Modeling a species presence/absence across a landscape is an increasingly
common procedure in wildlife research. However, rarely does this research
account for the autocorrelated nature of the data one might collect at,
say, point counts. I would like to incorporate the spatial dependency of
this sort of data to improve the output of my modeling efforts.
My interest is in predicting the historical presence of a species across
the state of Illinois. I could do this with autologistic regression with
covariates. However, I could also do this with some sort of indicator
kriging that incorporates covariates. The covariates in both cases would
be environmental variables such as patch size, cover type, proximity to
Is anyone aware why one method might be preferable to the other? I have
not seen any literature on indicator co-kriging though I'm sure it must
exist. Can anyone point me to relevant literature and to codes or programs
that would facilitate either procedure?
Wayne Thogmartin wthogma@...
Upland Wildlife/Habitat Project 618.453.5495 (office)
Mail Code 6504
Cooperative Wildlife Research Laboratory 618.453.6944 (fax)
Southern Illinois University
Carbondale, IL 62901-6504
"The dwarf standing on the shoulders of a giant sees farther than he who
*To post a message to the list, send it to ai-geostats@...
*As a general service to list users, please remember to post a summary
of any useful responses to your questions.
*To unsubscribe, send email to majordomo@...
with no subject and
"unsubscribe ai-geostats" in the message body.
DO NOT SEND Subscribe/Unsubscribe requests to the list!