NEW THIS YEAR! The schedule of technical sessions is in Sched.org which allows you to search within the schedule, filter the schedule to show sessions only occurring on a certain date, within a track, or in a room. You can also build your own schedule by creating a free account in Sched.org. Click here to return to the main Midwest Fish & Wildlife Conference website.
PLEASE NOTE: The schedule posted here is as of 1/25/16, and is subject to change. Please check back for updates.
AUTHORS:Brian A. Maurer*, Michigan State University
ABSTRACT: Data on fish and wildlife communities are often collected over long periods of time. Such data are meant to represent dynamical processes going on in a multidimensional state space defined by the abundances of all species in the community. The intent of any analysis of such data is to infer something about the processes that affect the community from the multivariate time series that represents the community. Dynamic factor analysis provides a unique approach to this goal. The basic concept is that the measures of abundance for each species in a community are indicators of some latent trend common to all species in the community. The approach begins by modeling these latent trends as simple random walks. For each latent trend, a “loading” is estimated for each species that represents the degree to which that species is associated with that trend. I will discuss the model building process for DFA and address a few computational issues. A brief discussion of R code implementation and results will be illustrated using data from a long term study of desert rodents.
Monday January 25, 2016 2:20pm - 2:40pm EST
Vandenberg B
