Response of a small felid of conservation concern to habitat fragmentation

Habitat loss and fragmentation are major drivers of biodiversity loss. A key question, particularly relevant to carnivore conservation, is to which extent species are able to survive in human-modified landscapes. Currently, conservationists are concerned about the impact habitat fragmentation may have on the long-term persistence of the forest-dwelling guiña (Leopardus guigna), given the increasingly modified landscapes in which they live. Here we evaluate the effect habitat cover, fragmentation and anthropogenic pressure have on the occupancy probability for guiñas in privately-owned forest fragments. We collected camera-trap data from 100 temperate rainforest sites in Chile and used single-season occupancy modeling to evaluate the influence of 13 parameters of landscape structure/anthropogenic pressure and four parameters of detection probability on the ocurrence of guiñas. The camera-trap survey data comprised 4168 camera-trap days and 112 independent records of guiñas. Surprisingly, fragmented (defined as having a high perimeter-to-area ratio) and moderately sized habitat patches best predicted site occupancy. Occupancy also increased where habitat patches were closer to continuous forest and nearer to buildings. Our results imply that guiñas can benefit from a high degree of edge type habitats in fragmented landscapes, capable of adapting to habitat fragmentation in the proximity to large continuous forest patches. This suggests that guiñas have a broader niche than previously believed. Additionally, the guiña is tolerant of human infrastructure. Further research is required to identify potential ecological traps, long-term source-sink dynamics, and the habitat loss/fragmentation threshold beyond which guiña populations are no longer viable.     

Heterogeneous capture rates in low density populations and consequences for capture-recapture analysis of camera-trap data

Closed population capture-recapture analysis of camera-trap data has become the conventional method for estimating the abundance of individually recognisable cryptic species living at low densities, such as large felids. Often these estimates are the only information available to guide wildlife managers and conservation policy. Capture probability of the target species using camera traps is commonly heterogeneous and low. Published studies often report overall capture probabilities as low as 0.03 and fail to report on the level of heterogeneity in capture probability. We used simulations to study the effects of low and heterogeneous capture probability on the reliability of abundance estimates using the M_h jack-knife estimator within a closed-population capture-recapture framework. High heterogeneity in capture probability was associated with under- and over-estimates of true abundance. The use of biased abundance estimates could have serious conservation management consequences. We recommend that studies present capture frequencies of all sampled individuals so that policy makers can assess the reliability of the abundance estimates.     

Camera-trap studies of maned wolf density in the Cerrado and the Pantanal of Brazil

The maned wolf (Chrysocyon brachyurus) is threatened by large-scale habitat loss, in particular due to conversion to agricultural land. This is the first published study on maned wolf density and the first test of individual identification from camera-trap photographs. We present results from two Brazilian regions: the Cerrado and the Pantanal. Using capture–recapture analysis of camera-trap data, we estimated densities per 100 square kilometers of 3.64 ± 0.77 individuals at the Cerrado site and 1.56 ± 0.77 individuals at the Pantanal site. Parallel radio-telemetry studies at the Pantanal site showed that maned wolves occupied home ranges of 39–58 km² (mean = 50.3 ± 7.67 km²). Our study in the Cerrado took place in a private farm with a mixture of agricultural land and native habitats, representative of the majority of the present-day Cerrado. Whereas many other mammalian species have suffered in the region, our results show that the maned wolf may cope better with this highly fragmented landscape than one might have feared. Finally, the paper briefly compares maned wolf density with density of puma (Puma concolor) in the Pantanal site.     

The effect of anthropic pressures and elevation on the large and medium-sized terrestrial mammals of the subtropical mountain forests (Yungas) of NW Argentina

We conducted a 55-day long camera-trap survey in the Yungas subtropical forest in NW Argentina, to assess the effect of human accessibility, conservation status of the area, domestic animals and elevation on the diversity and composition of the large and medium-sized native terrestrial mammal assemblage. We deployed 24 camera-trap stations at distances of ～2 km from each other. The study area is covered by continuous forest and has its center in the small community of Acambuco, in the Acambuco Provincial Reserve. The main economic activity in the area is oil/gas exploitation. Local residents raise cattle, hunt and use timber and non-timber forest products. The human impact was indirectly measured with an accessibility cost model. We used a multiple regression ANCOVA to assess the effect of elevation (range: 628–1170 masl), accessibility, protection status (reserve vs not) and frequency of records of domestic animals on the native mammal species richness and on a nonmetric multidimensional scaling (NMDS) ordination based on the frequency of records of the native mammals recorded at >3 camera-trap stations. We recorded 15 species of native mammals. Native mammal species richness decreased with elevation. Elevation was correlated with NMDS axes. Other predictive variables had no effect on species richness or the NMDS ordination, probably as a result of the relatively narrow range of conditions assessed in this study. The effect of elevation on mammal assemblages should be considered in landscape planning processes aimed at promoting biodiversity conservation.     

Semantic region of interest and species classification in the deep neural network feature domain

In this paper, we focus on animal object detection and species classification in camera-trap images collected in highly cluttered natural scenes. Using a deep neural network (DNN) model training for animal- background image classification, we analyze the input camera-trap images to generate a multi-level visual representation of the input image. We detect semantic regions of interest for animals from this representation using k-mean clustering and graph cut in the DNN feature domain. These animal regions are then classified into animal species using multi-class deep neural network model. According the experimental results, our method achieves 99.75% accuracy for classifying animals and background and 90.89% accuracy for classifying 26 animal species on the Snapshot Serengeti dataset, outperforming existing image classification methods.     

Evaluating Methods for Counting Cryptic Carnivores

ABSTRACT Numerous techniques have been proposed to estimate carnivore abundance and density, but few have been validated against populations of known size. We used a density estimate established by intensive monitoring of a population of radiotagged leopards (Panthera pardus) with a detection probability of 1.0 to evaluate efficacy of track counts and camera-trap surveys as population estimators. We calculated densities from track counts using 2 methods and compared performance of 10 methods for calculating the effectively sampled area for camera-trapping data. Compared to our reference density (7.33 ± 0.44 leopards/100 km²), camera-trapping generally produced more accurate but less precise estimates than did track counts. The most accurate result (6.97 ± 1.88 leopards/100 km²) came from camera-trap data with a sampled area buffered by a boundary strip representing the mean maximum distance moved by leopards outside the survey area (MMDMOSA) established by telemetry. However, contrary to recent suggestions, the traditional method of using half the mean maximum distance moved from photographic recaptures did not result in gross overestimates of population density (6.56 ± 1.92 leopards/100 km²) but rather displayed the next best performance after MMDMOSA. The only track-count method comparable to reference density employed a capture-recapture framework applied to data when individuals were identified from their tracks (6.45 ± 1.43 leopards/100 km²) but the underlying assumptions of this technique limit more widespread application. Our results demonstrate that if applied correctly, camera-trap surveys represent the best balance of rigor and cost-effectiveness for estimating abundance and density of cryptic carnivore species that can be identified individually.     

The last stronghold of Muñoa’s Pampas cat (Leopardus munoai) in Argentina?

The Muñoa’s Pampas cat (Leopardus munoai) is a relatively understudied species restricted to open savannas and grasslands of Southern Brazil, Uruguay and NE Argentina, that has been recently split from the broader “Pampas cat” species complex. Only three documented records of the species have been published since the year 2000 for the Argentine portion of its distribution, a situation that led to speculations regarding its conservation status in Argentina. We conducted an intensive camera-trap survey (2,067 camera-trapping stations and 15,560 camera-trapping days) to assess the presence of the Muñoa’s Pampas cat in an area of 275.3 km² within the Iberá National Park and the adjacent Iberá National Reserve, Corrientes province, Argentina. Four records of Muñoa’s Pampas cat were obtained, representing at least two adults and one young individual. Only one camera-trapping record of this species had been previously obtained in Argentina, during a survey carried out in 2009 in the same study area, despite an important camera-trapping effort in the Corrientes province. All camera-trap records of Muñoa’s Pampas cat in Argentina are concentrated in areas of temporally flooded grasslands locally known as “Malezales”, suggesting that this habitat type is critically important for the conservation of this rare felid. The Iberá National Park and the adjacent Iberá National Reserve provide the adequate framework for the conservation of an important piece of habitat for Muñoa’s Pampas cat and constitute a stronghold for the species in Argentina.     

Bag of tricks for long-tail visual recognition of animal species in camera-trap images

Camera traps are a method for monitoring wildlife and they collect a large number of pictures. The number of images collected of each species usually follows a long-tail distribution, i.e., a few classes have a large number of instances, while a lot of species have just a small percentage. Although in most cases these rare species are the ones of interest to ecologists, they are often neglected when using deep-learning models because these models require a large number of images for the training. In this work, a simple and effective framework called Square-Root Sampling Branch (SSB) is proposed, which combines two classification branches that are trained using square-root sampling and instance sampling to improve long-tail visual recognition, and this is compared to state-of-the-art methods for handling this task: square-root sampling, class-balanced focal loss, and balanced group softmax. To achieve a more general conclusion, the methods for handling long-tail visual recognition were systematically evaluated in four families of computer vision models (ResNet, MobileNetV3, EfficientNetV2, and Swin Transformer) and four camera-trap datasets with different characteristics. Initially, a robust baseline with the most recent training tricks was prepared and, then, the methods for improving long-tail recognition were applied. Our experiments show that square-root sampling was the method that most improved the performance for minority classes by around 15%; however, this was at the cost of reducing the majority classes' accuracy by at least 3%. Our proposed framework (SSB) demonstrated itself to be competitive with the other methods and achieved the best or the second-best results for most of the cases for the tail classes; but, unlike the square-root sampling, the loss in the performance of the head classes was minimal, thus achieving the best trade-off among all the evaluated methods. Our experiments also show that Swin Transformer can achieve high performance for rare classes without applying any additional method for handling imbalance, and attains an overall accuracy of 88.76% for the WCS dataset and 94.97% for Snapshot Serengeti using a location-based training/test partition. Despite the improvement in the tail classes' performance, our experiments highlight the need for better methods for handling long-tail visual recognition in camera-trap images, since state-of-the-art approaches achieve poor performance, especially in classes with just a few training instances.     

Patch Configuration Non-linearly Affects Sediment Loss across Scales in a Grazed Catchment in North-east Australia

Abstract A principle of the cross-scale interaction (CSI) framework is that disturbance-induced landscape changes resulting in coarser-grained spatial structure may non-linearly amplify transfer processes across scales. We studied suspended sediment losses at two spatial scales (0.24 m² plots and ca. 0.25 ha hillslopes of about 140 m in length) in a semiarid savanna landscape to determine whether the spatial structure of grassy and bare soil areas introduced a non-linear amplification of sediment loss. Sediment loss rates from 0.24 m² bare plots averaged 1.527 t ha⁻¹ y⁻¹, which was 23 times the loss rate from nearby grassy plots (0.066 t ha⁻¹ y⁻¹). These rates were then extrapolated linearly to two hillslopes separated by only 200 m and having similar total grass cover, slope and soil type but differing in the spatial structure of bare soil patches. The coarse-grained hillslope had a large bare patch on its lower slope, whereas the fine-grained hillslope had no bare soil patches when quantified at a 4 m grid-cell resolution. Measured sediment loss from the fine-grained hillslope averaged 0.050 t ha⁻¹ y⁻¹, whereas the average sediment loss from the coarse-grained hillslope was 2.133 t ha⁻¹ y⁻¹. By linearly extrapolating from the plot scale, the expected sediment loss for the fine-grained hillslope was 0.066 t ha⁻¹ y⁻¹, which is similar to that observed. The expected sediment loss for the coarse-grained hillslope was 0.855 t ha⁻¹ y⁻¹, where linear extrapolation assumed a 46:54 ratio of bare to grassy plots and that the spatial arrangement of plots does not affect sediment loss processes. For the coarse-grained hillslope observed sediment loss is 2.5 times greater than that expected by linear extrapolation from the plot scale. This result indicates a cross-scale interaction related to spatial configuration of patches. We suggest that there were non-linearities in hillslope ecohydrological transfer processes (runoff, erosion) across scales due to a specific patch configuration that greatly amplified sediment loss because the pattern failed to slow runoff and retain sediment before it entered a creek. This example supports the CSI framework and indicates the importance of considering the effect of spatial structure when predicting system dynamics at different scales.     

Lead induces oxidative stress and phenotypic markers of apoptosis in Saccharomyces cerevisiae

In the present work, the mode of cell death induced by Pb in Saccharomyces cerevisiae was studied. Yeast cells Pb-exposed, up to 6 h, loss progressively the capacity to proliferate and maintained the membrane integrity evaluated by the fluorescent probes bis(1,3-dibutylbarbituric acid trimethine oxonol) and propidium iodide. Pb-induced death is an active process, requiring the participation of cellular metabolism, since the simultaneous addition of cycloheximide attenuated the loss of cell proliferation capacity. Cells exposed to Pb accumulated intracelullarly reactive oxygen species (ROS), evaluated by 2′,7′-dichlorodihydrofluorescein diacetate. The addition of ascorbic acid (a ROS scavenger) strongly reduced the oxidative stress and impaired the loss of proliferation capacity in Pb-treated cells. Pb-exposed cells displayed nuclear morphological alterations, like chromatin fragmentation, as revealed by diaminophenylindole staining. Together, the data obtained indicate that yeast cells exposition to 1 mmol/l Pb results in severe oxidative stress which can be the trigger of programmed cell death by apoptosis.     

Body size and food web structure: testing the equiprobability assumption of the cascade model

The cascade model successfuly predicts many patterns in reported food webs. A key assumption of this model is the existence of a predetermined trophic hierarchy; prey are always lower in the hierarchy than their predators. At least three studies have suggested that, in animal food webs, this hierarchy can be explained to a large extent by body size relationships. A second assumption of the standard cascade model is that trophic links not prohibited by the hierarchy occur with equal probability. Using nonparametric contingency table analyses, we tested this ”equiprobability hypothesis” in 16 published animal food webs for which the adult body masses of the species had been estimated. We found that when the hierarchy was based on body size, the equiprobability hypothesis was rejected in favor of an alternative, ”predator-dominance” hypothesis wherein the probability of a trophic link varies with the identity of the predator. Another alternative to equiprobabilty is that the probability of a trophic link depends upon the ratio of the body sizes of the two species. Using nonparametric regression and liklihood ratio tests, we show that a size-ratio based model represents a significant improvement over the cascade model. These results suggest that models with heterogeneous predation probabilities will fit food web data better than the homogeneous cascade model. They also suggest a new way to bridge the gap between static and dynamic food web models. Received: 3 February 1999 / Accepted: 26 October 1999     

Predicting establishment success for alien reptiles and amphibians: a role for climate matching

We examined data comprising 1,028 successful and 967 failed introduction records for 596 species of alien reptiles and amphibians around the world to test for factors influencing establishment success. We found significant variations between families and between genera. The number of jurisdictions where a species was introduced was a significant predictor of the probability the species had established in at least one jurisdiction. All species that had been introduced to more than 10 jurisdictions (34 species) had established at least one alien population. We also conducted more detailed quantitative comparisons for successful (69 species) and failed (116 species) introductions to three jurisdictions (Great Britain, California and Florida) to test for associations with climate match, geographic range size, and history of establishment success elsewhere. Relative to failed species, successful species had better climate matches between the jurisdiction where they were introduced and their geographic range elsewhere in the world. Successful species were also more likely to have high establishment success rates elsewhere in the world. Cross-validations indicated our full model correctly categorized establishment success with 78–80% accuracy. Our findings may guide risk assessments for the import of live alien reptiles and amphibians to reduce the rate new species establish in the wild.     

Extreme temperatures,foundation species,and abrupt ecosystem change: an example from an iconic seagrass ecosystem

Extreme climatic events can trigger abrupt and often lasting change in ecosystems via the reduction or elimination of foundation (i.e., habitat‐forming) species. However, while the frequency/intensity of extreme events is predicted to increase under climate change, the impact of these events on many foundation species and the ecosystems they support remains poorly understood. Here, we use the iconic seagrass meadows of Shark Bay, Western Australia – a relatively pristine subtropical embayment whose dominant, canopy‐forming seagrass, Amphibolis antarctica, is a temperate species growing near its low‐latitude range limit – as a model system to investigate the impacts of extreme temperatures on ecosystems supported by thermally sensitive foundation species in a changing climate. Following an unprecedented marine heat wave in late summer 2010/11, A. antarctica experienced catastrophic (>90%) dieback in several regions of Shark Bay. Animal‐borne video footage taken from the perspective of resident, seagrass‐associated megafauna (sea turtles) revealed severe habitat degradation after the event compared with a decade earlier. This reduction in habitat quality corresponded with a decline in the health status of largely herbivorous green turtles (Chelonia mydas) in the 2 years following the heat wave, providing evidence of long‐term, community‐level impacts of the event. Based on these findings, and similar examples from diverse ecosystems, we argue that a generalized framework for assessing the vulnerability of ecosystems to abrupt change associated with the loss of foundation species is needed to accurately predict ecosystem trajectories in a changing climate. This includes seagrass meadows, which have received relatively little attention in this context. Novel research and monitoring methods, such as the analysis of habitat and environmental data from animal‐borne video and data‐logging systems, can make an important contribution to this framework.     

The adaptive significance of chromatophores in the Arctic under-ice amphipod Apherusa glacialis

Solar radiation is a crucial factor governing biological processes in polar habitats. Containing harmful ultraviolet radiation (UVR), it can pose a threat for organisms inhabiting surface waters of polar oceans. The present study investigated the physiological color change in the obligate sympagic amphipod Apherusa glacialis mediated by red-brown chromatophores, which cover the body and internal organs of the species. Short-term experimental exposure to photosynthetic active radiation (PAR) led to pigment dispersal in the chromatophores, resulting in darkening of the animal. Irradiation in the PAR range (400–700 nm) was identified as the main trigger with high light intensities evoking marked responses within 15 min. After exposure to high PAR, darkness led to a slow aggregation of pigments in the cell center after 24 h. Experiments revealed no statistically significant change in coloration of the animal when exposed to different background colors nor UV radiation. Our results point to a dose- and time-dependent photoprotective role of chromatophores in the amphipod, presuming a shielding effect from harmful radiation in a dispersed state. The reversible nature of the physiological color change enables the species to adapt dynamically to prevailing light conditions and thereby minimize the cost of increased conspicuousness toward visually hunting predators.     

Population extinctions in the Italian diurnal lepidoptera: an analysis of possible causes

In depth studies of patterns of extinction are fundamental to understand species vulnerability, in particular when population extinctions are not driven by habitat loss, but related to subtle changes in habitat quality and are due to ‘unknown causes’. We used a dataset containing over 160,000 non-duplicate individual records of occurrence (referred to 280 butterflies and 43 zygenid moths), and their relative extinction data, to carry out a twofold analysis. We identified ecological preferences that influence extinction probability, and we analysed if all species were equally vulnerable to the same factors. Our analyses revealed that extinctions were non-randomly distributed in space and time, as well as across species. Most of the extinctions were recorded in 1901–1950 and, as expected, populations at their range edges were more prone to become extinct for non-habitat-related causes. Ecological traits were not only unequally distributed between extinction and non-extinction events, but also not all ecological features had the same importance in driving population vulnerability. Hygrophilous and nemoral species were the most likely to experience population losses and the most prone to disappear even when their habitat remained apparently unchanged. Species vulnerability depends on both ecological requirements and threat type: in fact, each species showed a distinct pattern of vulnerability, depending on threats. We concluded that the analysis may be an important step to prevent butterfly declines: species that are strongly suffering due to ‘unknown changes’ are in clear and urgent need of more detailed auto-ecological studies.     

Quantifying site-level usage and certainty of absence for an invasive species through occupancy analysis of camera-trap data

Efficient implementation of management programs for invasive species depends on accurate surveillance for guiding prioritization of surveillance and control resources in space and time. Occupancy probabilities can be used to determine where surveillance should occur. Conversely, knowledge of the certainty of site-level absence is of special interest in situations where the objective is to completely remove populations despite substantial risk of re-invasion. Indeed, the decision to shift from emphasizing control activities over the full range to emphasizing reinvasion prevention, surveillance, and response near the borders, depends on accurate knowledge of absence across space. We used a dynamic occupancy model to monitor changes in the distribution of an invasive species, feral swine (Sus scrofa), based on camera-trap data collected as part of a management program from June 2014 to January 2016 in San Diego County, California. Site usage of feral swine declined overall. The most informative predictors of site usage were spatial (latitude and longitude). Site-level non-usage rates increased over time and in response to management removal efforts; and site-level usage rates were heavily impacted by having neighboring sites that were used. Combining the detection probability estimated from the occupancy model and Bayes Theorem, we demonstrated how certainty of local (site-level) absence can be estimated iteratively in time in areas with negative surveillance (no detections) data. Our framework provides a means for using management-based surveillance data to quantify certainty of site-level absence of an invasive species, allowing for adaptive prioritization of surveillance and control resources. Our approach is flexible for application to other species and types of surveillance (e.g., track-plates, eDNA).     

Wildlife Trade and the Emergence of Infectious Diseases

Most recent emerging infectious diseases have been zoonotic in origin. It is our contention that one of the factors responsible for such emergence is the trade in wildlife and bushmeat in particular. This article considers the effect of increasing diversity in the species hunted on the probability of global epidemics such as SARS. In particular, we develop a mathematical model of the probability of such an outbreak in terms of the number of species hunted, the number of susceptibles, and the rate of contact. Hence, we postulate that local biodiversity loss and increasing rates of animal trafficking, and trade and transportation of animals to large cities—where there is a greater potential for person-to-person transmission—may increase the probability of such outbreaks dramatically.     

Developing an empirical approach to optimal camera-trap deployment at mammal resting sites: evidence from a longitudinal study of an otter <Emphasis Type="Italic">Lutra lutra</Emphasis> holt

The study of nocturnal mammals relies on indirect evidence or invasive methods involving capture and tagging of individuals. Indirect methods are prone to error, while capture and tagging mammals have logistical and ethical considerations. Off-the-shelf camera traps are perceived as an accessible, non-intrusive method for direct data gathering, having many benefits but also potential biases. Here, using a 6-year camera-trap study of a Eurasian otter holt (den), we evaluate key parameters of study design. First, we analyse patterns of holt use in relation to researcher visits to maintain the camera traps. Then, using a dual camera-trap deployment, we compare the success of data capture from each camera-trap position in relation to the dual setup. Finally, we provide analyses to optimise minimum survey effort and camera-trap programming. Our findings indicate that otter presence and resting patterns were unaffected by the researcher visits. Results were significantly better using a close camera-trap emplacement than a distant. There was a higher frequency of otter activity at the holt during the natal and early rearing period which has implications for determining the minimum survey duration. Reducing video clip duration from 30 to 19 s would have included 95% of instances where sex could be identified, and saved 35–40% of memory storage. Peaks of otter activity were related to sunrise and sunset; exclusion of diurnal hours would have missed 11% of registrations. Camera-trap studies would benefit by adopting a similar framework of analyses in the preliminary stages or during a trial period to inform subsequent methodological refinements.     

Known knowns and unknowns in biology

Here we present a knowledge‐data framework based on the politico‐military statement by Donald Rumsfeld (below) which has, we believe, direct relevance to ecological conservation. Ecological examples of four of the identified categories are provided with discussion of the conservation risks to a species through knowledge or data loss and movement through the categories. We show that so‐called known knowns in terms of global biodiversity are not as accurately known as thought, despite 500 years or more of world‐wide collecting and recording of eukaryotic species. In addition, as fast as new species, living or fossil, are discovered (unknown unknowns), some of which have revolutionised concepts about the biology of particular taxa, meanwhile, sadly other living species are being extirpated, or are assumed to be so (unknown knowns). These often have a high probability of ultimately being rediscovered, especially if small and/or living in remote, under‐sampled regions. Furthermore, we suggest that in some cases it may be possible to predict the existence of known species in new habitats, or the existence of unknown co‐evolved animal species (known unknowns). We discuss how technological advances (e.g. molecular markers and DNA sequencing) are inflating current estimates of biodiversity by identifying the existence of cryptic species. We believe the knowledge‐data matrix provides another tool for conservation practitioners to focus data collection on bridging knowledge gaps for more effective conservation outcomes.