## Difference with Sébastien's report (Delmotte et al. 2010):

## - set probability of capture at Eu in 2000 and 2001 to be smaller than the other years (partial trapping)

## - didn't consider recapture probability in 1989 and 1993 because trap in Beauchamps was not working those years (and so in 2000 and 2001 as in the report)

## - adding a flow effect in pi_Eu. Considering a different temporal window depending on sea age. Flow data are standardized (ln(Q[t]) - mean(ln(Q)))/sd(ln(Q)) within the model. Residuals are standardized and followed.

########### - 15 june - 31 august for 1SW

########### - 15 april - 30 june for MSW

## - adding another effect of flow corresponding to the second peak of migration.Flow data are standardized within the model.

########### Same temporal window for 1SW and MSW: 1 octobre - 30 november

## - calculating calculating R² (% of variation explained by the two covariates in the probability of capture at Eu)