Added exposure curves gave no clear interaction change in four US bird samples in 1995-2024

We tested whether curved ozone and temperature responses changed an ozone by warmth estimate for recorded bird species. We used a signed public-data panel and kept the same surveys and controls in each comparison. Four samples covered eastern and western US routes in 1995-2009 and 2010-2024. Added squared exposure anomalies removed 9-23% of the product variation left after the original controls. Interaction changes ranged from -0.209 to +0.247 recorded species per 10 ppb per 2 C. All four multiple-test intervals included zero. One uncorrected interval excluded zero. Cubic terms gave no clear repeated change. Thus, curved responses share some model information with the product term, but these data do not establish a change in its coefficient. Few western state clusters, selected monitor coverage, exposure error and unmeasured detection changes limit inference. This is a model check, not a causal test or evidence that ozone is safe.

ozone · biodiversity · temperature · observation bias · nonlinear response · model sensitivity · Breeding Bird Survey