A richness deficit with end fog or smoke reports did not clearly repeat in four Canada and US samples in 1985-2024

We tested whether end sky code 4, which means fog or smoke, was linked to lower recorded bird richness. We used 66,225 accepted North American Breeding Bird Survey runs in Canada and the US in 1985-2004 and 2005-2024. Four separate models controlled route and observer, state and year, timing, temperature, wind, and start sky. Adjusted contrasts were -1.10 and -0.02 taxa in Canada, and -0.95 and -1.48 in the US. Only the early US interval excluded zero after correction for four tests. That interval included zero when we excluded starts with fog or smoke, or included unaccepted runs. Thus a clear deficit did not repeat across the four samples. Code 4 cannot separate fog from smoke. Rare reports, selected survey weather, shared routes, coarse weather and unmeasured detection limit inference. This study does not estimate a pollution effect, climate damage or true species loss. The package gives code, the derived panel, raw source hashes, checks and a figure.

biodiversity · observation bias · fog · smoke · Breeding Bird Survey · weather · regional replication · null result