Geology topics

J. Andrew Royle

Publications and source records attributed to J. Andrew Royle.

At least 91 records · Page 5Linked to original sources

USGS2017

Dynamic optimization of landscape connectivity embedding spatial-capture-recapture information

Maintaining landscape connectivity is increasingly important in wildlife conservation, especially for species experiencing the effects of habitat loss and fragmentation. We propose a novel approach to dynamically optimize landscape connectivity. Our approach is based on a mixed integer program formulation, embedding a spatial capture-recapture model that estimates the density, space usage, and landscape connectivity for a given species. Our method takes into account the fact that local animal density and connectivity change dynamically and non-linearly with different habitat protection plans. In order to scale up our encoding, we propose a sampling scheme via random partitioning of the search space using parity functions. We show that our method scales to realworld size problems and dramatically outperforms the solution quality of an expectation maximization approach and a sample average approximation approach.

Conference Paper↗

USGS2016

Scaling-up camera traps: monitoring the planet's biodiversity with networks of remote sensors

Countries committed to implementing the Convention on Biological Diversity's 2011–2020 strategic plan need effective tools to monitor global trends in biodiversity. Remote cameras are a rapidly growing technology that has great potential to transform global monitoring for terrestrial biodiversity and can be an important contributor to the call for measuring Essential Biodiversity Variables. Recent advances in camera technology and methods enable researchers to estimate changes in abundance and distribution for entire communities of animals and to identify global drivers of biodiversity trends. We suggest that interconnected networks of remote cameras will soon monitor biodiversity at a global scale, help answer pressing ecological questions, and guide conservation policy. This global network will require greater collaboration among remote-camera studies and citizen scientists, including standardized metadata, shared protocols, and security measures to protect records about sensitive species. With modest investment in infrastructure, and continued innovation, synthesis, and collaboration, we envision a global network of remote cameras that not only provides real-time biodiversity data but also serves to connect people with nature.

Frontiers in Ecology and the Environment↗

USGS2016

A multistate dynamic site occupancy model for spatially aggregated sessile communities

Estimation of transition probabilities of sessile communities seems easy in principle but may still be difficult in practice because resampling error (i.e. a failure to resample exactly the same location at fixed points) may cause significant estimation bias. Previous studies have developed novel analytical methods to correct for this estimation bias. However, they did not consider the local structure of community composition induced by the aggregated distribution of organisms that is typically observed in sessile assemblages and is very likely to affect observations. We developed a multistate dynamic site occupancy model to estimate transition probabilities that accounts for resampling errors associated with local community structure. The model applies a nonparametric multivariate kernel smoothing methodology to the latent occupancy component to estimate the local state composition near each observation point, which is assumed to determine the probability distribution of data conditional on the occurrence of resampling error. By using computer simulations, we confirmed that an observation process that depends on local community structure may bias inferences about transition probabilities. By applying the proposed model to a real data set of intertidal sessile communities, we also showed that estimates of transition probabilities and of the properties of community dynamics may differ considerably when spatial dependence is taken into account. Results suggest the importance of accounting for resampling error and local community structure for developing management plans that are based on Markovian models. Our approach provides a solution to this problem that is applicable to broad sessile communities. It can even accommodate an anisotropic spatial correlation of species composition, and may also serve as a basis for inferring complex nonlinear ecological dynamics.

Methods in Ecology and Evolution↗

USGS2016

Integrating occurrence and detectability patterns based on interview data: a case study for threatened mammals in Equatorial Guinea

Occurrence models that account for imperfect detection of species are increasingly used for estimating geographical range, for determining species-landscape relations and to prioritize conservation actions worldwide. In 2010, we conducted a large-scale survey in Río Muni, the mainland territory of Equatorial Guinea, which aimed to estimate the probabilities of occurrence and detection of threatened mammals based on environmental covariates, and to identify priority areas for conservation. Interviews with hunters were designed to record presence/absence data of seven species (golden cat, leopard, forest elephant, forest buffalo, western gorilla, chimpanzee and mandrill) in 225 sites throughout the region. We fitted single season occupancy models and recently developed models which also include false positive errors (i.e. species detected in places where it actually does not occur), which should provide more accurate estimates for most species, which are susceptible to mis-identification. Golden cat and leopard had the lowest occurrence rates in the region, whereas primates had the highest rates. All species, except gorilla, were affected negatively by human settlements. The southern half of Río Muni showed the highest occurrence of the species studied, and conservation strategies for ensuring the persistence of threatened mammals should be focused on this area.

Scientific Reports↗

USGS2016

Estimating species – area relationships by modeling abundance and frequency subject to incomplete sampling

Models and data used to describe species–area relationships confound sampling with ecological process as they fail to acknowledge that estimates of species richness arise due to sampling. This compromises our ability to make ecological inferences from and about species–area relationships. We develop and illustrate hierarchical community models of abundance and frequency to estimate species richness. The models we propose separate sampling from ecological processes by explicitly accounting for the fact that sampled patches are seldom completely covered by sampling plots and that individuals present in the sampling plots are imperfectly detected. We propose a multispecies abundance model in which community assembly is treated as the summation of an ensemble of species-level Poisson processes and estimate patch-level species richness as a derived parameter. We use sampling process models appropriate for specific survey methods. We propose a multispecies frequency model that treats the number of plots in which a species occurs as a binomial process. We illustrate these models using data collected in surveys of early-successional bird species and plants in young forest plantation patches. Results indicate that only mature forest plant species deviated from the constant density hypothesis, but the null model suggested that the deviations were too small to alter the form of species–area relationships. Nevertheless, results from simulations clearly show that the aggregate pattern of individual species density–area relationships and occurrence probability–area relationships can alter the form of species–area relationships. The plant community model estimated that only half of the species present in the regional species pool were encountered during the survey. The modeling framework we propose explicitly accounts for sampling processes so that ecological processes can be examined free of sampling artefacts. Our modeling approach is extensible and could be applied to a variety of study designs and allows the inclusion of additional environmental covariates.

Ecology and Evolution↗

USGS2016

Estimating population density and connectivity of American mink using spatial capture-recapture

Estimating the abundance or density of populations is fundamental to the conservation and management of species, and as landscapes become more fragmented, maintaining landscape connectivity has become one of the most important challenges for biodiversity conservation. Yet these two issues have never been formally integrated together in a model that simultaneously models abundance while accounting for connectivity of a landscape. We demonstrate an application of using capture–recapture to develop a model of animal density using a least-cost path model for individual encounter probability that accounts for non-Euclidean connectivity in a highly structured network. We utilized scat detection dogs ( Canis lupus familiaris ) as a means of collecting non-invasive genetic samples of American mink ( Neovison vison ) individuals and used spatial capture–recapture models (SCR) to gain inferences about mink population density and connectivity. Density of mink was not constant across the landscape, but rather increased with increasing distance from city, town, or village centers, and mink activity was associated with water. The SCR model allowed us to estimate the density and spatial distribution of individuals across a 388 km 2 area. The model was used to investigate patterns of space usage and to evaluate covariate effects on encounter probabilities, including differences between sexes. This study provides an application of capture–recapture models based on ecological distance, allowing us to directly estimate landscape connectivity. This approach should be widely applicable to provide simultaneous direct estimates of density, space usage, and landscape connectivity for many species.

Ecological Applications↗

USGS2016

Management decision making for fisher populations informed by occupancy modeling

Harvest data are often used by wildlife managers when setting harvest regulations for species because the data are regularly collected and do not require implementation of logistically and financially challenging studies to obtain the data. However, when harvest data are not available because an area had not previously supported a harvest season, alternative approaches are required to help inform management decision making. When distribution or density data are required across large areas, occupancy modeling is a useful approach, and under certain conditions, can be used as a surrogate for density. We collaborated with the New York State Department of Environmental Conservation (NYSDEC) to conduct a camera trapping study across a 70,096-km 2 region of southern New York in areas that were currently open to fisher ( Pekania [ Martes ] pennanti ) harvest and those that had been closed to harvest for approximately 65 years. We used detection–nondetection data at 826 sites to model occupancy as a function of site-level landscape characteristics while accounting for sampling variation. Fisher occupancy was influenced positively by the proportion of conifer and mixed-wood forest within a 15-km 2 grid cell and negatively associated with road density and the proportion of agriculture. Model-averaged predictions indicated high occupancy probabilities (>0.90) when road densities were low (<1 km/km 2 ) and coniferous and mixed forest proportions were high (>0.50). Predicted occupancy ranged 0.41–0.67 in wildlife management units (WMUs) currently open to trapping, which could be used to guide a minimum occupancy threshold for opening new areas to trapping seasons. There were 5 WMUs that had been closed to trapping but had an average predicted occupancy of 0.52 (0.07 SE), and above the threshold of 0.41. These areas are currently under consideration by NYSDEC for opening a conservative harvest season. We demonstrate the use of occupancy modeling as an aid to management decision making when harvest-related data are unavailable and when budgetary constraints do not allow for capture–recapture studies to directly estimate density.

Journal of Wildlife Management↗

USGS2016

Study of biological communities subject to imperfect detection: Bias and precision of community N -mixture abundance models in small-sample situations

Community N -mixture abundance models for replicated counts provide a powerful and novel framework for drawing inferences related to species abundance within communities subject to imperfect detection. To assess the performance of these models, and to compare them to related community occupancy models in situations with marginal information, we used simulation to examine the effects of mean abundance ( λ ¯ : 0.1, 0.5, 1, 5), detection probability ( p ¯ : 0.1, 0.2, 0.5), and number of sampling sites ( n site : 10, 20, 40) and visits ( n visit : 2, 3, 4) on the bias and precision of species-level parameters (mean abundance and covariate effect) and a community-level parameter (species richness). Bias and imprecision of estimates decreased when any of the four variables ( λ ¯ , p ¯ , n site , n visit ) increased. Detection probability p ¯ was most important for the estimates of mean abundance, while λ ¯ was most influential for covariate effect and species richness estimates. For all parameters, increasing n site was more beneficial than increasing n visit . Minimal conditions for obtaining adequate performance of community abundance models were n site ≥ 20, p¯ ≥ 0.2, and λ ¯ ≥ 0.5. At lower abundance, the performance of community abundance and community occupancy models as species richness estimators were comparable. We then used additive partitioning analysis to reveal that raw species counts can overestimate β diversity both of species richness and the Shannon index, while community abundance models yielded better estimates. Community N -mixture abundance models thus have great potential for use with community ecology or conservation applications provided that replicated counts are available.

Ecological Research↗

USGS2016

What are hierarchical models and how do we analyze them?

In this chapter we provide a basic definition of hierarchical models and introduce the two canonical hierarchical models in this book: site occupancy and N-mixture models. The former is a hierarchical extension of logistic regression and the latter is a hierarchical extension of Poisson regression. We introduce basic concepts of probability modeling and statistical inference including likelihood and Bayesian perspectives. We go through the mechanics of maximizing the likelihood and characterizing the posterior distribution by Markov chain Monte Carlo (MCMC) methods. We give a general perspective on topics such as model selection and assessment of model fit, although we demonstrate these topics in practice in later chapters (especially Chapters 5, 6, 7, and 10 Chapter 5 Chapter 6 Chapter 7 Chapter 10)

Book chapter↗

USGS2016

Modeling abundance using multinomial N-mixture models

Multinomial N-mixture models are a generalization of the binomial N-mixture models described in Chapter 6 to allow for more complex and informative sampling protocols beyond simple counts. Many commonly used protocols such as multiple observer sampling, removal sampling, and capture-recapture produce a multivariate count frequency that has a multinomial distribution and for which multinomial N-mixture models can be developed. Such protocols typically result in more precise estimates than binomial mixture models because they provide direct information about parameters of the observation process. We demonstrate the analysis of these models in BUGS using several distinct formulations that afford great flexibility in the types of models that can be developed, and we demonstrate likelihood analysis using the unmarked package. Spatially stratified capture-recapture models are one class of models that fall into the multinomial N-mixture framework, and we discuss analysis of stratified versions of classical models such as model Mb, Mh and other classes of models that are only possible to describe within the multinomial N-mixture framework.

Book chapter↗

USGS2016

Modeling abundance using hierarchical distance sampling

In this chapter, we provide an introduction to classical distance sampling ideas for point and line transect data, and for continuous and binned distance data. We introduce the conditional and the full likelihood, and we discuss Bayesian analysis of these models in BUGS using the idea of data augmentation, which we discussed in Chapter 7. We then extend the basic ideas to the problem of hierarchical distance sampling (HDS), where we have multiple point or transect sample units in space (or possibly in time). The benefit of HDS in practice is that it allows us to directly model spatial variation in population size among these sample units. This is a preeminent concern of most field studies that use distance sampling methods, but it is not a problem that has received much attention in the literature. We show how to analyze HDS models in both the unmarked package and in the BUGS language for point and line transects, and for continuous and binned distance data. We provide a case study of HDS applied to a survey of the island scrub-jay on Santa Cruz Island, California.

Book chapter↗

USGS2016

Incorporating imperfect detection into joint models of communites: A response to Warton et al.

Warton et al. [1] advance community ecology by describing a statistical framework that can jointly model abundances (or distributions) across many taxa to quantify how community properties respond to environmental variables. This framework specifies the effects of both measured and unmeasured (latent) variables on the abundance (or occurrence) of each species. Latent variables are random effects that capture the effects of both missing environmental predictors and correlations in parameter values among different species. As presented in Warton et al. , however, the joint modeling framework fails to account for the common problem of detection or measurement errors that always accompany field sampling of abundance or occupancy, and are well known to obscure species- and community-level inferences.

Trends in Ecology and Evolution↗

USGS2016

Estimating abundance: Chapter 27

This chapter provides a non-technical overview of ‘closed population capture–recapture’ models, a class of well-established models that are widely applied in ecology, such as removal sampling, covariate models, and distance sampling. These methods are regularly adopted for studies of reptiles, in order to estimate abundance from counts of marked individuals while accounting for imperfect detection. Thus, the chapter describes some classic closed population models for estimating abundance, with considerations for some recent extensions that provide a spatial context for the estimation of abundance, and therefore density. Finally, the chapter suggests some software for use in data analysis, such as the Windows-based program MARK, and provides an example of estimating abundance and density of reptiles using an artificial cover object survey of Slow Worms ( Anguis fragilis ).

Book chapter↗

USGS2016

Southeast regional and state trends in anuran occupancy from calling survey data (2001-2013) from the North American Amphibian Monitoring Program

We present the first regional trends in anuran occupancy for eight states of the southeastern United States, based on 13 y (2001–2013) of North American Amphibian Monitoring Program (NAAMP) data. The NAAMP is a longterm monitoring program in which observers collect anuran calling observation data at fixed locations along random roadside routes. We assessed occupancy trends for 14 species. We found weak evidence for a general regional pattern of decline in calling anurans within breeding habitats along roads in the southeastern USA over the last 13 y. Two species had positive regional trends with 95% posterior intervals that did not include zero ( Hyla cinerea and Pseudacris crucifer ). Five other species also showed an increasing trend, while eight species showed a declining trend, although 95% posterior intervals included zero. We also assessed state level trends for 107 species/state combinations. Of these, 14 showed a significant decline and 12 showed a significant increase in occupancy (i.e., credible intervals did not include zero for these 26 trends).

Florida, Georgia, Louisiana, Mississippi, North Ca↗

USGS2015

Spatial capture-recapture models allowing Markovian transience or dispersal

Spatial capture–recapture (SCR) models are a relatively recent development in quantitative ecology, and they are becoming widely used to model density in studies of animal populations using camera traps, DNA sampling and other methods which produce spatially explicit individual encounter information. One of the core assumptions of SCR models is that individuals possess home ranges that are spatially stationary during the sampling period. For many species, this assumption is unlikely to be met and, even for species that are typically territorial, individuals may disperse or exhibit transience at some life stages. In this paper we first conduct a simulation study to evaluate the robustness of estimators of density under ordinary SCR models when dispersal or transience is present in the population. Then, using both simulated and real data, we demonstrate that such models can easily be described in the BUGS language providing a practical framework for their analysis, which allows us to evaluate movement dynamics of species using capture–recapture data. We find that while estimators of density are extremely robust, even to pathological levels of movement (e.g., complete transience), the estimator of the spatial scale parameter of the encounter probability model is confounded with the dispersal/transience scale parameter. Thus, use of ordinary SCR models to make inferences about density is feasible, but interpretation of SCR model parameters in relation to movement should be avoided. Instead, when movement dynamics are of interest, such dynamics should be parameterized explicitly in the model.

Population Ecology↗

USGS2015

Density, distribution, and genetic structure of grizzly bears in the Cabinet-Yaak Ecosystem

The conservation status of the 2 threatened grizzly bear ( Ursus arctos ) populations in the Cabinet-Yaak Ecosystem (CYE) of northern Montana and Idaho had remained unchanged since designation in 1975; however, the current demographic status of these populations was uncertain. No rigorous data on population density and distribution or analysis of recent population genetic structure were available to measure the effectiveness of conservation efforts. We used genetic detection data from hair corral, bear rub, and opportunistic sampling in traditional and spatial capture–recapture models to generate estimates of abundance and density of grizzly bears in the CYE. We calculated mean bear residency on our sampling grid from telemetry data using Huggins and Pledger models to estimate the average number of bears present and to correct our superpopulation estimates for lack of geographic closure. Estimated grizzly bear abundance (all sex and age classes) in the CYE in 2012 was 48–50 bears, approximately half the population recovery goal. Grizzly bear density in the CYE (4.3–4.5 grizzly bears/1,000 km 2 ) was among the lowest of interior North American populations. The sizes of the Cabinet ( n = 22–24) and Yaak ( n = 18–22) populations were similar. Spatial models produced similar estimates of abundance and density with comparable precision without requiring radio-telemetry data to address assumptions of geographic closure. The 2 populations in the CYE were demographically and reproductively isolated from each other and the Cabinet population was highly inbred. With parentage analysis, we documented natural migrants to the Cabinet and Yaak populations by bears born to parents in the Selkirk and Northern Continental Divide populations. These events supported data from other sources suggesting that the expansion of neighboring populations may eventually help sustain the CYE populations. However, the small size, isolation, and inbreeding documented by this study demonstrate the need for comprehensive management designed to support CYE population growth and increased connectivity and gene flow with other populations.

Idaho, Montana↗

USGS2015

Population size and stopover duration estimation using mark–resight data and Bayesian analysis of a superpopulation model

We present a novel formulation of a mark–recapture–resight model that allows estimation of population size, stopover duration, and arrival and departure schedules at migration areas. Estimation is based on encounter histories of uniquely marked individuals and relative counts of marked and unmarked animals. We use a Bayesian analysis of a state–space formulation of the Jolly–Seber mark–recapture model, integrated with a binomial model for counts of unmarked animals, to derive estimates of population size and arrival and departure probabilities. We also provide a novel estimator for stopover duration that is derived from the latent state variable representing the interim between arrival and departure in the state–space model. We conduct a simulation study of field sampling protocols to understand the impact of superpopulation size, proportion marked, and number of animals sampled on bias and precision of estimates. Simulation results indicate that relative bias of estimates of the proportion of the population with marks was low for all sampling scenarios and never exceeded 2%. Our approach does not require enumeration of all unmarked animals detected or direct knowledge of the number of marked animals in the population at the time of the study. This provides flexibility and potential application in a variety of sampling situations (e.g., migratory birds, breeding seabirds, sea turtles, fish, pinnipeds, etc.). Application of the methods is demonstrated with data from a study of migratory sandpipers.

Biometrics↗

USGS2015

Estimating population size for Capercaillie ( Tetrao urogallus L.) with spatial capture-recapture models based on genotypes from one field sample

We conducted a survey of an endangered and cryptic forest grouse, the capercaillie Tetrao urogallus , based on droppings collected on two sampling occasions in eight forest fragments in central Switzerland in early spring 2009. We used genetic analyses to sex and individually identify birds. We estimated sex-dependent detection probabilities and population size using a modern spatial capture-recapture (SCR) model for the data from pooled surveys. A total of 127 capercaillie genotypes were identified (77 males, 46 females, and 4 of unknown sex). The SCR model yielded atotal population size estimate (posterior mean) of 137.3 capercaillies (posterior sd 4.2, 95% CRI 130–147). The observed sex ratio was skewed towards males (0.63). The posterior mean of the sex ratio under the SCR model was 0.58 (posterior sd 0.02, 95% CRI 0.54–0.61), suggesting a male-biased sex ratio in our study area. A subsampling simulation study indicated that a reduced sampling effort representing 75% of the actual detections would still yield practically acceptable estimates of total size and sex ratio in our population. Hence, field work and financial effort could be reduced without compromising accuracy when the SCR model is used to estimate key population parameters of cryptic species.

PLoS ONE↗