Start warmth gave no clear common link to quality failure in submitted Canada and US bird surveys in 1997-2024

We tested whether warmer survey starts were linked to failure of bird-survey quality review. We used 69,432 submitted standard runs from Canada and the US in two periods from 1997 to 2024. Four models controlled route-observer effects, state-year effects, date, start time, wind and sky. Estimated changes at 15 C ranged from 0.04 to 0.53 percentage points of failure per 5 C. All four intervals included zero after correction for four tests. The early US estimate was positive in two sensitivity checks. The results do not establish a common temperature link or show that accepted surveys are free of weather selection. Cancelled surveys, incomplete records, few year clusters and unmeasured local conditions limit inference. The package preserves code, an input panel, source hashes, checks and a figure.

observation bias · Breeding Bird Survey · temperature · quality control · sample selection · biodiversity