Marginalizing Bayesian population models - data for examples in the Grand Canyon region, southeastern Arizona, western Oregon USA - 1990-2015
Data for journal manuscript: A need for speed in Bayesian population models: a practical guide to marginalizing and recovering discrete latent states
Dates
Publication Date
2020-03-05
Start Date
1990
End Date
2015
Citation
Yackulic, C.B., Dzul, M.C, Reid, J.A., Sanderlin, J.S., Block, W.M., Ganey, J.L., Dodrill, M.J., and Yard, M.D., 2020, Marginalizing bayesian population models - data for examples in the Grand Canyon region, southeastern Arizona, and western Oregon, USA - 1990-2015: U.S. Geological Survey data release, https://doi.org/10.5066/P9JN5C0L.
Summary
These data were compiled here to fit various versions of Bayesian population models and compare their performance, primarily the time required to make inferences using different softwares and versions of code. The humpback chub data were collected by US Geological Survey and US Fish and Wildlife service in the Colorado and Little Colorado Rivers from April 2009 to October 2017. Adult fish were captured using hoop nets and electro-fishing, measured for total length and given individual marks using passive integrated transponders that were scanned when fish were recaptured. The other three datasets were collected by US Forest Service. Owl data for the N-occupancy model was collected between 1990 and 2015. Owl data for the two-species [...]
Summary
These data were compiled here to fit various versions of Bayesian population models and compare their performance, primarily the time required to make inferences using different softwares and versions of code. The humpback chub data were collected by US Geological Survey and US Fish and Wildlife service in the Colorado and Little Colorado Rivers from April 2009 to October 2017. Adult fish were captured using hoop nets and electro-fishing, measured for total length and given individual marks using passive integrated transponders that were scanned when fish were recaptured. The other three datasets were collected by US Forest Service. Owl data for the N-occupancy model was collected between 1990 and 2015. Owl data for the two-species example was collected between 1990 and 2011. Both owl data sets were collected in a ~1000 km2 area in the Roseburg District of the Bureau of Land Management in western Oregon, USA. Owl vocalizations (vocal lures) were used to detect barred owl or spotted owl pairs in 158 survey polygons spread throughout the study area. The avian community occupancy data were collected from 1991 to 1995 across 92 sites in the Chiricahua Mountains of southeastern Arizona, USA. 149 species were detected through repeated point counts in each year.
The purpose of these data are to provide real-world examples to compare the speeds of different approaches to Bayesian mark-recapture and occupancy models. The data sets were originally created for prior publications and more generally, for monitoring programs. The data were collected to ensure compliance with the Endangered Species Act and other relevant laws and regulations and to determine the impacts of fire and forest management on various bird species. The objective of providing data here is to allow users to run associated code and reproduce the gains in speed reported in the manuscript by marginalizing code and switching to Stan, a platform for statistical modeling and high-performance statistical computation (https://mc-stan.org/).
Rights
The author(s) of these data request that data users contact them regarding intended use and to assist with understanding limitations and interpretation. Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although these data and associated metadata have been reviewed for accuracy and completeness and approved for release by the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data for other purposes, nor on all computer systems, nor shall the act of distribution constitute any such warranty.
