I think it is definitely convincing to say that urban bias is increasing! The boxplots or 'effect of urban cover', those are some sort of modelled results? Or what does that mean?
Couple aesthetic things.
get id of the amphibians, butterflies, and birds label and we can replace with phylopics maybe. Or at least add phylopics to the end.
I would make amphibians 0, 1 on the y-axis only cause the other two are, so it would make sense to display with the same y-axis scale?
currently looks like this. Is this convincing enough?