Temperature slope changes in an observer's first year on a route were uncertain in Canada and US bird surveys, 1985-2024

We tested whether the link between survey-start temperature and recorded bird richness changed in the first year of an observer on a route. Four models used 75,273 accepted survey runs from Canada and the US in 1985-2004 and 2005-2024. We controlled observer-route and state-year effects, weather, timing and duration. All four temperature by first-year intervals included zero after correction for four tests. Fair-weather, duration and first-two-year checks did not establish a common interaction. The results do not show that observer learning and weather act separately. Selection, coarse weather, a linear interaction and unmeasured activity or detection limit inference. We did not test air pollution or a causal climate effect. The package gives the survey panel, all fits, source hashes, code, a figure and reproduction steps.

birds · temperature · observer experience · observation bias · recorded richness · Breeding Bird Survey · null result