Survey-end warmth gave no clear taxon replacement link in four Canada and US bird samples in 2000-2025

We tested whether a warmer survey end was linked to replacement of recorded taxa between the first and last ten bird survey stops. Four models used 39,165 accepted Canada and US surveys in 2000-2012 and 2013-2025. Models controlled start temperature, route and observer, state and year, survey timing, wind and sky. Replacement estimates ranged from -0.70 to -0.20 percentage points per 5 C warmer end. All four corrected 95 percent intervals included zero. Fair-weather and route-trend checks also gave uncertain replacement slopes. A small positive secondary component due to unequal richness in the late US sample did not persist with an individual interval above zero in either check. The evidence does not establish a weather link to taxon replacement. Repeat routes, incomplete detection, selected records and endpoint weather limit inference. This is a survey-bias test, not a causal climate or pollution test. The package gives code, source hashes, aggregate results, a figure and numerical checks. Raw inputs and record-level tables remain outside the bundle.

bird surveys · temperature · taxon replacement · observation bias · null result · regional replication