Manuela M. P. Huso, Dan Dalthorp, David Dail, and Lisa Madsen. 2015. Estimating wind-turbine-caused bird and bat fatality when zero carcasses are observed. Ecological Applications 25:1213–1225. http://dx.doi.org/10.1890/14-0764.1


Supplement

R code to calculate posterior distribution of M, total fatality at a wind facility when zero carcasses are observed.
Ecological Archives A025-072-S1.

Copyright


Authors
File list (downloads)
Description


Author(s)

Manuela M. P. Huso
U.S. Geological Survey
Forest and Rangeland Ecosystems Science Center
Corvallis, OR 97330, USA
E-mail: [email protected]

Daniel H. Dalthorp
U.S. Geological Survey
Forest and Rangeland Ecosystems Science Center
Corvallis, OR 97330, USA
E-mail: [email protected]

David Dail
Chase Bank
Columbus, OH 43240, USA
E-mail: [email protected]

Lisa J. Madsen
Department of Statistics
Oregon State University
Corvallis, OR 97331, USA
E-mail: [email protected]


File list

Xeq0_Rcode.R (MD5: 0189df9bc7f877f78943a34f377c950c)

Description

The Xeq0_Rcode.R file is an R script file that calculates the posterior distribution of M, the total fatality at a wind facility, assuming observed count of carcasses=0. The posterior distribution is a function of the prior, the overall probability of detecting a carcass killed at the facility (g), and the number of carcasses counted during the search process (x). Since we are assuming x = 0, we only need posterior distributions for each combination of g and prior. For a given prior, the code below calculates an array with a posterior for each value of g. There are three possibilities for g: fixed, uncertain (1.732x), highly uncertain (3x). The degree of uncertainty about g is given in terms of the width of the CI in terms of odds ratios: if Upr and Lwr are upper and lower bounds on CI for g, then odds(Upr)/odds(Lwr) is a measure of uncertainty about g.