Elusive mustelids-18 months in the search of near-threatened stoat (Mustela erminea) and weasel (M. nivalis) reveals low captures

Stoat (Mustela erminea) and weasel (M. nivalis) are hard to monitor as they are elusive of nature and leave few identifying marks in their surroundings. Stoat and weasel are both fully protected in Denmark and are thought to be widely distributed throughout the country. Despite this stoat and weasel were listed on the Danish Red List as Near Threatened in 2019, as their densities and population trends are unknown. Using a modified novel camera trapping device, the Double-Mostela, a wooden box comprising a tracking tunnel and two camera traps, we attempted to obtain density estimates based on identification of individual stoats and weasels. We deployed camera traps both inside Double-Mostela traps and externally in three different study areas in northern Zealand, Denmark, and tested commercial, American scent-based lures to attract stoat and weasel. We obtained very low seasonal trapping rates of weasel in two study areas, but in one study area, we obtained a seasonal trapping rate of stoat larger compared to another study using the Mostela. In one study area, both species were absent. We observed no effect of scent-based lures in attracting small mustelids compared to non-bait traps. Potential reasons behind low capture rates of weasel and stoat are suboptimal habitat placement and timing of deployment of the Double-Mostelas, land-use changes over the last 200 years, predation from larger predators, as well as unintended secondary poisoning with rodenticides. Due to the scarcity of weasel and stoat captures, we were unable to make density estimates based on identification of individuals; however, we identified potential features that could be used for identification and density estimates with more captures.

Population and Conservation Status of Bighorn Sheep in the State of Baja California, Mexico

The bighorn sheep in Mexico is classified as at-risk by the Mexican federal government. In the state of Baja California, wild sheep can be observed throughout the length of the state from the USA-Mexico border south to the Agua de Soda mountain range. This research aimed to document the historical trend of the bighorn population based on aerial surveys conducted in 1992, 1995, 1999, 2010, and 2021, and the abundance, distribution, and structure of bighorn sheep populations in Baja California, based on an aerial survey conducted from 8-14 November 2021, covering thirteen mountain ranges. The estimated sheep population in 2021 was based on the number of individuals observed; the sightability of the animals; the area sampled; and the total area of habitat available. In 30.5 flight hours, 456 bighorn sheep were observed, with an estimated population of 1697 ± 80 individuals. The observation rate was 16 sheep sighted per hour of flight, and the ram:ewe:lamb ratio was 62:100:19. When the results of the 2021 flight were compared to the results of the previous aerial surveys, there was a large variation between the data, which was related to the lack of consistency between the sampling designs used in each study. Nevertheless, a statistical test of the results of aerial surveys conducted in the state suggest that the Baja California bighorn sheep population remained stable between 1992 and 2021. This study highlights the need to standardize wild sheep aerial surveys by defining flight paths and establishing a consistent duration of flights. On the other hand, Baja California authorities should consider modifying the current conservation strategy for bighorn sheep to increase the species' population in the state by initiating community-based wildlife conservation programs in rural communities.

Evaluating genotyping-in-thousands by sequencing as a genetic monitoring tool for a climate sentinel mammal using non-invasive and archival samples

Genetic tools for wildlife monitoring can provide valuable information on spatiotemporal population trends and connectivity, particularly in systems experiencing rapid environmental change. Multiplexed targeted amplicon sequencing techniques, such as genotyping-in-thousands by sequencing (GT-seq), can provide cost-effective approaches for collecting genetic data from low-quality and quantity DNA samples, making them potentially useful for long-term wildlife monitoring using non-invasive and archival samples. Here, we developed a GT-seq panel as a potential monitoring tool for the American pika (Ochotona princeps) and evaluated its performance when applied to traditional, non-invasive, and archival samples, respectively. Specifically, we optimized a GT-seq panel (307 single nucleotide polymorphisms (SNPs)) that included neutral, sex-associated, and putatively adaptive SNPs using contemporary tissue samples (n = 77) from the Northern Rocky Mountains lineage of American pikas. The panel demonstrated high genotyping success (94.7%), low genotyping error (0.001%), and excellent performance identifying individuals, sex, relatedness, and population structure. We subsequently applied the GT-seq panel to archival tissue (n = 17) and contemporary fecal pellet samples (n = 129) collected within the Canadian Rocky Mountains to evaluate its effectiveness. Although the panel demonstrated high efficacy with archival tissue samples (90.5% genotyping success, 0.0% genotyping error), this was not the case for the fecal pellet samples (79.7% genotyping success, 28.4% genotyping error) likely due to the exceptionally low quality/quantity of recovered DNA using the approaches implemented. Overall, our study reinforced GT-seq as an effective tool using contemporary and archival tissue samples, providing future opportunities for temporal applications using historical specimens. Our results further highlight the need for additional optimization of sample and genetic data collection techniques prior to broader-scale implementation of a non-invasive genetic monitoring tool for American pikas.

An expert-based system to predict population survival rate from health data

Timely detection and understanding of causes for population decline are essential for effective wildlife management and conservation. Assessing trends in population size has been the standard approach, but we propose that monitoring population health could prove more effective. We collated data from 7 bottlenose dolphin (Tursiops truncatus) populations in the southeastern United States to develop a method for estimating survival probability based on a suite of health measures identified by experts as indices for inflammatory, metabolic, pulmonary, and neuroendocrine systems. We used logistic regression to implement the veterinary expert system for outcome prediction (VESOP) within a Bayesian analysis framework. We fitted parameters with records from 5 of the sites that had a robust network of responders to marine mammal strandings and frequent photographic identification surveys that documented definitive survival outcomes. We also conducted capture-mark-recapture (CMR) analyses of photographic identification data to obtain separate estimates of population survival rates for comparison with VESOP survival estimates. The VESOP analyses showed that multiple measures of health, particularly markers of inflammation, were predictive of 1- and 2-year individual survival. The highest mortality risk 1 year following health assessment related to low alkaline phosphatase (odds ratio [OR] = 10.2 [95% CI: 3.41-26.8]), whereas 2-year mortality was most influenced by elevated globulin (OR = 9.60 [95% CI: 3.88-22.4]); both are markers of inflammation. The VESOP model predicted population-level survival rates that correlated with estimated survival rates from CMR analyses for the same populations (1-year Pearson's r = 0.99, p = 1.52 × 10-5 ; 2-year r = 0.94, p = 0.001). Although our proposed approach will not detect acute mortality threats that are largely independent of animal health, such as harmful algal blooms, it can be used to detect chronic health conditions that increase mortality risk. Random sampling of the population is important and advancement in remote sampling methods could facilitate more random selection of subjects, obtainment of larger sample sizes, and extension of the approach to other wildlife species.

Elephants and algorithms: a review of the current and future role of AI in elephant monitoring

Artificial intelligence (AI) and machine learning (ML) present revolutionary opportunities to enhance our understanding of animal behaviour and conservation strategies. Using elephants, a crucial species in Africa and Asia's protected areas, as our focal point, we delve into the role of AI and ML in their conservation. Given the increasing amounts of data gathered from a variety of sensors like cameras, microphones, geophones, drones and satellites, the challenge lies in managing and interpreting this vast data. New AI and ML techniques offer solutions to streamline this process, helping us extract vital information that might otherwise be overlooked. This paper focuses on the different AI-driven monitoring methods and their potential for improving elephant conservation. Collaborative efforts between AI experts and ecological researchers are essential in leveraging these innovative technologies for enhanced wildlife conservation, setting a precedent for numerous other species.

Return to 1616: Multispecies Fauna Reconstruction Requires Thinking Outside the Box

Conservation translocations have become increasingly popular for 'rewilding' areas that have lost their native fauna. These multispecies translocations are complex and need to consider the requirements of each individual species as well as the influence of likely interactions among them. The Dirk Hartog Island National Park Ecological Restoration Project, Return to 1616, aspires to restore ecological function to Western Australia's largest island. Since 2012, pest animals have been eradicated, and conservation translocations of seven fauna species have been undertaken, with a further six planned. Here, we present a synthesis of the innovative approaches undertaken in restoring the former faunal assemblage of Dirk Hartog Island and the key learnings gathered as the project has progressed.

Hearing to the Unseen: AudioMoth and BirdNET as a Cheap and Easy Method for Monitoring Cryptic Bird Species

The efficient analyses of sound recordings obtained through passive acoustic monitoring (PAM) might be challenging owing to the vast amount of data collected using such technique. The development of species-specific acoustic recognizers (e.g., through deep learning) may alleviate the time required for sound recordings but are often difficult to create. Here, we evaluate the effectiveness of BirdNET, a new machine learning tool freely available for automated recognition and acoustic data processing, for correctly identifying and detecting two cryptic forest bird species. BirdNET precision was high for both the Coal Tit (Peripatus ater) and the Short-toed Treecreeper (Certhia brachydactyla), with mean values of 92.6% and 87.8%, respectively. Using the default values, BirdNET successfully detected the Coal Tit and the Short-toed Treecreeper in 90.5% and 98.4% of the annotated recordings, respectively. We also tested the impact of variable confidence scores on BirdNET performance and estimated the optimal confidence score for each species. Vocal activity patterns of both species, obtained using PAM and BirdNET, reached their peak during the first two hours after sunrise. We hope that our study may encourage researchers and managers to utilize this user-friendly and ready-to-use software, thus contributing to advancements in acoustic sensing and environmental monitoring.

Variation of hair cortisol in two herds of migratory caribou (Rangifer tarandus): implications for health monitoring

Migratory caribou (Rangifer tarandus sspp.) is an ecotype of conservation concern that is experiencing increased cumulative stressors associated with rapid climate change and development in Arctic Canada. Increasingly, hair cortisol concentrations (HCCs) are being used to monitor seasonal hypothalamic-pituitary-adrenal axis activity of ungulate populations; yet, the effect of key covariates for caribou (sex, season, sampling source, body location) are largely unknown. The objectives of this research were 4-fold: first, we assessed the impact of body location (neck, rump) sampling sites on HCC; second, we assessed key covariates (sex, sampling method, season) impacting HCCs of caribou; third, we investigated inter-population (Dolphin and Union (DU), Bluenose-East (BNE)) and inter-annual differences in HCC and fourth, we examined the association between HCCs and indices of biting insect activity on the summer range (oestrid index, mosquito index). We examined hair from 407 DU and BNE caribou sampled by harvesters or during capture-collaring operations from 2012 to 2020. Linear mixed-effect models were used to assess the effect of body location on HCC and generalized least squares regression (GLS) models were used to examine the impacts of key covariates, year and herd and indices of biting insect harassment. HCC varied significantly by body location, year, herd and source of samples (harvester vs capture). HCC was higher in samples taken from the neck and in the DU herd compared with the BNE, decreased linearly over time and was higher in captured versus hunted animals (P < 0.05). There was no difference in HCC between sexes, and indices of biting insect harassment in the previous year were not significantly associated with HCC. This study identifies essential covariates impacting the HCC of caribou that must be accounted for in sampling, monitoring and data interpretation.

An Acoustic Camera for Use on UAVs

Airborne acoustic surveillance would enable and ease several applications, including security surveillance, urban and industrial noise monitoring, rescue missions, and wildlife monitoring. Airborne surveillance with an acoustic camera mounted on an airship would provide the deployment flexibility and utility required by these applications. Nevertheless, and problematically for these applications, there is not a single acoustic camera mounted on an airship yet. We make significant advances towards solving this problem by designing and constructing an acoustic camera for direct mounting on the hull of a UAV airship. The camera consists of 64 microphones, a central processing unit, and software for data acquisition and processing dedicatedly developed for far-field low-level acoustic signal detection. We demonstrate a large-aperture mock-up camera operation on the ground, although all preparations have been made to integrate the camera onto an airship. The camera has an aperture of 2 m and has been designed for surveillance from a height up to 300 m, with a spatial resolution of 12 m.

A global community-sourced assessment of the state of conservation technology

Conservation technology holds the potential to vastly increase conservationists' ability to understand and address critical environmental challenges, but systemic constraints appear to hamper its development and adoption. Understanding of these constraints and opportunities for advancement remains limited. We conducted a global online survey of 248 conservation technology users and developers to identify perceptions of existing tools' current performance and potential impact, user and developer constraints, and key opportunities for growth. We also conducted focus groups with 45 leading experts to triangulate findings. The technologies with the highest perceived potential were machine learning and computer vision, eDNA and genomics, and networked sensors. A total of 95%, 94%, and 92% respondents, respectively, rated them as very helpful or game changers. The most pressing challenges affecting the field as a whole were competition for limited funding, duplication of efforts, and inadequate capacity building. A total of 76%, 67%, and 55% respondents, respectively, identified these as primary concerns. The key opportunities for growth identified in focus groups were increasing collaboration and information sharing, improving the interoperability of tools, and enhancing capacity for data analyses at scale. Some constraints appeared to disproportionately affect marginalized groups. Respondents in countries with developing economies were more likely to report being constrained by upfront costs, maintenance costs, and development funding (p = 0.048, odds ratio [OR] = 2.78; p = 0.005, OR = 4.23; p = 0.024, OR = 4.26), and female respondents were more likely to report being constrained by development funding and perceived technical skills (p = 0.027, OR = 3.98; p = 0.048, OR = 2.33). To our knowledge, this is the first attempt to formally capture the perspectives and needs of the global conservation technology community, providing foundational data that can serve as a benchmark to measure progress. We see tremendous potential for this community to further the vision they define, in which collaboration trumps competition; solutions are open, accessible, and interoperable; and user-friendly processing tools empower the rapid translation of data into conservation action. Article impact statement: Addressing financing, coordination, and capacity-building constraints is critical to the development and adoption of conservation technology.

A review of spatial capture-recapture: Ecological insights, limitations, and prospects

First described by Efford (2004), spatial capture-recapture (SCR) has become a popular tool in ecology. Like traditional capture-recapture, SCR methods account for imperfect detection when estimating ecological parameters. In addition, SCR methods use the information inherent in the spatial configuration of individual detections, thereby allowing spatially explicit estimation of population parameters, such as abundance, survival, and recruitment. Paired with advances in noninvasive survey methods, SCR has been applied to a wide range of species across different habitats, allowing for population- and landscape-level inferences with direct consequences for conservation and management. I conduct a literature review of SCR studies published since the first description of the method and provide an overview of their scope in terms of the ecological questions answered with this tool, taxonomic groups targeted, geography, spatio-temporal extent of analyses, and data collection methods. In addition, I review approaches for analytical implementation and provide an overview of parameters targeted by SCR studies and conclude with current limitations and future directions in SCR methods.

A Comprehensive Overview of Technologies for Species and Habitat Monitoring and Conservation

The range of technologies currently used in biodiversity conservation is staggering, with innovative uses often adopted from other disciplines and being trialed in the field. We provide the first comprehensive overview of the current (2020) landscape of conservation technology, encompassing technologies for monitoring wildlife and habitats, as well as for on-the-ground conservation management (e.g., fighting illegal activities). We cover both established technologies (routinely deployed in conservation, backed by substantial field experience and scientific literature) and novel technologies or technology applications (typically at trial stage, only recently used in conservation), providing examples of conservation applications for both types. We describe technologies that deploy sensors that are fixed or portable, attached to vehicles (terrestrial, aquatic, or airborne) or to animals (biologging), complemented with a section on wildlife tracking. The last two sections cover actuators and computing (including web platforms, algorithms, and artificial intelligence).

Low-frequency sampling rates are effective to record bottlenose dolphins

Acoustic monitoring in cetacean studies is an effective but expensive approach. This is partly because of the high sampling rate required by acoustic devices when recording high-frequency echolocation clicks. However, the proportion of echolocation clicks recorded at different frequencies is unknown for many species, including bottlenose dolphins. Here, we investigated the echolocation clicks of two subspecies of bottlenose dolphins in the western South Atlantic Ocean. The possibility of recording echolocation clicks at 24 and 48 kHz was assessed by two approaches. First, we considered the clicks in the frequency range up to 96 kHz. We found a loss of 0.95-13.90% of echolocation clicks in the frequency range below 24 kHz, and 0.01-0.42% below 48 kHz, to each subspecies. Then, we evaluated these recordings downsampled at 48 and 96 kHz and confirmed that echolocation clicks are recorded at these lower frequencies, with some loss. Therefore, despite reaching high frequencies, the clicks can also be recorded at lower frequencies because echolocation clicks from bottlenose dolphins are broadband. We concluded that ecological studies based on the presence-absence data are still effective for bottlenose dolphins when acoustic devices with a limited sampling rate are used.

Seismic localization of elephant rumbles as a monitoring approach

African elephants (Loxodonta africana) are sentient and intelligent animals that use a variety of vocalizations to greet, warn or communicate with each other. Their low-frequency rumbles propagate through the air as well as through the ground and the physical properties of both media cause differences in frequency filtering and propagation distances of the respective wave. However, it is not well understood how each mode contributes to the animals' abilities to detect these rumbles and extract behavioural or spatial information. In this study, we recorded seismic and co-generated acoustic rumbles in Kenya and compared their potential use to localize the vocalizing animal using the same multi-lateration algorithms. For our experimental set-up, seismic localization has higher accuracy than acoustic, and bimodal localization does not improve results. We conclude that seismic rumbles can be used to remotely monitor and even decipher elephant social interactions, presenting us with a tool for far-reaching, non-intrusive and surprisingly informative wildlife monitoring.

Automated location invariant animal detection in camera trap images using publicly available data sources

A time-consuming challenge faced by camera trap practitioners is the extraction of meaningful data from images to inform ecological management. An increasingly popular solution is automated image classification software. However, most solutions are not sufficiently robust to be deployed on a large scale due to lack of location invariance when transferring models between sites. This prevents optimal use of ecological data resulting in significant expenditure of time and resources to annotate and retrain deep learning models.We present a method ecologists can use to develop optimized location invariant camera trap object detectors by (a) evaluating publicly available image datasets characterized by high intradataset variability in training deep learning models for camera trap object detection and (b) using small subsets of camera trap images to optimize models for high accuracy domain-specific applications.We collected and annotated three datasets of images of striped hyena, rhinoceros, and pigs, from the image-sharing websites FlickR and iNaturalist (FiN), to train three object detection models. We compared the performance of these models to that of three models trained on the Wildlife Conservation Society and Camera CATalogue datasets, when tested on out-of-sample Snapshot Serengeti datasets. We then increased FiN model robustness by infusing small subsets of camera trap images into training.In all experiments, the mean Average Precision (mAP) of the FiN trained models was significantly higher (82.33%-88.59%) than that achieved by the models trained only on camera trap datasets (38.5%-66.74%). Infusion further improved mAP by 1.78%-32.08%.Ecologists can use FiN images for training deep learning object detection solutions for camera trap image processing to develop location invariant, robust, out-of-the-box software. Models can be further optimized by infusion of 5%-10% camera trap images into training data. This would allow AI technologies to be deployed on a large scale in ecological applications. Datasets and code related to this study are open source and available on this repository: https://doi.org/10.5061/dryad.1c59zw3tx.

U-Infuse: Democratization of Customizable Deep Learning for Object Detection

Image data is one of the primary sources of ecological data used in biodiversity conservation and management worldwide. However, classifying and interpreting large numbers of images is time and resource expensive, particularly in the context of camera trapping. Deep learning models have been used to achieve this task but are often not suited to specific applications due to their inability to generalise to new environments and inconsistent performance. Models need to be developed for specific species cohorts and environments, but the technical skills required to achieve this are a key barrier to the accessibility of this technology to ecologists. Thus, there is a strong need to democratize access to deep learning technologies by providing an easy-to-use software application allowing non-technical users to train custom object detectors. U-Infuse addresses this issue by providing ecologists with the ability to train customised models using publicly available images and/or their own images without specific technical expertise. Auto-annotation and annotation editing functionalities minimize the constraints of manually annotating and pre-processing large numbers of images. U-Infuse is a free and open-source software solution that supports both multiclass and single class training and object detection, allowing ecologists to access deep learning technologies usually only available to computer scientists, on their own device, customised for their application, without sharing intellectual property or sensitive data. It provides ecological practitioners with the ability to (i) easily achieve object detection within a user-friendly GUI, generating a species distribution report, and other useful statistics, (ii) custom train deep learning models using publicly available and custom training data, (iii) achieve supervised auto-annotation of images for further training, with the benefit of editing annotations to ensure quality datasets. Broad adoption of U-Infuse by ecological practitioners will improve ecological image analysis and processing by allowing significantly more image data to be processed with minimal expenditure of time and resources, particularly for camera trap images. Ease of training and use of transfer learning means domain-specific models can be trained rapidly, and frequently updated without the need for computer science expertise, or data sharing, protecting intellectual property and privacy.

Limitations, lack of standardization, and recommended best practices in studies of renewable energy effects on birds and bats

Increasing global energy demand is fostering the development of renewable energy as an alternative to fossil fuels. However, renewable energy facilities may adversely affect wildlife. Facility siting guidelines recommend or require project developers complete pre- and postconstruction wildlife surveys to predict risk and estimate effects of proposed projects. Despite this, there are no published studies that have quantified the types of surveys used or how survey types are standardized within and across facilities. We evaluated 628 peer-reviewed publications, unpublished reports, and citations, and we analyzed data from 525 of these sources (203 facilities: 193 wind and 10 solar) in the United States and Canada to determine the frequency of pre- and postconstruction surveys and whether that frequency changed over time; frequency of studies explicitly designed to allow before-after or impact-control analyses; and what types of survey data were collected during pre- and postconstruction periods and how those data types were standardized across periods and among facilities. Within our data set, postconstruction monitoring for wildlife fatalities and habitat use was a standard practice (n = 446 reports), but preconstruction estimation of baseline wildlife habitat use and mortality was less frequently reported (n = 84). Only 22% (n = 45) of the 203 facilities provided data from both pre- and postconstruction, and 29% (n = 59) had experimental study designs. Of 108 facilities at which habitat-use surveys were conducted, only 3% estimated of detection probability. Thus, the available data generally preclude comparison of biological data across construction periods and among facilities. Use of experimental study designs and following similar field protocols would improve the knowledge of how renewable energy affects wildlife. Article Impact Statement Many surveys at wind and solar facilities provide limited information on wildlife use and fatality rates.

Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears

Emerging technologies support a new era of applied wildlife research, generating data on scales from individuals to populations. Computer vision methods can process large datasets generated through image-based techniques by automating the detection and identification of species and individuals. With the exception of primates, however, there are no objective visual methods of individual identification for species that lack unique and consistent body markings. We apply deep learning approaches of facial recognition using object detection, landmark detection, a similarity comparison network, and an support vector machine-based classifier to identify individuals in a representative species, the brown bear Ursus arctos. Our open-source application, BearID, detects a bear's face in an image, rotates and extracts the face, creates an "embedding" for the face, and uses the embedding to classify the individual. We trained and tested the application using labeled images of 132 known individuals collected from British Columbia, Canada, and Alaska, USA. Based on 4,674 images, with an 80/20% split for training and testing, respectively, we achieved a facial detection (ability to find a face) average precision of 0.98 and an individual classification (ability to identify the individual) accuracy of 83.9%. BearID and its annotated source code provide a replicable methodology for applying deep learning methods of facial recognition applicable to many other species that lack distinguishing markings. Further analyses of performance should focus on the influence of certain parameters on recognition accuracy, such as age and body size. Combining BearID with camera trapping could facilitate fine-scale behavioral research such as individual spatiotemporal activity patterns, and a cost-effective method of population monitoring through mark-recapture studies, with implications for species and landscape conservation and management. Applications to practical conservation include identifying problem individuals in human-wildlife conflicts, and evaluating the intrapopulation variation in efficacy of conservation strategies, such as wildlife crossings.

Acoustic localization of terrestrial wildlife: Current practices and future opportunities

Autonomous acoustic recorders are an increasingly popular method for low-disturbance, large-scale monitoring of sound-producing animals, such as birds, anurans, bats, and other mammals. A specialized use of autonomous recording units (ARUs) is acoustic localization, in which a vocalizing animal is located spatially, usually by quantifying the time delay of arrival of its sound at an array of time-synchronized microphones. To describe trends in the literature, identify considerations for field biologists who wish to use these systems, and suggest advancements that will improve the field of acoustic localization, we comprehensively review published applications of wildlife localization in terrestrial environments. We describe the wide variety of methods used to complete the five steps of acoustic localization: (1) define the research question, (2) obtain or build a time-synchronizing microphone array, (3) deploy the array to record sounds in the field, (4) process recordings captured in the field, and (5) determine animal location using position estimation algorithms. We find eight general purposes in ecology and animal behavior for localization systems: assessing individual animals' positions or movements, localizing multiple individuals simultaneously to study their interactions, determining animals' individual identities, quantifying sound amplitude or directionality, selecting subsets of sounds for further acoustic analysis, calculating species abundance, inferring territory boundaries or habitat use, and separating animal sounds from background noise to improve species classification. We find that the labor-intensive steps of processing recordings and estimating animal positions have not yet been automated. In the near future, we expect that increased availability of recording hardware, development of automated and open-source localization software, and improvement of automated sound classification algorithms will broaden the use of acoustic localization. With these three advances, ecologists will be better able to embrace acoustic localization, enabling low-disturbance, large-scale collection of animal position data.

Development and validation of protein biomarkers of health in grizzly bears

Large carnivores play critical roles in the maintenance and function of natural ecosystems; however, the populations of many of these species are in decline across the globe. Therefore, there is an urgent need to develop novel techniques that can be used as sensitive conservation tools to detect new threats to the health of individual animals well in advance of population-level effects. Our study aimed to determine the expression of proteins related to energetics, reproduction and stress in the skin of grizzly bears (Ursus arctos) using a liquid chromatography and multiple reaction monitoring mass spectrometry assay. We hypothesized that a suite of target proteins could be measured using this technique and that the expression of these proteins would be associated with biological (sex, age, sample location on body) and environmental (geographic area, season, sample year) variables. Small skin biopsies were collected from free-ranging grizzly bears in Alberta, Canada, from 2013 to 2019 (n = 136 samples from 111 individuals). Over 700 proteins were detected in the skin of grizzly bears, 19 of which were chosen as targets because of their established roles in physiological function. Generalized linear mixed model analysis was used for each target protein. Results indicate that sample year influenced the majority of proteins, suggesting that physiological changes may be driven in part by responses to changes in the environment. Season influenced the expression of proteins related to energetics, reproduction and stress, all of which were lower during fall compared to early spring. The expression of proteins related to energetics and stress varied by geographic area, while the majority of proteins that were affected by biological attributes (age class, sex and age class by sex interaction) were related to reproduction and stress. This study provides a novel method by which scientists and managers can further assess and monitor physiological function in wildlife.

ClassifyMe: A Field-Scouting Software for the Identification of Wildlife in Camera Trap Images

We present ClassifyMe a software tool for the automated identification of animal species from camera trap images. ClassifyMe is intended to be used by ecologists both in the field and in the office. Users can download a pre-trained model specific to their location of interest and then upload the images from a camera trap to a laptop or workstation. ClassifyMe will identify animals and other objects (e.g., vehicles) in images, provide a report file with the most likely species detections, and automatically sort the images into sub-folders corresponding to these species categories. False Triggers (no visible object present) will also be filtered and sorted. Importantly, the ClassifyMe software operates on the user's local machine (own laptop or workstation)-not via internet connection. This allows users access to state-of-the-art camera trap computer vision software in situ, rather than only in the office. The software also incurs minimal cost on the end-user as there is no need for expensive data uploads to cloud services. Furthermore, processing the images locally on the users' end-device allows them data control and resolves privacy issues surrounding transfer and third-party access to users' datasets.

Density-dependent space use affects interpretation of camera trap detection rates

Camera traps (CTs) are an increasingly popular tool for wildlife survey and monitoring. Estimating relative abundance in unmarked species is often done using detection rate as an index of relative abundance, which assumes that detection rate has a positive linear relationship with true abundance. This assumption may be violated if movement behavior varies with density, but the degree to which movement behavior is density-dependent across taxa is unclear. The potential confounding of population-level relative abundance indices by movement would depend on how regularly, and by what magnitude, movement rate and home-range size vary with density. We conducted a systematic review and meta-analysis to quantify relationships between movement rate, home-range size, and density, across terrestrial mammalian taxa. We then simulated animal movements and CT sampling to test the effect of contrasting movement scenarios on CT detection rate indices. Overall, movement rate and home-range size were negatively correlated with density and positively correlated with one another. The strength of the relationships varied significantly between taxa and populations. In simulations, detection rates were related to true abundance but underestimated change, particularly for slower moving species with small home ranges. In situations where animal space use changes markedly with density, we estimate that up to thirty percent of a true change in relative abundance may be missed due to the confounding effect of movement, making trend estimation more difficult. The common assumption that movement remains constant across densities is therefore violated across a wide range of mammal species. When studying unmarked species using CT detection rates, researchers and managers should explicitly consider that such indices of relative abundance reflect both density and movement. Practitioners interpreting changes in camera detection rates should be aware that observed differences may be biased low relative to true changes in abundance. Further information on animal movement, or methods that do not depend on assumptions of density-independent movement, may be required to make robust inferences on population trends.

Design patterns for wildlife-related camera trap image analysis

This paper describes and explains design patterns for software that supports how analysts can efficiently inspect and classify camera trap images for wildlife-related ecological attributes. Broadly speaking, a design pattern identifies a commonly occurring problem and a general reusable design approach to solve that problem. A developer can then use that design approach to create a specific software solution appropriate to the particular situation under consideration. In particular, design patterns for camera trap image analysis by wildlife biologists address solutions to commonly occurring problems they face while inspecting a large number of images and entering ecological data describing image attributes. We developed design patterns for image classification based on our understanding of biologists' needs that we acquired over 8 years during development and application of the freely available Timelapse image analysis system. For each design pattern presented, we describe the problem, a design approach that solves that problem, and a concrete example of how Timelapse addresses the design pattern. Our design patterns offer both general and specific solutions related to: maintaining data consistency, efficiencies in image inspection, methods for navigating between images, efficiencies in data entry including highly repetitious data entry, and sorting and filtering image into sequences, episodes, and subsets. These design patterns can inform the design of other camera trap systems and can help biologists assess how competing software products address their project-specific needs along with determining an efficient workflow.

Wildlife surveillance using deep learning methods

Wildlife conservation and the management of human-wildlife conflicts require cost-effective methods of monitoring wild animal behavior. Still and video camera surveillance can generate enormous quantities of data, which is laborious and expensive to screen for the species of interest. In the present study, we describe a state-of-the-art, deep learning approach for automatically identifying and isolating species-specific activity from still images and video data.We used a dataset consisting of 8,368 images of wild and domestic animals in farm buildings, and we developed an approach firstly to distinguish badgers from other species (binary classification) and secondly to distinguish each of six animal species (multiclassification). We focused on binary classification of badgers first because such a tool would be relevant to efforts to manage Mycobacterium bovis (the cause of bovine tuberculosis) transmission between badgers and cattle.We used two deep learning frameworks for automatic image recognition. They achieved high accuracies, in the order of 98.05% for binary classification and 90.32% for multiclassification. Based on the deep learning framework, a detection process was also developed for identifying animals of interest in video footage, which to our knowledge is the first application for this purpose.The algorithms developed here have wide applications in wildlife monitoring where large quantities of visual data require screening for certain species.

Snap happy: camera traps are an effective sampling tool when compared with alternative methods

Camera traps have become a ubiquitous tool in ecology and conservation. They are routinely deployed in wildlife survey and monitoring work, and are being advocated as a tool for planetary-scale biodiversity monitoring. The camera trap's widespread adoption is predicated on the assumption of its effectiveness, but the evidence base for this is lacking. Using 104 past studies, we recorded the qualitative overall recommendations made by study authors (for or against camera traps, or ambiguous), together with quantitative data on the effectiveness of camera traps (e.g. number of species detected or detection probabilities) relative to 22 other methods. Most studies recommended the use of camera traps overall and they were 39% more effective based on the quantitative data. They were significantly more effective compared with live traps (88%) and were otherwise comparable in effectiveness to other methods. Camera traps were significantly more effective than other methods at detecting a large number of species (31% more) and for generating detections of species (91% more). This makes camera traps particularly suitable for broad-spectrum biodiversity surveys. Film camera traps were found to be far less effective than digital models, which has led to an increase in camera trap effectiveness over time. There was also evidence from the authors that the use of attractants with camera traps reduced their effectiveness (counter to their intended effect), while the quantitative data indicated that camera traps were more effective in closed than open habitats. Camera traps are a highly effective wildlife survey tool and their performance will only improve with future technological advances. The images they produce also have a range of other benefits, for example as digital voucher specimens and as visual aids for outreach. The evidence-base supports the increasing use of camera traps and underlines their suitability for meeting the challenges of global-scale biodiversity monitoring.

Animal Scanner: Software for classifying humans, animals, and empty frames in camera trap images

Camera traps are a popular tool to sample animal populations because they are noninvasive, detect a variety of species, and can record many thousands of animal detections per deployment. Cameras are typically set to take bursts of multiple photographs for each detection and are deployed in arrays of dozens or hundreds of sites, often resulting in millions of photographs per study. The task of converting photographs to animal detection records from such large image collections is daunting, and made worse by situations that generate copious empty pictures from false triggers (e.g., camera malfunction or moving vegetation) or pictures of humans. We developed computer vision algorithms to detect and classify moving objects to aid the first step of camera trap image filtering-separating the animal detections from the empty frames and pictures of humans. Our new work couples foreground object segmentation through background subtraction with deep learning classification to provide a fast and accurate scheme for human-animal detection. We provide these programs as both Matlab GUI and command prompt developed with C++. The software reads folders of camera trap images and outputs images annotated with bounding boxes around moving objects and a text file summary of results. This software maintains high accuracy while reducing the execution time by 14 times. It takes about 6 seconds to process a sequence of ten frames (on a 2.6 GHZ CPU computer). For those cameras with excessive empty frames due to camera malfunction or blowing vegetation automatically removes 54% of the false-triggers sequences without influencing the human/animal sequences. We achieve 99.58% on image-level empty versus object classification of Serengeti dataset. We offer the first computer vision tool for processing camera trap images providing substantial time savings for processing large image datasets, thus improving our ability to monitor wildlife across large scales with camera traps.

Quantifying the ecological success of a community-based wildlife conservation area in Tanzania

In Tanzania, community-based natural resource management of wildlife occurs through the creation of Wildlife Management Areas (WMAs). WMAs consist of multiple villages designating land for wildlife conservation, and sharing a portion of subsequent tourism revenues. Nineteen WMAs are currently operating, encompassing 7% of Tanzania’s land area, with 19 more WMAs planned. The ecological success or failure of WMAs for wildlife conservation has yet to be quantified. We defined ecological success in this case as significantly greater densities of wildlife and significantly lower densities of livestock in the WMA relative to the control site, after the WMA was established. We used 4 years of distance sampling surveys conducted 6 times per year for wild and domestic ungulates to quantify wildlife and livestock densities before and after the establishment and implementation of management efforts at Randilen WMA, relative to a control site on adjacent land of similar vegetation and habitat types. We documented similarity between the sites before WMA establishment, when both sites were managed by the same authority. After WMA establishment, we documented significantly higher densities of resident wildlife (giraffes and dik-diks) and lower densities of cattle in the WMA, relative to the control site, indicating short-term ecological success. Continued monitoring is necessary to determine longer-term effects, and to evaluate management decisions.

Non-invasive genetic censusing and monitoring of primate populations

Knowing the density or abundance of primate populations is essential for their conservation management and contextualizing socio-demographic and behavioral observations. When direct counts of animals are not possible, genetic analysis of non-invasive samples collected from wildlife populations allows estimates of population size with higher accuracy and precision than is possible using indirect signs. Furthermore, in contrast to traditional indirect survey methods, prolonged or periodic genetic sampling across months or years enables inference of group membership, movement, dynamics, and some kin relationships. Data may also be used to estimate sex ratios, sex differences in dispersal distances, and detect gene flow among locations. Recent advances in capture-recapture models have further improved the precision of population estimates derived from non-invasive samples. Simulations using these methods have shown that the confidence interval of point estimates includes the true population size when assumptions of the models are met, and therefore this range of population size minima and maxima should be emphasized in population monitoring studies. Innovations such as the use of sniffer dogs or anti-poaching patrols for sample collection are important to ensure adequate sampling, and the expected development of efficient and cost-effective genotyping by sequencing methods for DNAs derived from non-invasive samples will automate and speed analyses.

Nutritional state reveals complex consequences of risk in a wild predator-prey community

Animal populations are regulated by the combined effects of top-down, bottom-up and abiotic processes. Ecologists have struggled to isolate these mechanisms because their effects on prey behaviour, nutrition, security and fitness are often interrelated. We monitored how forage, non-consumptive effects (NCEs), consumptive predation and climatic conditions influenced the demography and nutritional state of a wild prey population during predator recolonization. Combined measures of nutrition, survival and population growth reveal that predators imposed strong effects on the prey population through interacting non-consumptive and consumptive effects, and forage mechanisms. Predation was directly responsible for adult survival, while declining recruitment was attributed to predation risk-sensitive foraging, manifested in poor female nutrition and juvenile recruitment. Substituting nutritional state into the recruitment model through a shared term reveals that predation risk-sensitive foraging was nearly twice as influential as summer forage conditions. Our findings provide a novel, mechanistic insight into the complex means by which predators and forage conditions affect prey populations, and point to a need for more ecological studies that integrate behaviour, nutrition and demography. This line of inquiry can provide further insight into how NCEs interactively contribute to the dynamics of terrestrial prey populations; particularly, how predation risk-sensitive foraging has the potential to stabilize predator-prey coexistence.

The changing contribution of top-down and bottom-up limitation of mesopredators during 220 years of land use and climate change

Apex predators may buffer bottom-up driven ecosystem change, as top-down suppression may dampen herbivore and mesopredator responses to increased resource availability. However, theory suggests that for this buffering capacity to be realized, the equilibrium abundance of apex predators must increase. This raises the question: will apex predators maintain herbivore/mesopredator limitation, if bottom-up change relaxes resource constraints? Here, we explore changes in mesopredator (red fox Vulpes vulpes) abundance over 220 years in response to eradication and recovery of an apex predator (Eurasian lynx Lynx lynx), and changes in land use and climate which are linked to resource availability. A three-step approach was used. First, recent data from Finland and Sweden were modelled to estimate linear effects of lynx density, land use and winter temperature on fox density. Second, lynx density, land use and winter temperature was estimated in a 22 650 km² focal area in boreal and boreo-nemoral Sweden in the years 1830, 1920, 2010 and 2050. Third, the models and estimates were used to project historic and future fox densities in the focal area. Projected fox density was lowest in 1830 when lynx density was high, winters cold and the proportion of cropland low. Fox density peaked in 1920 due to lynx eradication, a mesopredator release boosted by favourable bottom-up changes - milder winters and cropland expansion. By 2010, lynx recolonization had reduced fox density, but it remained higher than in 1830, partly due to the bottom-up changes. Comparing 1830 to 2010, the contribution of top-down limitation decreased, while environment enrichment relaxed bottom-up limitation. Future scenarios indicated that by 2050, lynx density would have to increase by 79% to compensate for a projected climate-driven increase in fox density. We highlight that although top-down limitation in theory can buffer bottom-up change, this requires compensatory changes in apex predator abundance. Hence apex predator recolonization/recovery to historical levels would not be sufficient to compensate for widespread changes in climate and land use, which have relaxed the resource constraints for many herbivores and mesopredators. Variation in bottom-up conditions may also contribute to context dependence in apex predator effects.

Conservation and monitoring of a persecuted African lion population by Maasai warriors

Although Africa has many threatened species and biological hot spots, there are few citizen science schemes, particularly in rural communities, and there has been limited evaluation of existing programs. We engaged traditional Maasai warriors (pastoralist men aged 15 to 35) in community-based conservation and demographic monitoring of a persecuted African lion (Panthera leo) population. Through direct engagement, we investigated whether a citizen science approach employing local warriors, who had no formal education, could produce reliable data on the demographics, predation, and movements of a species with which their communities have been in conflict for generations. Warriors were given benefits such as literacy training and skill enhancement and engaged in the monitoring of the lions. The trained warriors reported on lion sign across an area nearly 4000 km(2) . Scientists worked together with the warriors to verify their reports and gather observations on the lion population. Using the verified reports and collected observations, we examined our scientific knowledge relative to the lion population preceding and during the citizen science program. Our observations showed that data quality and quantity improved with the involvement and training of the participants. Furthermore, because they engaged in conservation and gained personal benefits, the participants came to appreciate a species that was traditionally their foe. We believe engaging other local communities in biodiversity conservation and monitoring may be an effective conservation approach in rural Africa.

Are we getting the full picture? Animal responses to camera traps and implications for predator studies

Camera trapping is widely used in ecological studies. It is often considered nonintrusive simply because animals are not captured or handled. However, the emission of light and sound from camera traps can be intrusive. We evaluated the daytime and nighttime behavioral responses of four mammalian predators to camera traps in road‐based, passive (no bait) surveys, in order to determine how this might affect ecological investigations. Wild dogs, European red foxes, feral cats, and spotted‐tailed quolls all exhibited behaviors indicating they noticed camera traps. Their recognition of camera traps was more likely when animals were approaching the device than if they were walking away from it. Some individuals of each species retreated from camera traps and some moved toward them, with negative behaviors slightly more common during the daytime. There was no consistent response to camera traps within species; both attraction and repulsion were observed. Camera trapping is clearly an intrusive sampling method for some individuals of some species. This may limit the utility of conclusions about animal behavior obtained from camera trapping. Similarly, it is possible that behavioral responses to camera traps could affect detection probabilities, introducing as yet unmeasured biases into camera trapping abundance surveys. These effects demand consideration when utilizing camera traps in ecological research and will ideally prompt further work to quantify associated biases in detection probabilities.

Unmanned Aerial Vehicles (UAVs) and Artificial Intelligence Revolutionizing Wildlife Monitoring and Conservation

Surveying threatened and invasive species to obtain accurate population estimates is an important but challenging task that requires a considerable investment in time and resources. Estimates using existing ground-based monitoring techniques, such as camera traps and surveys performed on foot, are known to be resource intensive, potentially inaccurate and imprecise, and difficult to validate. Recent developments in unmanned aerial vehicles (UAV), artificial intelligence and miniaturized thermal imaging systems represent a new opportunity for wildlife experts to inexpensively survey relatively large areas. The system presented in this paper includes thermal image acquisition as well as a video processing pipeline to perform object detection, classification and tracking of wildlife in forest or open areas. The system is tested on thermal video data from ground based and test flight footage, and is found to be able to detect all the target wildlife located in the surveyed area. The system is flexible in that the user can readily define the types of objects to classify and the object characteristics that should be considered during classification.

Towards an Automated Acoustic Detection System for Free Ranging Elephants

The human-elephant conflict is one of the most serious conservation problems in Asia and Africa today. The involuntary confrontation of humans and elephants claims the lives of many animals and humans every year. A promising approach to alleviate this conflict is the development of an acoustic early warning system. Such a system requires the robust automated detection of elephant vocalizations under unconstrained field conditions. Today, no system exists that fulfills these requirements. In this paper, we present a method for the automated detection of elephant vocalizations that is robust to the diverse noise sources present in the field. We evaluate the method on a dataset recorded under natural field conditions to simulate a real-world scenario. The proposed method outperformed existing approaches and robustly and accurately detected elephants. It thus can form the basis for a future automated early warning system for elephants. Furthermore, the method may be a useful tool for scientists in bioacoustics for the study of wildlife recordings.

Testing the consistency of wildlife data types before combining them: the case of camera traps and telemetry

Wildlife data gathered by different monitoring techniques are often combined to estimate animal density. However, methods to check whether different types of data provide consistent information (i.e., can information from one data type be used to predict responses in the other?) before combining them are lacking. We used generalized linear models and generalized linear mixed-effects models to relate camera trap probabilities for marked animals to independent space use from telemetry relocations using 2 years of data for fishers (Pekania pennanti) as a case study. We evaluated (1) camera trap efficacy by estimating how camera detection probabilities are related to nearby telemetry relocations and (2) whether home range utilization density estimated from telemetry data adequately predicts camera detection probabilities, which would indicate consistency of the two data types. The number of telemetry relocations within 250 and 500 m from camera traps predicted detection probability well. For the same number of relocations, females were more likely to be detected during the first year. During the second year, all fishers were more likely to be detected during the fall/winter season. Models predicting camera detection probability and photo counts solely from telemetry utilization density had the best or nearly best Akaike Information Criterion (AIC), suggesting that telemetry and camera traps provide consistent information on space use. Given the same utilization density, males were more likely to be photo-captured due to larger home ranges and higher movement rates. Although methods that combine data types (spatially explicit capture–recapture) make simple assumptions about home range shapes, it is reasonable to conclude that in our case, camera trap data do reflect space use in a manner consistent with telemetry data. However, differences between the 2 years of data suggest that camera efficacy is not fully consistent across ecological conditions and make the case for integrating other sources of space-use data.

Wireless Sensor Network deployment for monitoring wildlife passages

Wireless Sensor Networks (WSNs) are being deployed in very diverse application scenarios, including rural and forest environments. In these particular contexts, specimen protection and conservation is a challenge, especially in natural reserves, dangerous locations or hot spots of these reserves (i.e., roads, railways, and other civil infrastructures). This paper proposes and studies a WSN based system for generic target (animal) tracking in the surrounding area of wildlife passages built to establish safe ways for animals to cross transportation infrastructures. In addition, it allows target identification through the use of video sensors connected to strategically deployed nodes. This deployment is designed on the basis of the IEEE 802.15.4 standard, but it increases the lifetime of the nodes through an appropriate scheduling. The system has been evaluated for the particular scenario of wildlife monitoring in passages across roads. For this purpose, different schemes have been simulated in order to find the most appropriate network operational parameters. Moreover, a novel prototype, provided with motion detector sensors, has also been developed and its design feasibility demonstrated. Original software modules providing new functionalities have been implemented and included in this prototype. Finally, main performance evaluation results of the whole system are presented and discussed in depth.