Disease Ecology and Modeling Active
The USGS National Wildlife Health Center (NWHC) provides quantitative support and technical assistance to state and federal wildlife managers and partners to better understand or predict the impact of disease on wildlife populations.
The USGS is a world-leader in developing statistical tools and predictive models for wildlife diseases including those that impact humans and domestic animals. Human, agricultural, and wildlife health are interlinked; thus risk assessment, prediction, and management of wildlife diseases are important for our nation’s health and economy. However, examining disease dynamics, monitoring and evaluating wildlife health are difficult and expensive endeavors. These challenges often make wildlife health datasets “messy”, and extracting the signal from the noise in these datasets is often challenging or impossible using traditional methods. The NWHC has advanced skills in statistics and mathematics, and develops and applies novel quantitative methods to overcome the complexities of wildlife health data, and ultimately better understand the ecology and spread of wildlife diseases.
The projects conducted by the NWHC, as described below, represent a substantial leap forward in statistical modeling of wildlife diseases. Continued work will focus on further development of these and other novel techniques and application of these statistical innovations to current and future disease conditions on the landscape, especially using additional diseases as case studies. Ultimately, we believe these approaches will be used across a wide array of diseases to understand the critical drivers of disease and its spread, predict future disease states, and determine appropriate interventions for the protection and improvement of human, domestic animal and ecosystem health.
Spatial Analytical Techniques for Predictive Modeling of Emerging Diseases
To be able to predict the spread and impact of biological threats – both invasive species and emerging infectious diseases - we need to first understand how these threats spread across the landscape. Understanding the processes that allow for their growth and spread is critical to permit development of effective management actions for protecting the health of humans, animals and the environment, and to forecast the risk and spread of these threats to new populations and regions. However, developing models to understand disease processes is challenging owing to the scale and large uncertainty inherent in wildlife disease systems as well as the lack of suitable analytical tools. This is complicated by incomplete or passive surveillance efforts, which have known bias and could lead to false conclusions. To overcome these obstacles, the NWHC, in collaboration with the Wisconsin Department of Natural Resources, Colorado State University Co-op unit, and Kansas State University, developed a statistical framework that uses advanced statistics to predict the growth and spread of disease within wildlife populations while correcting for sampling biases and properly incorporating uncertainty. The NWHC applied this statistical framework to model the spatial and temporal changes in Chronic Wasting Disease (CWD) using surveillance data collected by the Wisconsin Department of Natural Resources. These efforts clearly demonstrated its usefulness in describing historic disease patterns as well as forecasting future growth and spread of CWD. Further development of these exciting modeling tools will occur in 2018 and 2019. In particular, future work will use these new analytical techniques to evaluate the effectiveness of historic disease management actions, rigorously assess the origin of disease outbreaks, and link individual animal dynamics to system-wide outcomes. Ultimately, we believe these approaches will be used across a wide array of diseases to understand the critical drivers of disease and its spread, predict future disease states, and determine appropriate interventions for the protection and improvement of human, domestic animal and ecosystem health.
Power to the People: Developing Accessible Tools for Statistical Modeling
There is a wealth of new analytical techniques being developed for analyzing ecological data, including investigations of fish and wildlife diseases. Unfortunately, these developments are often highly technical and require computer programming acumen, and often remain inaccessible to the average scientist as well as fish and wildlife managers. This inaccessibility prevents both scientists and managers from learning when and how to apply appropriate statistical tools, which slows the progress of science and the effectiveness of management programs. To address this problem, the NWHC in collaboration with the University of Montana have begun development of web applications that make sophisticated statistical modeling techniques easily accessible to the average scientist. Beginning in 2016, an application was developed that focused on providing an efficient means of gathering and quantifying expert opinions. Expert elicitation is an important scientific technique for understanding problems which are not well studied or are novel (e.g., new emerging diseases). This application provided the ability to quickly gather and quantify a large number of experts’ opinions for further sophisticated statistical analyses.
In 2017, the NWHC developed a new application for designing and conducting weighted surveillance programs for CWD, which is a type of surveillance that focuses on detecting new disease foci in a cost-efficient and statistically rigorous manner. Weighted surveillance is gaining in popularity among wildlife agencies trying to manage the spread of CWD with limited resources, but tools for its application are not readily available. Currently, additional applications are being developed for the modeling of population demography and disease processes. The fruits of these research efforts will be used by the scientific and wildlife management community at large. This research aligns with the USGS Ecosystems Mission Area goal to develop scientific and statistically reliable methods and protocols to assess the status and trends of the Nation's biological resources.
Check out the web application at: https://popr.cfc.umt.edu/CWD/
Wisconsin CWD Deer Population Study
Chronic wasting disease is a fatal disease affecting deer, elk and moose. These species are highly valued by society for conservation and hunting. Evidence is mounting that CWD causes declines in affected big game populations, however, no one has conducted a long-term study to demonstrate these declines while accounting for other influences such as predators and habitat conditions. The NWHC is working in collaboration with the Wisconsin Department of Natural Resources and the University of Wisconsin to investigate the long-term population-level impacts of CWD on white-tailed deer using an ecosystem-level approach. Through intensive field research and advanced statistical modeling, this project will measure the impacts of this disease on free-ranging deer populations using integrated population models. The data collected will help state game management agencies determine what aspects of the disease process could be targeted for effective disease control efforts, help evaluate potential management actions, and predict future disease intensity and impacts on white-tailed deer populations.
Quantitative Applications in Disease Ecology
The NWHC is working on developing new statistical and mathematical techniques and packaging them within user-friendly tools. Some examples of new tools in development are web applications to analyze and interpret complex data, assess risk of future or ongoing disease outbreaks, estimate the effects of disease on individuals, populations, and ecosystems, and evaluate potential management solutions. The results from this project are broadly applicable to a variety of wildlife diseases, but current focus is on development of new statistical methods to predict the likelihood of virus isolation from samples collected for avian influenza surveillance and predicting the spread of CWD in the Midwest. This study represents a cooperative ecosystems studies unit partnership with the University of Wisconsin Department of Statistics.
Partnerships in Emerging Wildlife Disease Epidemiology and Modeling
The NWHC routinely provides technical assistance to state, federal, and international wildlife managers who want to better understand or predict the impact of disease on wildlife populations using the advanced skills we have in statistics and mathematics. For example, the NWHC developed a model for CWD in Colorado that used several different datasets and incorporated sociological data as well as population level data to get better estimates of herd size and health. By providing necessary quantitative support to wildlife health problems, this technical assistance has directly and positively impacted wildlife health management across the nation.
Below are publications related to ecology and modeling.
Concordance in diagnostic testing for respiratory pathogens of bighorn sheep
Effects of wind energy generation and white-nose syndrome on the viability of the Indiana bat
Estimating the short-term recovery potential of little brown bats in the eastern United States in the face of White-nose syndrome
Mortality patterns and detection bias from carcass data: An example from wolf recovery in Wisconsin
Identifying priority chronic wasting disease surveillance areas for mule deer in Montana
Integrated survival analysis using an event-time approach in a Bayesian framework
A stage-structured, spatially explicit migration model for Myotis bats: mortality location affects system dynamics
Estimating the spatial distribution of wintering little brown bat populations in the eastern United States
Spatial and temporal patterns in concentrations of perfluorinated compounds in bald eagle nestlings in the Upper Midwestern United States
Perfluorinated chemicals (PFCs) are of concern due to their widespread use, persistence in the environment, tendency to accumulate in animal tissues, and growing evidence of toxicity. Between 2006 and 2011 we collected blood plasma from 261 bald eagle nestlings in six study areas from the upper Midwestern United States. Samples were assessed for levels of 16 different PFCs. We used regression anal
Using auxiliary information to improve wildlife disease surveillance when infected animals are not detected: A Bayesian approach
Underestimating the effects of spatial heterogeneity due to individual movement and spatial scale: infectious disease as an example
The effect of swab sample choice on the detection of avian influenza in apparently healthy wild ducks
Below are news stories associated with this project.
- Overview
The USGS National Wildlife Health Center (NWHC) provides quantitative support and technical assistance to state and federal wildlife managers and partners to better understand or predict the impact of disease on wildlife populations.
The USGS is a world-leader in developing statistical tools and predictive models for wildlife diseases including those that impact humans and domestic animals. Human, agricultural, and wildlife health are interlinked; thus risk assessment, prediction, and management of wildlife diseases are important for our nation’s health and economy. However, examining disease dynamics, monitoring and evaluating wildlife health are difficult and expensive endeavors. These challenges often make wildlife health datasets “messy”, and extracting the signal from the noise in these datasets is often challenging or impossible using traditional methods. The NWHC has advanced skills in statistics and mathematics, and develops and applies novel quantitative methods to overcome the complexities of wildlife health data, and ultimately better understand the ecology and spread of wildlife diseases.
The projects conducted by the NWHC, as described below, represent a substantial leap forward in statistical modeling of wildlife diseases. Continued work will focus on further development of these and other novel techniques and application of these statistical innovations to current and future disease conditions on the landscape, especially using additional diseases as case studies. Ultimately, we believe these approaches will be used across a wide array of diseases to understand the critical drivers of disease and its spread, predict future disease states, and determine appropriate interventions for the protection and improvement of human, domestic animal and ecosystem health.
Spatial Analytical Techniques for Predictive Modeling of Emerging Diseases
To be able to predict the spread and impact of biological threats – both invasive species and emerging infectious diseases - we need to first understand how these threats spread across the landscape. Understanding the processes that allow for their growth and spread is critical to permit development of effective management actions for protecting the health of humans, animals and the environment, and to forecast the risk and spread of these threats to new populations and regions. However, developing models to understand disease processes is challenging owing to the scale and large uncertainty inherent in wildlife disease systems as well as the lack of suitable analytical tools. This is complicated by incomplete or passive surveillance efforts, which have known bias and could lead to false conclusions. To overcome these obstacles, the NWHC, in collaboration with the Wisconsin Department of Natural Resources, Colorado State University Co-op unit, and Kansas State University, developed a statistical framework that uses advanced statistics to predict the growth and spread of disease within wildlife populations while correcting for sampling biases and properly incorporating uncertainty. The NWHC applied this statistical framework to model the spatial and temporal changes in Chronic Wasting Disease (CWD) using surveillance data collected by the Wisconsin Department of Natural Resources. These efforts clearly demonstrated its usefulness in describing historic disease patterns as well as forecasting future growth and spread of CWD. Further development of these exciting modeling tools will occur in 2018 and 2019. In particular, future work will use these new analytical techniques to evaluate the effectiveness of historic disease management actions, rigorously assess the origin of disease outbreaks, and link individual animal dynamics to system-wide outcomes. Ultimately, we believe these approaches will be used across a wide array of diseases to understand the critical drivers of disease and its spread, predict future disease states, and determine appropriate interventions for the protection and improvement of human, domestic animal and ecosystem health.
Power to the People: Developing Accessible Tools for Statistical Modeling
There is a wealth of new analytical techniques being developed for analyzing ecological data, including investigations of fish and wildlife diseases. Unfortunately, these developments are often highly technical and require computer programming acumen, and often remain inaccessible to the average scientist as well as fish and wildlife managers. This inaccessibility prevents both scientists and managers from learning when and how to apply appropriate statistical tools, which slows the progress of science and the effectiveness of management programs. To address this problem, the NWHC in collaboration with the University of Montana have begun development of web applications that make sophisticated statistical modeling techniques easily accessible to the average scientist. Beginning in 2016, an application was developed that focused on providing an efficient means of gathering and quantifying expert opinions. Expert elicitation is an important scientific technique for understanding problems which are not well studied or are novel (e.g., new emerging diseases). This application provided the ability to quickly gather and quantify a large number of experts’ opinions for further sophisticated statistical analyses.
In 2017, the NWHC developed a new application for designing and conducting weighted surveillance programs for CWD, which is a type of surveillance that focuses on detecting new disease foci in a cost-efficient and statistically rigorous manner. Weighted surveillance is gaining in popularity among wildlife agencies trying to manage the spread of CWD with limited resources, but tools for its application are not readily available. Currently, additional applications are being developed for the modeling of population demography and disease processes. The fruits of these research efforts will be used by the scientific and wildlife management community at large. This research aligns with the USGS Ecosystems Mission Area goal to develop scientific and statistically reliable methods and protocols to assess the status and trends of the Nation's biological resources.
Check out the web application at: https://popr.cfc.umt.edu/CWD/
Wisconsin CWD Deer Population Study
Chronic wasting disease is a fatal disease affecting deer, elk and moose. These species are highly valued by society for conservation and hunting. Evidence is mounting that CWD causes declines in affected big game populations, however, no one has conducted a long-term study to demonstrate these declines while accounting for other influences such as predators and habitat conditions. The NWHC is working in collaboration with the Wisconsin Department of Natural Resources and the University of Wisconsin to investigate the long-term population-level impacts of CWD on white-tailed deer using an ecosystem-level approach. Through intensive field research and advanced statistical modeling, this project will measure the impacts of this disease on free-ranging deer populations using integrated population models. The data collected will help state game management agencies determine what aspects of the disease process could be targeted for effective disease control efforts, help evaluate potential management actions, and predict future disease intensity and impacts on white-tailed deer populations.
Quantitative Applications in Disease Ecology
The NWHC is working on developing new statistical and mathematical techniques and packaging them within user-friendly tools. Some examples of new tools in development are web applications to analyze and interpret complex data, assess risk of future or ongoing disease outbreaks, estimate the effects of disease on individuals, populations, and ecosystems, and evaluate potential management solutions. The results from this project are broadly applicable to a variety of wildlife diseases, but current focus is on development of new statistical methods to predict the likelihood of virus isolation from samples collected for avian influenza surveillance and predicting the spread of CWD in the Midwest. This study represents a cooperative ecosystems studies unit partnership with the University of Wisconsin Department of Statistics.
Partnerships in Emerging Wildlife Disease Epidemiology and Modeling
The NWHC routinely provides technical assistance to state, federal, and international wildlife managers who want to better understand or predict the impact of disease on wildlife populations using the advanced skills we have in statistics and mathematics. For example, the NWHC developed a model for CWD in Colorado that used several different datasets and incorporated sociological data as well as population level data to get better estimates of herd size and health. By providing necessary quantitative support to wildlife health problems, this technical assistance has directly and positively impacted wildlife health management across the nation.
- Publications
Below are publications related to ecology and modeling.
Filter Total Items: 29Concordance in diagnostic testing for respiratory pathogens of bighorn sheep
Reliable diagnostic tests are essential for disease investigation and management. This is particularly true for diseases of free-ranging wildlife where sampling is logistically difficult precluding retesting. Clinical assays for wildlife diseases frequently vary among laboratories because of lack of appropriate standardized commercial kits. Results of diagnostic testing may also be called into queAuthorsDaniel P. Walsh, E. Frances Cassirer, Michael D. Bonds, Daniel R. Brown, William H. Edwards, Glen C. Weiser, Mark L. Drew, Robert E. Briggs, Karen A. Fox, Michael W. Miller, Sudarvili Shanthalingam, Subramaniam Srikumaran, Thomas E. BesserEffects of wind energy generation and white-nose syndrome on the viability of the Indiana bat
Wind energy generation holds the potential to adversely affect wildlife populations. Species-wide effects are difficult to study and few, if any, studies examine effects of wind energy generation on any species across its entire range. One species that may be affected by wind energy generation is the endangered Indiana bat (Myotis sodalis), which is found in the eastern and midwestern United StateAuthorsRichard A. Erickson, Wayne E. Thogmartin, James E. Diffendorfer, Robin E. Russell, Jennifer A. SzymanskiEstimating the short-term recovery potential of little brown bats in the eastern United States in the face of White-nose syndrome
White-nose syndrome (WNS) was first detected in North American bats in New York in 2006. Since that time WNS has spread throughout the northeastern United States, southeastern Canada, and southwest across Pennsylvania and as far west as Missouri. Suspect WNS cases have been identified in Minnesota and Iowa, and the causative agent of WNS (Pseudogymnoascus destructans) has recently been detected inAuthorsRobin E. Russell, Wayne E. Thogmartin, Richard A. Erickson, Jennifer A. Szymanski, Karl TinsleyMortality patterns and detection bias from carcass data: An example from wolf recovery in Wisconsin
We developed models and provide computer code to make carcass recovery data more useful to wildlife managers. With these tools, wildlife managers can understand the spatial, temporal (e.g., across time periods, seasons), and demographic patterns in mortality causes from carcass recovery datasets. From datasets of radio-collared and non-collared carcasses, managers can calculate the detection biasAuthorsJennifer L. Stenglein, Timothy R. Van Deelen, Adrian P. Wydeven, David J. Mladenoff, Jane E. Wiedenhoft, Nancy K. Businga, Julia A. Langenberg, Nancy J. Thomas, Dennis M. HeiseyIdentifying priority chronic wasting disease surveillance areas for mule deer in Montana
Chronic wasting disease (CWD) is a fatal prion disease that affects a variety of ungulate species including mule deer (Odocoileus hemionus). As of 2014, no CWD cases had been reported in free-ranging ungulates in Montana. However, nearby cases in Canada, Wyoming, and the Dakotas indicated that the disease was encroaching on Montana's borders. Mule deer are native and common throughout Montana, andAuthorsRobin E. Russell, Justin Gude, N.J. Anderson, Jennifer M. RamseyIntegrated survival analysis using an event-time approach in a Bayesian framework
Event-time or continuous-time statistical approaches have been applied throughout the biostatistical literature and have led to numerous scientific advances. However, these techniques have traditionally relied on knowing failure times. This has limited application of these analyses, particularly, within the ecological field where fates of marked animals may be unknown. To address these limitationsAuthorsDaniel P. Walsh, VJ Dreitz, Dennis M. HeiseyA stage-structured, spatially explicit migration model for Myotis bats: mortality location affects system dynamics
Bats are ecologically and economically important species because they consume insects, transport nutrients, and pollinate flowers. Many species of bats, including those in the Myotis genus, are facing population decline and increased extinction risk. Despite these conservation concerns, few models exist for providing insight into the population dynamics of bats in a spatially explicit context.AuthorsRichard A. Erickson, Wayne E. Thogmartin, Robin E. Russell, James E. Diffendorfer, Jennifer A. SzymanskiEstimating the spatial distribution of wintering little brown bat populations in the eastern United States
Depicting the spatial distribution of wildlife species is an important first step in developing management and conservation programs for particular species. Accurate representation of a species distribution is important for predicting the effects of climate change, land-use change, management activities, disease, and other landscape-level processes on wildlife populations. We developed models to eAuthorsRobin E. Russell, Karl Tinsley, Richard A. Erickson, Wayne E. Thogmartin, Jennifer A. SzymanskiSpatial and temporal patterns in concentrations of perfluorinated compounds in bald eagle nestlings in the Upper Midwestern United States
Perfluorinated chemicals (PFCs) are of concern due to their widespread use, persistence in the environment, tendency to accumulate in animal tissues, and growing evidence of toxicity. Between 2006 and 2011 we collected blood plasma from 261 bald eagle nestlings in six study areas from the upper Midwestern United States. Samples were assessed for levels of 16 different PFCs. We used regression anal
AuthorsWilliam T. Route, Robin E. Russell, Andrew B. Lindstrom, Mark J. Strynor, Rebecca L. KeyUsing auxiliary information to improve wildlife disease surveillance when infected animals are not detected: A Bayesian approach
There are numerous situations in which it is important to determine whether a particular disease of interest is present in a free-ranging wildlife population. However adequate disease surveillance can be labor-intensive and expensive and thus there is substantial motivation to conduct it as efficiently as possible. Surveillance is often based on the assumption of a simple random sample, but this cAuthorsDennis M. Heisey, Christopher S. Jennelle, Robin E. Russell, Daniel P. WalshUnderestimating the effects of spatial heterogeneity due to individual movement and spatial scale: infectious disease as an example
Many ecological and epidemiological studies occur in systems with mobile individuals and heterogeneous landscapes. Using a simulation model, we show that the accuracy of inferring an underlying biological process from observational data depends on movement and spatial scale of the analysis. As an example, we focused on estimating the relationship between host density and pathogen transmission. ObsAuthorsPaul C. Cross, Damien Caillaud, Dennis M. HeiseyThe effect of swab sample choice on the detection of avian influenza in apparently healthy wild ducks
Historically, avian influenza viruses have been isolated from cloacal swab specimens, but recent data suggest that the highly pathogenic avian influenza (HPAI) H5N1 virus can be better detected from respiratory tract specimens. To better understand how swab sample type affects the detection ability of low pathogenic avian influenza (LPAI) viruses we collected and tested four swab types: oropharyngAuthorsHon S. Ip, Robert J. Dusek, Dennis M. Heisey - News
Below are news stories associated with this project.