M-Surge - External covariate

[L.Schmaltz]

Hi,

My aim is to study the age effect on “breeding attendance” (or probability to continue the breeding process) at a colony for greater flamingos along the breeding process and explore links with environmental variables.

I’m considering a multistates model applied to the resighting of banded flamingos at a colony during one breeding season, with the following states:
« 1 » flamingo with an egg
« 2 » Flamingo with chick on island
« 3 » Flamingo with chick in creche out of island.

The season of reproduction is divided in 16 time intervals of 10 days (April to September). (For example for each time interval, a flamingo seen with an egg at least once during these 10 days will be noted as state “1”)

I considered also 3 groups.

After my selection model, my best model considered an additive group effect only on the “survival” in state 1 (S1 – here more a probability to continue to incubate then). All “survival” probabilities considered are time dependent.

In GEMACO:

ForSurvival:(47):[f(1).[g+t]]+[f(2,3).t]

ForTransition(29):[f(1).to(2).t(2_15)]+[f(2).to(2).t(3_15)]+[f(1,2).to(3).t(9_15)]+[f(3).to(3).t(10_15)]+others

ForCapture (27):[to(1).t(1_10)]+[to(2).t(2_10)]+[to(3).t(9_15)]+others


Looking at environmental variables during the study period, I observed a good correspondence between the variation of S1 over time and the occurrence of rain on the breeding site.

I would like to explore if heavy rainfalls during the incubation period can explained variations of S1 over time.

In GEMACO, is the following sentence correct? [f(1).[i+t*x]]+[f(2,3).t]

How I can include the group effect, and conserve in the same time the additivity between group and time [g+t]?

How to explore the magnitude of the effect of the covariate over the 3 different groups?

Thank you in advance,

Lucie.

[CHOQUET]

Hello Lucie,

> In GEMACO, is the following sentence correct? [f(1).[i+t*x]]+[f(2,3).t]

YES

> How I can include the group effect, and conserve in the same time the additivity between group and time [g+t]?

[f(1).[g+t*x]]+[f(2,3).t] should work

> How to explore the magnitude of the effect of the covariate over the 3 different groups?

[f(1).g.[i+t*x]]+[f(2,3).t] lead to three interceps and three slopes (one for each group).

Rémi