(almost) complete population

Dear friends,

I am in strange situation now:
I analyze data about decline of bats on the island, and I have got data
about almost complete population (no substantial migration). However,
there are only 7 time replicates.
I would like to link the decrease in number of bats with weather
conditions in those 7 times (years), and improve my fit with the fact
that I have data about vast majority of the population (clearly not a
tiny sample). Is there any way how to get the advantage form this fact,
e.g. how to be able to use this information in order to use more than 1
weather predictor reliably and avoid overfitting?

Any advice?

With thanks in advance,
best,
Martin Weiser