Survey-start warmth gave no clear next-year observer continuity link in Canada and US bird surveys in 1985-2023

We tested whether warmer survey starts predicted failure to record the same observer on the same bird survey route in the next year. Four Canada and US models used 73,910 accepted runs in 1985-2003 and 2005-2023. Models controlled route, state-year, date, clock time, weather and observer experience. Slopes ranged from -0.45 to +0.54 percentage points per 5 C. All four adjusted intervals included zero. End-temperature, fair-weather and no-tenure checks did not establish a common link. The outcome combines missing accepted surveys and observer changes. It is not confirmed volunteer loss. Selected survey weather, short route histories, missing reasons for nonreturn and unmeasured local change limit inference. We did not test pollution or causal climate effects. The package gives the selected panel, source hashes, code, all results and numerical checks.

observation bias · weather · temperature · observer continuity · Breeding Bird Survey · null result