Show simple item record

dc.contributor.authorPacifici, Jamian Krishna
dc.date.accessioned2014-03-04T20:22:05Z
dc.date.available2014-03-04T20:22:05Z
dc.date.issued2011-08
dc.identifier.otherpacifici_jamian_k_201108_phd
dc.identifier.urihttp://purl.galileo.usg.edu/uga_etd/pacifici_jamian_k_201108_phd
dc.identifier.urihttp://hdl.handle.net/10724/27555
dc.description.abstractConservation and management of rare species is one of the most challenging tasks confronting natural resource managers. Species are classified as rare for several reasons: (1) very few individuals are known to exist, (2) the species is widely distributed resulting in low densities, (3) the species has a clumped distribution and/or (4) the species has very low detection rates (elusive behavior, difficult to catch/observe). They are often most negatively affected by environmental perturbation (more specifically human alterations) making conservation and management extremely challenging. The Ivory-billed Woodpecker (Campephilus principalis), if extant (Fitzpatrick et al. 2005; Hill et al. 2006; Jackson 2006), may be the most rare and elusive bird species in the United States and thus presents a great challenge for designing efficient and effective surveys. In this dissertation I present results from a large-scale effort to estimate occupancy rates for the Ivory-billed Woodpecker. In addition I used this case study to highlight several important problems and shortfalls common to many studies involving rare species. These shortfalls motivated the development of several new approaches that provide advances in rare species modeling. First, I developed a framework for allocating effort that provides a probabilistic approach to sampling, allowing for improved accuracy in estimating occupancy probability. This approach was found to have a much lower predictive error rate compared to traditional approaches such as single-season occupancy estimation especially when there was a large amount of spatial heterogeneity in habitat and detection probability was low. Second, I developed a hierarchical model that integrates adaptive cluster sampling and occupancy estimation, which allowed for additional effort to be placed at adjacent sites after a known detection. I found this model to outperform traditional occupancy modeling and provide excellent coverage under a variety of conditions. Future improvements in conservation and management of rare species will be accomplished through a variety of techniques and approaches. Ultimately, I believe the most operative approach will be the integration of unique and innovative methods of data collection coupled with models that identify and subsequently estimate the most important vital rates responsible for driving population dynamics.
dc.languageeng
dc.publisheruga
dc.rightspublic
dc.subjectAdaptive sampling
dc.subjectBayesian hierarchical modeling
dc.subjectDesign-based estimation
dc.subjectDetection probability
dc.subjectIvory-billed woodpecker
dc.subjectModel-based estimation
dc.subjectOccupancy estimation
dc.subjectRare species
dc.titleConservation and management of rare species
dc.title.alternativethe development of adaptive models to reduce uncertainty influencing decision making
dc.typeDissertation
dc.description.degreePhD
dc.description.departmentDaniel B. Warnell School of Forestry and Natural Resources
dc.description.majorForest Resources
dc.description.advisorRobert J. Cooper
dc.description.advisorMichael Conroy
dc.description.committeeRobert J. Cooper
dc.description.committeeMichael Conroy
dc.description.committeeJames Peterson
dc.description.committeeNicole Lazar
dc.description.committeeJohn Drake


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record