Single season occupancy Covariate

[Deepak]

Dear all,
We are analysing data on single season occupancy for turtles in India.
If I have a covariate like disturbance, how do I incorporate in presence analysis? Say if I have varying degree of disturbance like use of forest trail, fire wood collection, hunting turtles and nearest hamlet from our sampling trails, or a combination of two or all the four disturbance factors how do we quantify and incorporate this information in the covariate.

Including it as just "1" "0" presence absence as site covariate doesn't make any sense . Can anyone suggest how do we incorporate this information.

One more thing when I went through the Single season Occupancy with site covariates. I didn't understand the idea of reference habitat, while designing habitat as a covariate. To be more clear how do one interpret the result if there is any effect of the reference habitat? Can anyone clarify my doubt.

Regards,
Deepak

[darryl]

Hi Deepak,
You could either use multiple variables to represent each aspect of possible disturbance and enter different combinations of these into the model as see which combination provides you with the 'best' model. Or look at the combination of variables and categorise areas into different groups, eg 1=high use trail, hunting, near hamlet; 2 = high use trail, no hunting, far from hamlet; 3 = low use trail, no hunting, far from hamlet; etc. You could then use the categories as a categorical (or discrete valued) covariate.

Covariate values don't' have to be just 0-1, look in the 'overview' in the help menu. Admittedly it's somewhat out of date, but info there still relevant.

When you talk about a reference habitat, I presume you're talking about a situation when you're using dummy or indicator variables (ie 0-1 covariates) to indicate habitat type at sites. If you're not sure how to use them I suggest you do some more background reading on simple regression or generalised linear models. In fact I'm guessing a quick google of 'dummy variables' will provide you with something.

Cheers
Darryl