Geology topics

Trevor J. Hefley

Publications and source records attributed to Trevor J. Hefley.

18 recordsLinked to original sources

USGS2026

Ecological diffusion models are still useful in ecology

Context Connecting animal movement across scales—both among spatial scales and between individual and population level processes—is a central theme in ecology, including the sub-fields of movement and landscape ecology. However, most modeling frameworks operate at either the individual (Lagrangian) or population (Eulerian) level. Objectives We aimed to review how a partial differential equation (PDE) known as the ecological diffusion equation (EDE) offers connections among the scales of animal movement and how it can be embedded in commonly used statistical frameworks (e.g., hierarchical models) and fit to multiple types of data, such as count, presence-only, presence-absence, and telemetry data. In doing so, we also highlight emerging and promising avenues of research related to statistical EDE models. Methods We present how the EDE is derived from a simple set of first principles, can be used to represent both individual animal movement and spatiotemporal population dynamics, and can be fit to data by being applied in a hierarchical statistical framework. Finally, we review literature on recent applications and advances related to EDE models. Results In presenting the derivation of the EDE and reviewing recent advances and applications of EDE models, we established that the EDE can enable a mechanistic understanding of animal movement across scales, as well as the integration of multiple types of data. We also identified a number of topics where EDE models can be advanced to gain additional ecological insight. Conclusions Ecological diffusion is a process that emerges naturally from first principles governing animal movement and can be used to model both individual and population-level spatiotemporal dynamics. Statistical EDE models are related to other modeling frameworks, such as species distribution models, occupancy models, and step-selection analysis. The flexibility of hierarchical statistical modeling allows combining different types of data collected on movement and population dynamics, and the theoretical underpinnings of the EDE allow it to connect ecological processes occurring at fine and coarse spatial scales, as well as individual animal movement and population dynamics.

Landscape Ecology↗

USGS2025

Simple bagged movement models for telemetry data

Determining which statistical methods are appropriate for data is both user and data dependent and prone to change as new methodology becomes available. This process encompasses model ideation, model selection, and determining appropriate use of statistical methods. Literature on models for animal movement emerging in the past two decades has yielded a rich collection of statistical methods garnering much deserved positive attention. Among such efforts, there is limited investigation of the broader place for simple machine learning methodology in animal movement modeling. We propose a bagged (i.e., bootstrap aggregated) animal movement model using simple, off-the-shelf machine learning algorithms. The model is intuitive, retains statistical inference about characteristics of animal movement (i.e., estimated from model-based summary statistics), and only requires knowledge of elementary statistical and machine learning analysis to understand. We show by simulation that our model can provide unbiased estimates of pertinent characteristics of animal movement (e.g., daily displacement) in the presence of large and realistic location error. We believe that increasing accessible literature on simple machine learning animal movement models provides valuable pedagogical and practical support for researchers using statistical models to study animal movement.

Ecology and Evolution↗

USGS2024

Estimation of contact time among animals from telemetry data

Continuous processes in most applications are measured discretely with error. This complicates the task of detecting intersections and the number of intersections between two continuous processes (i.e., when the processes have the same value). Intersections of continuous processes are scientifically important but challenging to estimate from data. For example, in the field of animal ecology, intersections of the paths of moving animals tracked with satellite technologies can be used to understand disease transmission. We illustrate how to quantify contact between animals using telemetry data (i.e., the recorded locations of an animal over time). We introduce our method to quantify contact time with accessible concepts from introductory stochastic process literature, such as Brownian motion. Then, we provide two data examples using white-tailed deer ( Odocoileus virginianus ) and mule deer ( Odocoileus hemionus) telemetry data in a region with high prevalence of chronic wasting disease. Our work provides a needed connection between existing model-based literature for animal movement and rule-based literature for animal interaction. Further, our work illustrates a unique statistical problem receiving minimal attention with broad applicability in human and livestock tracking.

The American Statistician↗

USGS2024

Treed Gaussian processes for animal movement modeling

Wildlife telemetry data may be used to answer a diverse range of questions relevant to wildlife ecology and management. One challenge to modeling telemetry data is that animal movement often varies greatly in pattern over time, and current continuous-time modeling approaches to handle such nonstationarity require bespoke and often complex models that may pose barriers to practitioner implementation. We demonstrate a novel application of treed Gaussian process (TGP) modeling, a Bayesian machine learning approach that automatically captures the nonstationarity and abrupt transitions present in animal movement. The machine learning formulation of TGPs enables modeling to be nearly automated, while their Bayesian formulation allows for the derivation of movement descriptors with associated uncertainty measures. We demonstrate the use of an existing R package to implement TGPs using the familiar Markov chain Monte Carlo algorithm. We then use estimated movement trajectories to derive movement descriptors that can be compared across individuals and populations. We applied the TGP model to a case study of lesser prairie-chickens ( Tympanuchus pallidicinctus ) to demonstrate the benefits of TGP modeling and compared distance traveled and residence times across lesser prairie-chicken individuals and populations. For broad usability, we outline all steps necessary for practitioners to specify relevant movement descriptors (e.g., turn angles, speed, contact points) and apply TGP modeling and trajectory comparison to their own telemetry datasets. Combining the predictive power of machine learning and the statistical inference of Bayesian methods to model movement trajectories allows for the estimation of statistically comparable movement descriptors from telemetry studies. Our use of an accessible R package allows practitioners to model trajectories and estimate movement descriptors, facilitating the use of telemetry data to answer applied management questions.

Ecology and Evolution↗

USGS2023

Model-based surveillance system design under practical constraints with application to white-nose syndrome

Infectious diseases are powerful ecological forces structuring ecosystems, causing devastating economic impacts and disrupting society. Successful prevention and control of pathogens requires knowledge of the current scope and severity of disease, as well as the ability to forecast future disease dynamics. Assessment of the current situation as well as prediction of the future conditions, rely on spatially referenced information regarding the presence or absence of a pathogen, and the prevalence of the pathogen in the population. In particular, knowledge about the location of the disease front is foundational for deploying disease countermeasures to prevent further disease spread and focusing control efforts to reduce disease intensity in affected areas. In this paper, we develop a model-based approach to designing sampling strategies for wildlife disease surveillance at the disease front. Specifically, we use a mechanistic spatio-temporal model based on an underlying partial differential equation to track the disease dynamics and predict the disease prevalence in the future. We also devise an optimal surveillance system design at the disease front that takes into account the practical constraints of sampling. We evaluate the effectiveness of our proposed design via a simulation study and demonstrate the application of the proposed approach by designing a surveillance strategy for the pathogen that causes white-nose syndrome.

Environmental and Ecological Statistics↗

USGS2023

Bayesian forecasting of disease spread with little or no local data

Rapid and targeted management actions are a prerequisite to efficiently mitigate disease outbreaks. Targeted actions, however, require accurate spatial information on disease occurrence and spread. Frequently, targeted management actions are guided by non-statistical approaches that define the affected area by a pre-determined distance surrounding a small number of disease detections. As an alternative, we present a long-recognized but underutilized Bayesian technique that uses limited local data and informative priors to make statistically valid predictions and forecasts about disease occurrence and spread. As a case study, we use limited local data that were available after the detection of chronic wasting disease in Michigan, U.S. along with information rich priors obtained from a previous study in a neighboring state. Using these limited local data and informative priors, we generate statistically valid predictions of disease occurrence and spread for the Michigan study area. This Bayesian technique is conceptually and computationally simple, relies on little to no local data, and is competitive with non-statistical distance-based metrics in all performance evaluations. Bayesian modeling has added benefits because it allows practitioners to generate immediate forecasts of future disease conditions and provides a principled framework to incorporate new data as they accumulate. We contend that the Bayesian technique offers broad-scale benefits and opportunities to make statistical inference across a diversity of data-deficient systems, not limited to disease.

Scientific Reports↗

USGS2021

Long-term Pseudogymnoascus destructans surveillance data reveal factors contributing to pathogen presence

The disease white-nose syndrome (WNS) was first recognized in upstate New York in 2006 and has since spread across much of the United States (U.S.), causing severe mortality in several North American bat species. To aid in the identification and monitoring of at-risk bat populations, we evaluate factors associated with the presence of the causative fungal agent of WNS, Pseudogymnoascus destructans ( Pd ), in the continental United States. We obtained Pd samples through hibernaculum surveys conducted from 2013 to 2020, with all samples analyzed at the U.S. Geological Survey National Wildlife Health Center. Using generalized additive models, we estimated the likelihood of Pd presence under three different hypotheses: human-mediated, species-mediated, and hibernaculum type. In addition to hypothesis-related predictor variables, a subset of models included a smoothed nonseparable effect of longitude and latitude and a smoothed effect of time since study onset to account for spatial and temporal autocorrelation. Under all hypotheses, models indicated probability of Pd detection is best described by the smoothed nonseparable effect of longitude and latitude and a smoothed effect of time since onset of this study. After accounting for spatial and temporal autocorrelations, only hibernaculum type significantly affected Pd presence, with mines and culverts/tunnels less likely to contain Pd compared with caves. Reduced likelihood of Pd presence in mines and culverts/tunnels bodes well for bats of the western and southern United States, where use of these hibernaculum types is more common. While our findings can help guide monitoring and management efforts, the potential for long-distance dispersal combined with variation in community composition and hibernation ecology between the western and eastern United States necessitates the continued monitoring of Pd presence.

Ecosphere↗

USGS2021

Recovering individual-level spatial inference from aggregated binary data

Binary regression models are commonly used in disciplines such as epidemiology and ecology to determine how spatial covariates influence individuals. In many studies, binary data are shared in a spatially aggregated form to protect privacy. For example, rather than reporting the location and result for each individual that was tested for a disease, researchers may report that a disease was detected or not detected within geopolitical units. Often, the spatial aggregation process obscures the values of response variables, spatial covariates, and locations of each individual, which makes recovering individual-level inference difficult. We show that applying a series of transformations, including a change of support, to a bivariate point process model allows researchers to recover individual-level inference for spatial covariates from spatially aggregated binary data. The series of transformations preserves the convenient interpretation of desirable binary regression models that are commonly applied to individual-level data. Using a simulation experiment, we compare the performance of our proposed method under varying types of spatial aggregation against the performance of standard approaches using the original individual-level data. We illustrate our method by modeling individual-level probability of infection using a data set that has been aggregated to protect an at-risk and endangered species of bats. Our simulation experiment and data illustration demonstrate the utility of the proposed method when access to original non-aggregated data is impractical or prohibited.

Northeast and Midwest United States↗

USGS2019

Bringing Bayesian models to life

Bringing Bayesian Models to Life empowers the reader to extend, enhance, and implement statistical models for ecological and environmental data analysis. We open the black box and show the reader how to connect modern statistical models to computer algorithms. These algorithms allow the user to fit models that answer their scientific questions without needing to rely on automated Bayesian software. We show how to handcraft statistical models that are useful in ecological and environmental science including: linear and generalized linear models, spatial and time series models, occupancy and capture-recapture models, animal movement models, spatio-temporal models, and integrated population-models.

Book↗

USGS2018

Animal movement models for migratory individuals and groups

Animals often exhibit changes in their behaviour during migration. Telemetry data provide a way to observe geographic position of animals over time, but not necessarily changes in the dynamics of the movement process. Continuous‐time models allow for statistical predictions of the trajectory in the presence of measurement error and during periods when the telemetry device did not record the animal's position. However, continuous‐time models capable of mimicking realistic trajectories with sufficient detail are computationally challenging to fit to large datasets. Furthermore, basic continuous‐time model specifications (e.g. Brownian motion) lack realism in their ability to capture nonstationary dynamics. We present a unified class of animal movement models that are computationally efficient and provide a suite of approaches for accommodating nonstationarity in continuous trajectories due to migration and interactions among individuals. Our approach uses process convolutions to allow for flexibility in the movement process while facilitating implementation and incorporating location uncertainty. We show how to nest convolution models to incorporate interactions among migrating individuals to account for nonstationarity and provide inference about dynamic migratory networks. We demonstrate these approaches in two case studies involving migratory birds. Specifically, we used process convolution models with temporal deformation to account for heterogeneity in individual greater white‐fronted goose migrations in Europe and Iceland, and we used nested process convolutions to model dynamic migratory networks in sandhill cranes in North America. The approach we present accounts for various forms of temporal heterogeneity in animal movement and is not limited to migratory applications. Furthermore, our models rely on well‐established principles for modelling‐dependent data and leverage modern approaches for modelling dynamic networks to help explain animal movement and social interaction.

Methods in Ecology and Evolution↗

USGS2017

Bias correction of bounded location errors in presence-only data

Location error occurs when the true location is different than the reported location. Because habitat characteristics at the true location may be different than those at the reported location, ignoring location error may lead to unreliable inference concerning species–habitat relationships. We explain how a transformation known in the spatial statistics literature as a change of support (COS) can be used to correct for location errors when the true locations are points with unknown coordinates contained within arbitrary shaped polygons. We illustrate the flexibility of the COS by modelling the resource selection of Whooping Cranes ( Grus americana ) using citizen contributed records with locations that were reported with error. We also illustrate the COS with a simulation experiment. In our analysis of Whooping Crane resource selection, we found that location error can result in up to a five-fold change in coefficient estimates. Our simulation study shows that location error can result in coefficient estimates that have the wrong sign, but a COS can efficiently correct for the bias.

Methods in Ecology and Evolution↗

USGS2017

When mechanism matters: Bayesian forecasting using models of ecological diffusion

Ecological diffusion is a theory that can be used to understand and forecast spatio-temporal processes such as dispersal, invasion, and the spread of disease. Hierarchical Bayesian modelling provides a framework to make statistical inference and probabilistic forecasts, using mechanistic ecological models. To illustrate, we show how hierarchical Bayesian models of ecological diffusion can be implemented for large data sets that are distributed densely across space and time. The hierarchical Bayesian approach is used to understand and forecast the growth and geographic spread in the prevalence of chronic wasting disease in white-tailed deer ( Odocoileus virginianus ). We compare statistical inference and forecasts from our hierarchical Bayesian model to phenomenological regression-based methods that are commonly used to analyse spatial occurrence data. The mechanistic statistical model based on ecological diffusion led to important ecological insights, obviated a commonly ignored type of collinearity, and was the most accurate method for forecasting.

Ecology Letters↗

USGS2017

The Bayesian group lasso for confounded spatial data

Generalized linear mixed models for spatial processes are widely used in applied statistics. In many applications of the spatial generalized linear mixed model (SGLMM), the goal is to obtain inference about regression coefficients while achieving optimal predictive ability. When implementing the SGLMM, multicollinearity among covariates and the spatial random effects can make computation challenging and influence inference. We present a Bayesian group lasso prior with a single tuning parameter that can be chosen to optimize predictive ability of the SGLMM and jointly regularize the regression coefficients and spatial random effect. We implement the group lasso SGLMM using efficient Markov chain Monte Carlo (MCMC) algorithms and demonstrate how multicollinearity among covariates and the spatial random effect can be monitored as a derived quantity. To test our method, we compared several parameterizations of the SGLMM using simulated data and two examples from plant ecology and disease ecology. In all examples, problematic levels multicollinearity occurred and influenced sampling efficiency and inference. We found that the group lasso prior resulted in roughly twice the effective sample size for MCMC samples of regression coefficients and can have higher and less variable predictive accuracy based on out-of-sample data when compared to the standard SGLMM.

Journal of Agricultural, Biological, and Environme↗

USGS2017

An integrated data model to estimate spatiotemporal occupancy, abundance, and colonization dynamics

Ecological invasions and colonizations occur dynamically through space and time. Estimating the distribution and abundance of colonizing species is critical for efficient management or conservation. We describe a statistical framework for simultaneously estimating spatiotemporal occupancy and abundance dynamics of a colonizing species. Our method accounts for several issues that are common when modeling spatiotemporal ecological data including multiple levels of detection probability, multiple data sources, and computational limitations that occur when making fine-scale inference over a large spatiotemporal domain. We apply the model to estimate the colonization dynamics of sea otters ( Enhydra lutris ) in Glacier Bay, in southeastern Alaska.

Ecology↗

USGS2017

A dynamic spatio-temporal model for spatial data

Analyzing spatial data often requires modeling dependencies created by a dynamic spatio-temporal data generating process. In many applications, a generalized linear mixed model (GLMM) is used with a random effect to account for spatial dependence and to provide optimal spatial predictions. Location-specific covariates are often included as fixed effects in a GLMM and may be collinear with the spatial random effect, which can negatively affect inference. We propose a dynamic approach to account for spatial dependence that incorporates scientific knowledge of the spatio-temporal data generating process. Our approach relies on a dynamic spatio-temporal model that explicitly incorporates location-specific covariates. We illustrate our approach with a spatially varying ecological diffusion model implemented using a computationally efficient homogenization technique. We apply our model to understand individual-level and location-specific risk factors associated with chronic wasting disease in white-tailed deer from Wisconsin, USA and estimate the location the disease was first introduced. We compare our approach to several existing methods that are commonly used in spatial statistics. Our spatio-temporal approach resulted in a higher predictive accuracy when compared to methods based on optimal spatial prediction, obviated confounding among the spatially indexed covariates and the spatial random effect, and provided additional information that will be important for containing disease outbreaks.

Wisconsin↗

USGS2016

When can the cause of a population decline be determined?

Inferring the factors responsible for declines in abundance is a prerequisite to preventing the extinction of wild populations. Many of the policies and programmes intended to prevent extinctions operate on the assumption that the factors driving the decline of a population can be determined. Exogenous factors that cause declines in abundance can be statistically confounded with endogenous factors such as density dependence. To demonstrate the potential for confounding, we used an experiment where replicated populations were driven to extinction by gradually manipulating habitat quality. In many of the replicated populations, habitat quality and density dependence were confounded, which obscured causal inference. Our results show that confounding is likely to occur when the exogenous factors that are driving the decline change gradually over time. Our study has direct implications for wild populations, because many factors that could drive a population to extinction change gradually through time.

Ecology Letters↗

USGS2015

Whooping crane stopover site use intensity within the Great Plains

Whooping cranes ( Grus americana ) of the Aransas-Wood Buffalo population migrate twice each year through the Great Plains in North America. Recovery activities for this endangered species include providing adequate places to stop and rest during migration, which are generally referred to as stopover sites. To assist in recovery efforts, initial estimates of stopover site use intensity are presented, which provide opportunity to identify areas across the migration range used more intensively by whooping cranes. We used location data acquired from 58 unique individuals fitted with platform transmitting terminals that collected global position system locations. Radio-tagged birds provided 2,158 stopover sites over 10 migrations and 5 years (2010–14). Using a grid-based approach, we identified 1,095 20-square-kilometer grid cells that contained stopover sites. We categorized occupied grid cells based on density of stopover sites and the amount of time cranes spent in the area. This assessment resulted in four categories of stopover site use: unoccupied, low intensity, core intensity, and extended-use core intensity. Although provisional, this evaluation of stopover site use intensity offers the U.S. Fish and Wildlife Service and partners a tool to identify landscapes that may be of greater conservation significance to migrating whooping cranes. Initially, the tool will be used by the U.S. Fish and Wildlife Service and other interested parties in evaluating the Great Plains Wind Energy Habitat Conservation Plan.

Alberta, Iowa, Kansas, Manitoba, Minnesota, Missou↗

USGS2015

On the existence of maximum likelihood estimates for presence-only data

Presence-only data can be used to determine resource selection and estimate a species’ distribution. Maximum likelihood is a common parameter estimation method used for species distribution models. Maximum likelihood estimates, however, do not always exist for a commonly used species distribution model – the Poisson point process. We demonstrate the issue with conventional maximum likelihood mathematically, using a data example, and a simulation experiment and show alternative estimation methods. We found that when habitat preferences are strong or the number of presence-only locations is small, by chance, maximum likelihood coefficient estimates for the Poisson point process model may not exist. We found that several alternative estimation methods can produce reliable estimates, but results will depend on the chosen method. It is important to identify conditions for which maximum likelihood estimates are unlikely to be identifiable from presence-only data. In data sets where the maximum likelihood estimates do not exist, penalized likelihood and Bayesian methods will produce coefficient estimates, but these are sensitive to the choice of estimation procedure and prior or penalty term. When sample size is small or it is thought that habitat preferences are strong, we propose a suite of estimation procedures researchers can consider using.

Methods in Ecology and Evolution↗