild boar range should be considered. There may be a risk of B. suis spillover from wild boars in Switzerland, which could increase in the future. Data on the occurrence of interactions between pigs and wild boars are needed to assess this risk.