REVIEW: Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes

Journal of Applied Ecology, Journal Year: 2015, Volume and Issue: 52(3), P. 675 - 685

Published: March 24, 2015

Summary Reliable assessment of animal populations is a long‐standing challenge in wildlife ecology. Technological advances have led to widespread adoption camera traps ( CT s) survey distribution, abundance and behaviour. As for any method, trapping must contend with sources sampling error such as imperfect detection. Early applications focused on density estimation naturally marked species, but there growing interest broad‐scale surveys unmarked communities. Nevertheless, inferences based detection indices are controversial, the suitability alternatives occupancy debatable. We reviewed 266 studies published between 2008 2013. recorded study objectives methodologies, evaluating consistency protocols designs, extent which considered error, linkages analytical assumptions species Nearly two‐thirds surveyed more than one majority used response variables that ignored (e.g. presence–absence, relative abundance). Many opportunistic did not explicitly report details design deployment could affect conclusions. Most estimating capture–recapture methods spatially explicit becoming prominent. Few estimated focusing instead modelling or measures abundance. While detectability, most define key components framework site) discuss potential violations model site closure). Studies using relied equal expected relationships measured responses underlying ecological processes movement). Synthesis . The rapid represents an exciting transition methodology. remain optimistic about technology's promise, call consideration abundance, movement by cameras, including thorough reporting methodological assumptions. Such transparency will facilitate efforts evaluate improve reliability trap surveys, ultimately leading stronger helping meet modern needs effective inquiry biodiversity monitoring.

Language: Английский

---

Risky business or simple solution – Relative abundance indices from camera-trapping

Biological Conservation, Journal Year: 2013, Volume and Issue: 159, P. 405 - 412

Published: Jan. 23, 2013

Language: Английский

---

Accounting for imperfect detection and survey bias in statistical analysis of presence‐only data

Global Ecology and Biogeography, Journal Year: 2014, Volume and Issue: 23(12), P. 1472 - 1484

Published: Aug. 8, 2014

Abstract Aim During the past decade ecologists have attempted to estimate parameters of species distribution models by combining locations presence observed in opportunistic surveys with spatially referenced covariates occurrence. Several statistical been proposed for analysis presence‐only data, but these largely ignored effects imperfect detection and survey bias. In this paper I describe a model‐based approach data that accounts errors individuals biased selection locations. Innovation develop hierarchical, model allows be analysed conjunction acquired independently planned surveys. One component specifies spatial within bounded, geographic region as realization point process. A second two kinds observations, encountered during Main conclusions Using mathematical proof simulation‐based comparisons, demonstrate biases induced or can reduced eliminated using hierarchical analyse counts show relatively small number high‐quality (from surveys) used leverage information which usually broad coverage may not informative both occurrence detectability individuals. Because variety sampling protocols surveys, is widely applicable. addition, since point‐process formulated at level an individual, it extended account biological interactions between temporal changes their distributions.

Language: Английский

---

Advances and applications of occupancy models

Methods in Ecology and Evolution, Journal Year: 2013, Volume and Issue: 5(12), P. 1269 - 1279

Published: July 24, 2013

Summary The past decade has seen an explosion in the development and application of models aimed at estimating species occurrence occupancy dynamics while accounting for possible non‐detection or misidentification. We discuss some recent estimation methods biological systems that motivated their development. Collectively, these offer tremendous flexibility, but simultaneously place added demands on investigator. Unlike many mark–recapture scenarios, investigators utilizing have ability, responsibility, to define sample units (i.e. sites), replicate sampling occasions, time period over which is assumed be static even criteria constitute ‘detection’ a target species. Subsequent inference interpretation model parameters depend definitions ability meet assumptions. demonstrate relevance by highlighting applications from single system (an amphibian–pathogen system) situations where use been criticized. Finally, we suggest future research

Language: Английский

---

A guide to Bayesian model checking for ecologists

Ecological Monographs, Journal Year: 2018, Volume and Issue: 88(4), P. 526 - 542

Published: May 15, 2018

Abstract Checking that models adequately represent data is an essential component of applied statistical inference. Ecologists increasingly use hierarchical Bayesian in their research. The appeal this modeling paradigm undeniable, as researchers can build and fit embody complex ecological processes while simultaneously accounting for observation error. However, ecologists tend to be less focused on checking model assumptions assessing potential lack when applying methods than more traditional modes inference such maximum likelihood. There are also multiple ways the models, each which has strengths weaknesses. For instance, P values relatively easy compute, but well known conservative, producing biased toward 0.5. Alternatively, lesser approaches checking, prior predictive checks, cross‐validation probability integral transforms, pivot discrepancy measures may produce accurate characterizations goodness‐of‐fit not ecologists. In addition, a suite visual targeted diagnostics used examine violations different at levels hierarchy, check residual temporal or spatial autocorrelation. review, we synthesize existing literature guide through many available options checking. We illustrate procedures with several case studies including (1) analysis simulated spatiotemporal count data, (2) N‐mixture estimating abundance sea otters from aircraft, (3) hidden Markov describe attendance patterns California lion mothers rookery. find commonly based posterior detect extreme inadequacy, often do subtle cases fit. Tests (including “sampled value”) appear better suited have overall performance. conclude necessary ensure scientific founded. As discovery, it should accompany most analyses presented literature.

Language: Английский

---

An empirical evaluation of camera trap study design: How many, how long and when?

Methods in Ecology and Evolution, Journal Year: 2020, Volume and Issue: 11(6), P. 700 - 713

Published: Feb. 4, 2020

Abstract Camera traps deployed in grids or stratified random designs are a well‐established survey tool for wildlife but there has been little evaluation of study design parameters. We used an empirical subsampling approach involving 2,225 camera deployments run at 41 areas around the world to evaluate three aspects trap (number sites, duration and season sampling) their influence on estimation ecological metrics (species richness, occupancy detection rate) mammals. found that 25–35 sites were needed precise estimates species depending scale study. The precision species‐level (ψ) was highly sensitive level, with <20 common (ψ > 0.75) species, more than 150 likely rare < 0.25) species. Species rates difficult estimate precisely grid level due spatial heterogeneity, presumably driven by unaccounted habitat variability factors within area. Running site 2 weeks most efficient detecting new 3–4 local rate, no gains observed after 1 month. Metrics all mammal communities seasonality, 37%–50% we examined fluctuating significantly over year. This effect pronounced temperate where seasonally varied relative abundance average factor 4–5, some completely absent one hibernation migration. recommend following guidelines efficiently obtain arrays: each 3–5 across 40–60 per array. comparisons be model based include covariates help account small‐scale variation. Furthermore, times must which could have strong impacts both tropical sites.

Language: Английский

---

A gentle introduction to camera‐trap data analysis

African Journal of Ecology, Journal Year: 2018, Volume and Issue: 56(4), P. 740 - 749

Published: Nov. 29, 2018

Abstract Camera traps are increasingly used to study wildlife ecology and inform conservation, but valid inference depends on appropriate data analysis. This article introduces the most common analytical approaches for camera‐trap data. generally as point‐based sampling devices, many methods require spatial independence of stations temporal subsequent records. Photographic rates species should be interpreted with care, because they confound abundance/use detectability. Occupancy models estimate occurrence while accounting imperfect detection can reveal species–habitat associations. Capture–recapture abundance probability from individual detection/nondetection applicable individually recognizable species. Spatial capture–recapture extends this framework by animal movement location relative trap array. is particularly useful often wide‐ranging typically studied camera presents possibilities modelling population processes. Several have been developed that cannot identified; all heavily rely model assumptions. Finally, time stamps records describe activity pattern interactions between Considering usefulness trapping, we expect ongoing development

Language: Английский

---

Spatiotemporal hierarchical modelling of species richness and occupancy using camera trap data

Journal of Applied Ecology, Journal Year: 2015, Volume and Issue: 52(2), P. 413 - 421

Published: Jan. 28, 2015

Summary Over the last two decades, a large number of camera trap surveys have been carried out around world and traps proposed as an ideal tool for inventorying monitoring medium to large‐sized terrestrial vertebrates. However, few studies analysed data at community level. We developed multi‐session multi‐species occupancy model that allows us obtain estimates species richness combining from multiple (sessions). By estimating presence session‐level modelling detection probability each sessions nested random effects, we could improve parameter session, especially with sparse data. variants our model: one was binary latent states while other used Royle–Nichols formulation relationship between abundance. applied both models eight south‐eastern Peru including six study sites, 263 stations 17 423 days. Sites covered protected areas, logging concession Brazil nut concessions. included habitat ( terra firme vs. floodplain) covariate trail off‐trail detection. Among‐camera heterogeneity serious problem variant had much better fit than binary‐state variant. Both resulted in similar showing most sites contained intact mammal communities. Detection probabilities values were more variable across within species. Three showed preference four or avoidance trails. Synthesis applications . Our provides improved set. is ideally suited integrating numbers sets investigate regional and/or temporal patterns distribution composition communities relation natural anthropogenic factors monitor over time.

Language: Английский

---

Human presence and human footprint have non-equivalent effects on wildlife spatiotemporal habitat use

Biological Conservation, Journal Year: 2019, Volume and Issue: 241, P. 108383 - 108383

Published: Dec. 19, 2019

Language: Английский

---

Disturbance type and species life history predict mammal responses to humans

Global Change Biology, Journal Year: 2021, Volume and Issue: 27(16), P. 3718 - 3731

Published: April 22, 2021

Abstract Human activity and land use change impact every landscape on Earth, driving declines in many animal species while benefiting others. Species ecological life history traits may predict success human‐dominated landscapes such that only with “winning” combinations of will persist disturbed environments. However, this link between successful coexistence humans remains obscured by the complexity anthropogenic disturbances variability among study systems. We compiled detection data for 24 mammal from 61 populations across North America to quantify effects (1) direct presence people (2) human footprint (landscape modification) occurrence levels. Thirty‐three percent exhibited a net negative response (i.e., reduced or activity) increasing and/or populations, whereas 58% were positively associated disturbance. apparent benefits tended decrease disappear at higher disturbance levels, indicative thresholds species’ capacity tolerate exploit landscapes. strong predictors their responses footprint, favoring smaller, less carnivorous, faster‐reproducing species. The positive distributed more randomly respect trait values, winners losers range body sizes dietary guilds. Differential some highlight importance considering these two forms separately when estimating impacts wildlife. Our approach provides insights into complex mechanisms through which activities shape communities globally, revealing drivers loss larger predators human‐modified

Language: Английский

---