Weird DSR results for Nest success

questions concerning analysis/theory using program MARK

Weird DSR results for Nest success

Postby gilu » Wed May 28, 2008 4:16 am

Hi all,

I am currently doing a multi year/multi site nest success analysis. Annual variation in nestsuccess is high, and our a priori hypothesis is that predation is the most important source of this variation. In addition to predation (measured as predatortracks/km for each area and year) we have a number of bird specific and environmental variables).

First model tries showed that indeed predation is the single most important variable, the predation model b0+b1*pred being the only single covariate model outcompeting the model including only year. Now, what we are interested in is what other variables affect the risk of depredation.

I have now a set of candidate models, each of which contains predation plus a maximum of two covariates (I chose only those models that outcompete the predation only model). Then I use model averaging for estimating average betas for the best variables and have a ranking list of best variables. The results look good and also make perfect biological sense. I get mean DSR's a little over 0.98, which also makes sense.

However, when I plot the predation covariate (over the range of observed values) against DSR (using the logit link as this is what I used) I get a curve of DSR that goes below 0.9 already at low predator densities and finally reaches approximately 0.3 at maximum predator density.

Since these are daily survival rates, this does not make sense to me. Considering an average incubation time of 25 days for the species in question, any DSR below 0.9 is too low to explain the observed frequency of nest success. At very low predator densities about 90% of nests have survived, but even at max. densities about 50% survive. These values are not mayfield corrected though, but even then the DSR across observed predator densities seems to go much too low.

Or do I simply misinterprete DSR?

Any help or thoughts are highly welcome!

cheers

Gilu
gilu
 
Posts: 4
Joined: Wed Apr 23, 2008 6:38 am

Postby gilu » Wed May 28, 2008 6:41 am

The problem has been solved :roll: :oops: :lol:
gilu
 
Posts: 4
Joined: Wed Apr 23, 2008 6:38 am


Return to analysis help

Who is online

Users browsing this forum: No registered users and 2 guests

cron