What do we still need to faff about with, wrt model fits?
negative binomial
looking at other outliers when harbour porpoises are removed (identify), if mass strandings consider removing?
trying extra factor smooths, for s(storms, body size), s(geomagnets, species) etc. Or at least seeing if different species/body size do seem to have different responses
backwards model selection just for interest/for the appendix/reviewers
What do we still need to faff about with, wrt model fits?
identify), if mass strandings consider removing?