I'm using eBird count data to try and look at changes in bird distributions over time throughout the urban gradient. I'm an undergrad so my stats level is pretty limited. I took the relative count (observed count/total count in a year) of observations from 2003 to 2023. I'm trying to run a GLM poisson right now but it is not going very well. Any advice on this model or another model would be greatly appreciated.
Edit: I didn't transform my data, it is not normal
I tried running this code:
glm_interactions<-glmer(RelCount~ urban_category*year+ latitude + longitude + number_observers + day_of_year + (1|county) + (1|protocol_type), family="poisson", data = DF)
But it gave me
50+ warnings (1: In (function (fr, X, reTrms, family, nAGQ = 1L, verbose = 0L, ... :
non-integer x = 0.000321)