Perfect Separation Detected Results Not Available. Compute the diagonal of the hat matrix Parameters-----observed. Begingroup You may well have separation based on the warning possibly on the z p values but it is not what you describe in your 1st paragraph.
Conversely as enounced in 3 besides the global mis-alignments fixed by. Unlike in ordinary least-squares regression for modeling a normally distributed response when a logistic model perfectly or nearly perfectly predicts the response that is separates the response levels unique maximum likelihood estimates do not exist. Take a look at this blog post by Andrew Gelman and you will see that it is not always a problem of perfect separation but sometimes a bug with glm.
Since supplementary food use nest initiation date are both covariates the implication is that you have some collinearity confounding.
So check first with other software like Stata. In the case of a canonical link function both are the same. Its analogous to a perfect prediction in a linear regression. Take a look at this blog post by Andrew Gelman and you will see that it is not always a problem of perfect separation but sometimes a bug with glm.