Nest Survival Model Question

[mikad]

Hello all,

I am using nest survival model "nest" in Mark with the RMark interface. I was wondering if there is a way to incorporate nests which were found fully depredated. For example, when searching for turtle nests, I found several excavated nests in which the eggs were all destroyed before they were found. I'd like to find a way to incorporate this into my modeling, but am unaware of any way to make this work using the Mark "nest" data structure because the last day the nest was present is unknown.

Thank you in advance for the help. If I need to provide additional information, please let me know.

-M

I posted to the wrong sub forum originally. Sorry about that.

[roobydooby]

Hi mikad,
Just to make sure I understand your question correctly, you want to separate failures into two categories, those that failed due to predation and those that failed due to some other cause, correct?

If so, this is also something I've tried to do with my data. The short answer is, no, in the nest survival module you cannot separate failures by type to look at, say, how covariates vary in their ability to affect different failure types. You could build two model sets depending on what you want to do with your end product (i.e., one with depredated nests and successful nests, and the other with nest failing to other reasons and successful nests). Another option suggested to me that I have been working with (but, have not yet mastered) is re-formatting your data to analyze it as multi-state data. You can read about it here:

Devineau et al. 2014. Increased flexibility for modeling telemetry and nest-survival data using the Mulitstate Framework. JWM 78:224-230.

You'll also need to read ch. 10 in the MARK book. Hope that helps.

[mikad]

Hello Roobydooby,

Thank you for your response. Let me clarify my question. I do not wish to separate out my nest failures into different categories because I believe that all of my nest failures (at least in this case) were due to predation events. My issue is that because I found the nests fully depredated, I cannot determine the date the nests were "last present" in order to fit the nests into the nest survival data framework. Therefore, I haven't included them in my data analysis.

Thank you for the paper citation! I will look into multistate framework as an alternative option.


Cheers,

Mika