Incorporating rainfall seasons in multiseason models

[rk1177]

Hello. I am having trouble trying to incorporate seasons into my models for presence.


The design is a 100 site sampling transect that was sampled 3 times a year for 3 years. So a 3 -3 arrangement with 900 data points. Within this, the samples were taken during different seasons based on rainfall: wet dry and transitional.
So this makes for a 1-1 wet, 1-2 dry, 1-3 transtional, 2-1 wet, etc.

In addition each sampling site has a tree family tied to it, which I have dummy coded as a site covariate.

My goal is to determine if the insects show any preference on where they are found (family of tree) based on season (rainfall) and if they are migrating within a year. More specifically, would I use the wet, dry, transitional information as a site covariate or survey covariate? Coupled with that would be how to code for these depending on the most appropriate covariate type.

Thank you for any help/direction you can give.

(Posted on behalf of L. Barringer)

[mferreira]

Hi

I believe that it should be a co-survey variable. Use one for each level.

[rk1177]

Thanks for the response.

I'm not familiar with "co-survey variables", however. Could you please elaborate?

Cheers,
LB