Assessing the camera trap methodologies used to estimate density of unmarked populations

Journal of Applied Ecology, Journal Year: 2021, Volume and Issue: 58(8), P. 1583 - 1592

Published: May 17, 2021

Abstract Population density estimations are essential for wildlife management and conservation. Camera traps have become a promising cost‐effective tool, which several methods been described to estimate population when individuals unrecognizable (i.e. unmarked populations). However, comparative tests of their applicability performance scarce. Here, we compared three based on camera without individual recognition: Random Encounter Model (REM), Staying Time (REST) Distance Sampling with (CT‐DS). Comparisons were carried out in terms consistency one another, precision cost‐effectiveness. We considered six natural populations wide range densities, species different behavioural traits (red deer Cervus elaphus , wild boar Sus scrofa red fox Vulpes vulpes ). In these populations, obtained independent estimates as reference. The densities estimated ranged from 0.23 individuals/km 2 (fox) 34.87 deer). did not find significant differences values by the five but REM has tendency generate higher average than REST CT‐DS. Regarding independents’ results significantly any population, CT‐DS population. was between methods, coefficients variation 0.28 (REST), 0.36 (REM) 0.42 method required lowest human effort. Synthesis applications . Our show that all examined can work well, each having particular strengths weaknesses. Broadly, could be recommended scenarios high abundance, (CT‐DS) those low abundance while trap is optimal, it applied less risk bias. This broadens trapping estimating using information exclusively traps. strengthens case scientifically provide reference managers, including within multi‐species monitoring programmes.

Language: Английский

---

Next-Generation Camera Trapping: Systematic Review of Historic Trends Suggests Keys to Expanded Research Applications in Ecology and Conservation

Frontiers in Ecology and Evolution, Journal Year: 2021, Volume and Issue: 9

Published: Feb. 26, 2021

Camera trapping is an effective non-invasive method for collecting data on wildlife species to address questions of ecological and conservation interest. We reviewed 2,167 camera trap (CT) articles from 1994 2020. Through the lens technological diffusion, we assessed trends in: (1) CT adoption measured by published research output, (2) topic, taxonomic, geographic diversification composition applications, (3) sampling effort, spatial extent, temporal duration studies. Annual publications have grown 81-fold since 1994, increasing at a rate 1.26 (SE = 0.068) per year 2005, but with decelerating growth 2017. Topic, richness studies increased encompass 100% topics, 59.4% ecoregions, 6.4% terrestrial vertebrates. However, declines in article rates accretion plateaus Shannon's H topics major taxa studied suggest upper limits further as currently practiced. Notable compositional changes included decrease capture-recapture, recent spatial-capture-recapture, increases occupancy, interspecific interactions, automated image classification. Mammals were dominant taxon studied; within mammalian orders carnivores exhibited unimodal peak whereas primates, rodents lagomorphs steadily increased. Among biogeographic realms observed decreases Oceania Nearctic, Afrotropic Palearctic, peaks Indomalayan Neotropic. days, area sampled increased, much greater 0.90 quantile compared median. Next-generation are poised expand knowledge valuable ecology posing previously infeasible unprecedented spatiotemporal scales, array species, wider variety environments. Converting potential into broad-based application will require transferable models classification, sharing among users across multiple platforms coordinated manner. Further taxonomic likely modifications that permit more efficient smaller improvements modeling unmarked populations. Environmental can benefit engineering solutions ease traditionally challenging sites.

Language: Английский

---

Towards a best‐practices guide for camera trapping: assessing differences among camera trap models and settings under field conditions

Journal of Zoology, Journal Year: 2021, Volume and Issue: 316(3), P. 197 - 208

Published: Nov. 21, 2021

Abstract Camera trapping is a widely used tool in wildlife research and conservation, plethora of makes models camera traps have emerged. However, insufficient attention has been paid to testing their performance, particularly under field conditions. In this study, we comparatively tested five the most frequently trap (Bushnell, KeepGuard, Ltl Acorn, Reconyx Scoutguard) identify key factors behind probability detection (i.e. that successfully capturing usable photograph an animal passing through view) trigger speed time delay between instant at which motion detected, picture taken). We 45 cameras (nine devices each make) with infrared flash experiment continuous remote video was parallel (as gold‐standard) discover animals entered zone. The period (day/night), distance cameras, model, species, deployment height activation sensitivity were significantly related detection. This lower during night than day. There greater detecting given species when set its shoulder height. interaction affected speed, meaning closer zone, higher substantial differences among species. probably by movement speed. conclusion, study shows performance settings, signifying caution required making direct comparisons results obtained different experiments, or designing new ones. These provide empirical guidelines for best practices highlight relevance experiments traps.

Language: Английский

---

Improving the accessibility and transferability of machine learning algorithms for identification of animals in camera trap images: MLWIC2

Ecology and Evolution, Journal Year: 2020, Volume and Issue: 10(19), P. 10374 - 10383

Published: Sept. 16, 2020

Abstract Motion‐activated wildlife cameras (or “camera traps”) are frequently used to remotely and noninvasively observe animals. The vast number of images collected from camera trap projects has prompted some biologists employ machine learning algorithms automatically recognize species in these images, or at least filter‐out that do not contain These approaches often limited by model transferability, as a trained one location might work well for the same different locations. Furthermore, methods require advanced computational skills, making them inaccessible many biologists. We 3 million 18 studies 10 states across United States America train two deep neural networks, recognizes 58 species, “species model,” determines if an image is empty it contains animal, “empty‐animal model.” Our empty‐animal had accuracies 96.8% 97.3%, respectively. models performed on out‐of‐sample datasets, 91% accuracy Canada (accuracy range 36%–91% all datasets) achieved 91%–94% datasets continents. software addresses limitations using classify traps. By including several locations, our potentially applicable North America. also found can facilitate removal without animals globally. provide R package (MLWIC2: Machine Learning Wildlife Image Classification R), which Shiny Applications allow scientists with minimal programming experience use new six network architectures varying depths.

Language: Английский

---

Índice de abundancia relativa y tasa de encuentro con trampas cámara

Mammalogy Notes, Journal Year: 2024, Volume and Issue: 10(1), P. 389 - 389

Published: Feb. 2, 2024

El monitoreo de fauna silvestre se basa en conteos directos o indirectos animales sus rastros, unidades muestreo (cámaras, transectos, trampas, redes, grabadores, u otro). Los por unidad esfuerzo expresan como tasa encuentro, fotográfica, captura, etc. Cuando asume que la está relacionada con el tamaño poblacional, entonces es considerada un índice abundancia relativa (IAR). cuales son empleados alternativa a las estimaciones absolutas densidad. IAR utilizados para monitorear cambio una población través del tiempo, bien comparar poblaciones misma especie localidades diferentes. Con incremento uso cámaras trampa ha popularizado cálculo los todas especies fotografiadas área estudio. Sin embargo, debe tener precaución esta interpretación ya están sesgados detectabilidad varía entre especies. En este artículo 1) reviso definiciones, supuestos y limitaciones IAR; 2) explica diferencia conceptual tasas encuentro; 3) enfatiza importancia probabilidad detección factor afecta ende 4) sugiere usar solo temporal espacialmente, mientras encuentro usarla especies; 5) sugiero algunas alternativas análisis estadísticos basados modelos jerárquicos.

---

Continuous daily sampling of airborne eDNA detects all vertebrate species identified by camera traps

Environmental DNA, Journal Year: 2024, Volume and Issue: 6(4)

Published: July 1, 2024

Abstract Ongoing pressures on global biodiversity require conservation action that is not possible without effective biomonitoring. Terrestrial vertebrate surveys are commonly performed using camera traps, a time‐intensive method known to miss many small or arboreal species and birds. Recent advances have shown airborne eDNA be potentially suitable technique more effectively monitor communities in time‐ cost‐effective manner. Here, we test whether commercially available air samplers collect particles 24/7 during 1‐week period can used detect the presence of vertebrates through eDNA. The results compared trap records at three locations with differing habitats Netherlands. Simultaneous sampling different for 3 weeks resulted detection 154 taxa, which majority were birds mammals (113 33 species, respectively), along four fish amphibian species. All observed traps also retrieved via eDNA, although every day sampling. Burkard spore trap, routinely pollen monitoring, showed highest number only samples when mammal was detected it remained undetected We unique indicative habitat they living. However, could account for. multitude found indicate sensitivity method; however, subsequent studies should prioritize validation these findings alternative biomonitoring approaches.

Language: Английский

---

Camera trap-based estimates reveal spatial variability in African clawless otter population densities and behaviour

Oryx, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 14

Published: Feb. 21, 2025

Abstract Estimating the population size of shy and elusive species is challenging but necessary to inform appropriate conservation actions for threatened or declining species. Using camera-trap surveys conducted during 2017–2021, we estimated compared African clawless otter Aonyx capensis densities activity times in six conserved areas southern Africa. We used two different models estimate densities: random encounter distance sampling. Our results highlight a general pattern higher narrower confidence intervals using found substantial variation between study areas, with model estimates ranging 0.9 4.2 otters/km 2 . sampling supported relative density obtained from were generally lower more variable, 0.8 4.0 significant differences patterns, populations either being nocturnal, mostly nocturnal cathemeral. As all experience little human disturbance, our suggest that there are large natural variations patterns regions. When converted metrics comparable previous studies, numbers than previously reported. This highlights need broader spatial coverage assessments future studies assess potential environmental drivers spatial, potentially temporal, patterns.

Language: Английский

---

One size does not fit all: A novel approach for determining the Realised Viewshed Size for remote camera traps

Methods in Ecology and Evolution, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 27, 2025

Abstract Camera traps (CTs) have become cemented as an important tool of wildlife research, yet their utility is now extending beyond academics, CTs can contribute to inclusive place‐based management. From advances in analytics and technology, CT‐based density estimates are emerging field research. Most methods require estimate the size viewshed monitored by each CT, a parameter that may be highly variable difficult quantify. We developed tested standardized analytical method allowing us predict probability photographic capture it varies within CT viewshed. investigated how changes due environmental influences (vegetation structure, ambient temperature, speed subject time day), addition internal factors from themselves (sensitivity settings, number photographs taken brand). then summarize these spatial kernels into Realised Viewshed Size (RVS)—the corrected for use denominator photograph‐based Random Encounter Staying Time (REST) or Front (TIFC) estimators. found RVS values heavily influenced location‐specific structure), technological delays associated with themselves, (refractory period) settings. computed using our methodology substantially smaller than sizes reported literature. Imprecision surrounding areas propagate bias when implementing Our change practitioners consider estimators thus increasing reliability estimation, contributing more accessible management practices.

Language: Английский

---

Estimating social network metrics from single-file movements in Barbary macaques, Macaca sylvanus

Animal Behaviour, Journal Year: 2025, Volume and Issue: unknown, P. 123146 - 123146

Published: April 1, 2025

Language: Английский

---

Best practices to account for capture probability and viewable area in camera‐based abundance estimation

Remote Sensing in Ecology and Conservation, Journal Year: 2022, Volume and Issue: 9(1), P. 152 - 164

Published: Aug. 26, 2022

Abstract A suite of recently developed statistical methods to estimate the abundance and density unmarked animals from camera traps require accurate estimates area sampled by each camera. Although viewshed is fundamental achieving estimates, there are no established guidelines for collecting this information in field. Furthermore, while complexities detection process motion sensor photography generally acknowledged, viewable (the common factor between time lapse photography) on its own has been underemphasized. We establish a set terminology identify component parts area, contrast photographic capture measurements photography, review estimating use case study demonstrate importance estimates. Time combined with allow researchers assume that probability equals 1. Motion requires measuring distances animal fitting distance sampling curve account <1.

Language: Английский

---