measure of encounter / competition risk

[laura labarge]

Hi Chris,

I was wondering if you could shed some light on a ctmm-related question I have?

I'm working on a project where I'm looking at spatial variation in a wildlife behavior (communication related) and using an RSF-like approach with models for each individual. My aim is to assess what social environment variables (odds of encountering other competing males and potentially receptive females) might affect where this behavior occurs.

So far, my predictors for location-specific competition (or mate) encounter risk have been CDF values extracted from rasterized AKDEs (separate predictors for each female and potentially competing range-resident male that occur within the study area to reflect how individualized these relationships likely are).

But after discussing with others who know a bit more than I do, I'm thinking using the CDF might not be the best approach - mainly because the study animals all overlap considerably (despite being solitary).

My question is, in this instance would you continue to use the CDF as a proxy for the odds of encountering another individual? We've thought of some alternatives (e.g. creating a distance raster to each competitor's and female's 50% highest use area).

I also shied away from creating CDEs using encounter() because we have so few actual encounters in the dataset to verify these. But maybe extracting the CDF of joint encounter distributions would actually be best?

Really appreciate any input you might have,

Laura

[Christen Fleming]

Hi Laura,

If you want an overall measure of how likely two individuals are to encounter each other, then there is a quantity related to the CDE that is proportional to the overall encounter rate. I don't have this currently exported, but could do so in a couple of weeks, if that would help.

There are also some distance functions in the package: help('distance') .

Best,

Chris

[laura labarge]

Hi Chris,

Thanks so much for your response! A measure of encounter likelihood sounds exactly like something we need!

I will check those distance functions out in the meantime!

Thanks very much!

Laura

[Christen Fleming]

Hi Laura,

I pushed a rates() function to the GitHub development branch of the package, which you can install via devtools::install_github("ctmm-initiative/ctmm")

This gives a relative estimate of the pairwise encounter rates (with CIs), under the assumption that the movements are not correlated when they aren't encountering each other.

Best,

Chris

[laura labarge]

Thanks so much Chris! Sorry for the delayed response but this looks great!