Derived parameters from a multi-season model

announcements (new versions, changes, bugs, installation problems...) related to program PRESENCE

Derived parameters from a multi-season model

Postby manuel_spinola » Sun Sep 25, 2016 8:43 am

Dear group members,

I am running multi-season occupancy model with the sample data NSO_pg209.csv

Untransformed Estimates of coefficients for covariates (Beta's)
======================================================================
estimate std.error
A1 psi1 : 0.509683 0.285328
B1 gam(1) : -1.524275 0.292907
C1 eps(1) : -1.796367 0.268769
D1 P[1-1] : 0.366350 0.163064
D2 P[2-1] : 0.089924 0.162215
D3 P[3-1] : -0.374344 0.182986
D4 P[4-1] : -0.469012 0.174247
D5 P[5-1] : 0.146605 0.157488

Real parameter estimates
=======================================================================

parameter site estimate std.error 95% conf. interval
psi1 1 site1 : 0.62473 0.06689 0.49362 - 0.75584

parameter site estimate std.error 95% conf. interval
gam(1) 1 site1 : 0.17883 0.04301 0.09453 - 0.26314
gam(2) 1 site1 : 0.17883 0.04301 0.09453 - 0.26314
gam(3) 1 site1 : 0.17883 0.04301 0.09453 - 0.26314
gam(4) 1 site1 : 0.17883 0.04301 0.09453 - 0.26314

parameter site estimate std.error 95% conf. interval
eps(1) 1 site1 : 0.14229 0.03280 0.07800 - 0.20659
eps(2) 1 site1 : 0.14229 0.03280 0.07800 - 0.20659
eps(3) 1 site1 : 0.14229 0.03280 0.07800 - 0.20659
eps(4) 1 site1 : 0.14229 0.03280 0.07800 - 0.20659

And my derived parameters

DERIVED parameters - psi2,psi3,psi4,...

Site psi(t) Std.err 95% conf. interval
1 site1 psi( 2): 0.6029 0.0514 0.5021 - 0.7038
1 site1 psi( 3): 0.5882 0.0515 0.4872 - 0.6891
1 site1 psi( 4): 0.5781 0.0570 0.4665 - 0.6897
1 site1 psi( 5): 0.5713 0.0626 0.4486 - 0.6940


DERIVED parameters - lam2,lam3,lam4,...

Site lam(t) Std.err 95% conf. interval
1 site1 lam( 2): 0.9651 0.0507 0.8658 - 1.0644
1 site1 lam( 3): 0.9755 0.0370 0.9030 - 1.0480
1 site1 lam( 4): 0.9829 0.0265 0.9309 - 1.0349
1 site1 lam( 5): 0.9882 0.0188 0.9514 - 1.0250

My question 2 questions,

1. How can I have a decreasing psi when colonization is larger than extinction?
2. How lambda is increasing if I have a decreasing psi?

Best,

Manuel
manuel_spinola
 
Posts: 4
Joined: Sat Sep 24, 2016 7:41 pm

Re: Derived parameters from a multi-season model

Postby darryl » Sun Sep 25, 2016 3:50 pm

1. Because colonization only happens at unoccupied places while extinction happens at occupied places, so as first year occupancy is > 0.5 there's more places that can get affected by extinciton. You need to compare occ*ext vs (1-occ)*col to determine if occupancy going to go up or down overall.

2. lambda is going up, but still less than 1 so occupancy decreasing; just not decreasing by as much.

All of these behaviour is because with constant extinction and colonization you have a stationary Markov process so the derived estimates of psi and tending to the equilibrium value of col/(col+ext)

Cheers
Darryl
darryl
 
Posts: 495
Joined: Thu Jun 12, 2003 3:04 pm
Location: Dunedin, New Zealand

Re: Derived parameters from a multi-season model

Postby manuel_spinola » Mon Sep 26, 2016 4:52 pm

Thank you very much Darryl.

Manuel
manuel_spinola
 
Posts: 4
Joined: Sat Sep 24, 2016 7:41 pm


Return to software problems/news

Who is online

Users browsing this forum: No registered users and 14 guests