Workshop on Use of Ecological Inference Methods for Estimation Problems Involving Political Conflict

USGS scientists Jim Hines and Jim Nichols presented a workshop on occupancy modeling to political scientists at Princeton University, April 19-20. The motivation stemmed from an earlier workshop at Columbia University in late January addressing the issue of nondetection in fields ranging from ecology to epidemiology to natural disasters to political science. During the course of this January workshop, Nichols noted that many of these diverse estimation problems could be addressed using models developed by ecologists to deal specifically with nondetection. The Princeton workshop represented an effort to apply these models to a specific type of inference problem, estimation of the number or proportion of locations (could be towns within a country, countries, political regions, etc.) at which political conflict occurred over a specified time interval. Data for these efforts are compiled from web-based reports of various news agencies, but there is widespread recognition that all conflicts are not detected. Political scientists have historically noted the problem of nondetection (sometimes) but have almost universally ignored it during analysis. Hines and Nichols lectured about the underlying logic of replication-based approaches to inference about underlying processes in the face of nondetection and demonstrated software developed to implement these procedures. Both the methods and the associated software were developed by USGS scientists and collaborators. An example application was developed with political scientists from Pennsylvania State University, and results were both interesting and useful. The initial manuscript of a planned series of papers is now being developed to explain use of these ecological methods to address this cross-discipline problem of nondetection.

Linking River Management to Species Conservation Using Dynamic Landscape Models

"Interdisciplinary USGS research was highlighted this week in an “early view” online publication of the international journal River Research and Applications. The paper titled “Linking River Management to Species Conservation Using Dynamic Landscape Models” describes development of a modeling approach to simulate effects of streamflow alteration on fish occupancy dynamics within stream networks. New in this work is the integration of landscape-scale variation in hydrology, geographic context and metapopulation processes to build a spatially-explicit modeling approach for comparing biological outcomes of alternative water management scenarios. The paper’s fourteen authors represent biology, geography and water disciplines from nine science centers or offices, and one university. The paper may be downloaded from http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1535-1467/earlyview."

here's the paper - hot off the press.

Estimating the Binomial N

Dr. William A. Link, mathematical statistician at Patuxent Wildlife Research Center, will present a Washington Statistical Society seminar on “Estimating the Binomial N,” at 1:30pm on Wednesday, April 18, 2012 at the Bureau of Labor Statistics. The seminar deals with estimating the size of a population based on incomplete counts. The binomial model used has two parameters, N (population size) and p (expected proportion counted). In this seminar Dr. Link addresses the challenging problem of estimating N based solely on replicate counts X, without auxiliary data about p. The problem has a long and not terribly encouraging history among statisticians. This seminar reviews some of the difficulties, examines recent developments based on model-based replication, and suggests diagnostic criteria allowing confidence in their application.

Patuxent Wildlife Research Center convenes structured decision making workshop to help NPS develop decision for management of the endangered Shenandoah salamander under the threat of climate change

Event Date(s): 4/10/2012 - 4/12/2012

In many national parks, high-elevation biota are severely threatened by climate change. An assessment of the impacts of climate change is required for efficient spending of funds, proper management of rare and endangered species, and effective conservation of National Park Service biological resources. This project proposes to combine refined habitat and competition models for the Shenandoah salamander (Plethodon shenandoah) with downscaled projections of future climates to anticipate changes in the salamander’s distribution under a variety of climate scenarios. The decision facing National Park Service managers is to choose among management actions which may mitigate the potential negative effects of climate change. The team will be coordinated by Evan Grant, Principle Investigator of the NE Amphibian Research and Monitoring Initiative, and will include decision makers at multiple levels of management, including park, regional and WASO level NPS staff as well as USGS decision modelers. The workshop is being held at the FWS NCTC facility.

Invasive Species and Climate Change: Does Immunology have a Role?

Dr. Caldwell Hahn of Patuxent Wildlife Research Center has been invited by the Interdepartmental Graduate Program at Penn State University to give a presentation at the Spring 2012 Seminar Series. This annual seminar series is open to the public and geared to stimulating discussion about how scientists engage with the public. This year the theme of the series is “Saving a Changing world: Ecology in the Public Eye”. Dr. Hahn will be speaking on April 23 on the topic of” Invasive Species and Climate Change: Does Immunology have a Role?”

Radar Analysis of Fall Bird Migration Stopover Sites
in the Northeastern U.S.

USGS Patuxent Wildlife Research Center biologist Deanna Dawson and Dr. Jeff Buler, University of Delaware, will present a seminar on their project ‘Radar Analysis of Fall Bird Migration Stopover Sites in the Northeastern U.S.’, on April 26 at the U.S. Fish & Wildlife Service Region 5 office. The project used data collected by 16 of the weather surveillance radars in the National Weather Service’s national network to identify ‘important’ stopover sites for landbirds that migrate nocturnally, and developed models to predict potentially important stopover sites in portions of Region 5 not sampled by the radars. The analyses indicate that preserving patches of natural habitat, particularly deciduous forests, in developed or agricultural landscapes and along major coastlines should be a priority for conservation plans addressing the habitat requirements of migratory landbirds in the northeastern U.S.

Archives