Bootstrapping with Nest Survival Model?

questions concerning analysis/theory using program MARK

Bootstrapping with Nest Survival Model?

Postby MTMJY » Fri Jan 05, 2018 8:30 pm

Hello,

I am hoping the Mark community can help me with a question I have regarding the estimation of turkey poult survival:

I am working to estimate poult survival to 4 weeks of age based on frequent counts of unmarked poults located with a radio-marked hen. Constraints in the field (e.g. poor lighting, brood mixing, broods flushing before being located) result in counts where detection is either very reliable (brood observed roosted next to the marked hen) or completely unreliable (brood flushes before we are close enough to get a visual). This, combined with large brood sizes, makes the Young Survival from Marked Adults an inappropriate framework for this question.

As such, my intent is to discard "useless" counts and treat data from counts where I am confident we detected the entire brood as ragged telemetry data within the nest survival module. At least a few studies have used this approach (e.g. Garrick et al. 2017, Amundson and Arnold 2011). I've run some of my preliminary models without issue.

The issue lies in dealing with the lack of independence between brood mates. I intended to estimate overdispersion using bootstrap simulations within Mark. However, the nest survival model is not listed among the supported data types for simulation. As I mentioned above, I believe others have used this method as I am proposing, so I am a bit confused. Is it possible this functionality was once available but no longer exists for this model type?

Thanks much,
Michael
MTMJY
 
Posts: 6
Joined: Sat Dec 02, 2017 4:42 pm
Location: Montana

Re: Bootstrapping with Nest Survival Model?

Postby tlyons4 » Tue Feb 13, 2018 7:44 pm

5.6 in the MARK manual
tlyons4
 
Posts: 19
Joined: Thu Nov 30, 2017 7:29 pm

Re: Bootstrapping with Nest Survival Model?

Postby cooch » Tue Feb 13, 2018 8:27 pm

Assuming you want a GOF for nest survival, this isn't available in MARK. But, the following post from Jay Rotella (who wrote the nest survival chapter) should provide some guidance:

You might take a look at:
Sturdivant, R.X., J.J. Rotella, R.E. Russell. 2007. A smoothed residual based goodness-of-fit statistic for nest-survival models. Studies in Avian Biology 34:45-54.

and at various SAS code I provide for nest-survival goodness-of-fit testing at:
http://www.montana.edu/rotella/nestsurv/

Although I'm sure some users would appreciate being able to implement these tests in R, I only have code for implementing these GOF tests in SAS for mixed-effects or fixed-effect-only models.
cooch
 
Posts: 1628
Joined: Thu May 15, 2003 4:11 pm
Location: Cornell University

Re: Bootstrapping with Nest Survival Model?

Postby cooch » Tue Feb 13, 2018 8:31 pm

It is also worth having a read of Appendix G in the MARK book, concerning data bootstrapping, which can be applied to 'nest data' in a fashion.
cooch
 
Posts: 1628
Joined: Thu May 15, 2003 4:11 pm
Location: Cornell University

Re: Bootstrapping with Nest Survival Model?

Postby MTMJY » Tue Mar 13, 2018 12:23 pm

cooch wrote:It is also worth having a read of Appendix G in the MARK book, concerning data bootstrapping, which can be applied to 'nest data' in a fashion.


Thanks Evan,

I should have been more clear in my initial post. I attempted to use the "Bootstrap Data" option as is outlined in Appendix G, but in doing so, I get the following message: "Simulation not implemented for data type Nest Success." This is what has lead to my confusion, as I believe others have previously used the approach outlined in the appendix and by Bishop et al. 2008 on this type of data. Any clarification you could offer would be much appreciated.
MTMJY
 
Posts: 6
Joined: Sat Dec 02, 2017 4:42 pm
Location: Montana

Re: Bootstrapping with Nest Survival Model?

Postby gwhite » Tue Mar 13, 2018 12:31 pm

What you want is the bootstrap data option under the Simulation tab, not the parametric bootstrap under the Test tab.
gwhite
 
Posts: 329
Joined: Fri May 16, 2003 9:05 am

Re: Bootstrapping with Nest Survival Model?

Postby MTMJY » Tue Mar 13, 2018 12:37 pm

gwhite wrote:What you want is the bootstrap data option under the Simulation tab, not the parametric bootstrap under the Test tab.


Yes, and when selecting "Simulations --> Bootstrap Data..." that is when I get the error I mentioned above.
MTMJY
 
Posts: 6
Joined: Sat Dec 02, 2017 4:42 pm
Location: Montana

Re: Bootstrapping with Nest Survival Model?

Postby gwhite » Tue Mar 13, 2018 4:35 pm

Looks like I've never implemented that feature for nest survival models. Sorry.

Gary
gwhite
 
Posts: 329
Joined: Fri May 16, 2003 9:05 am

Re: Bootstrapping with Nest Survival Model?

Postby etanner » Tue Apr 17, 2018 5:17 pm

I'm curious if the original poster has found a solution to this problem. I have run into the same issue in both the known fate and nest survival model and was hoping it may have been resolved by now before moving onto another method.

Thanks.
etanner
 
Posts: 3
Joined: Fri Mar 30, 2018 3:44 pm

Re: Bootstrapping with Nest Survival Model?

Postby gwhite » Tue Apr 17, 2018 5:20 pm

Stay tuned. I'm getting ready to work on getting this programmed.

Gary
gwhite
 
Posts: 329
Joined: Fri May 16, 2003 9:05 am

Next

Return to analysis help

Who is online

Users browsing this forum: Google [Bot] and 10 guests