Equal partial R2 values of 0.77-3.44% can remove prior ozone slopes for recorded bird richness in four US samples in 1995-2024

We tested the sensitivity of prior-season ozone slopes for recorded bird richness to an omitted factor. A published US bird survey panel gave negative slopes in four East and West, early and late samples after weather and survey controls. All four adjusted intervals included zero. A hypothetical factor with equal partial R2 values of 0.77-3.44% for ozone and richness can remove each slope. The pooled slope was -0.472 species per 10 ppb (95% interval -0.745 to -0.200). Its equal-strength threshold was 2.17%. Measured survey and weather blocks gave smaller calibration bias bounds than the pooled slope. These thresholds describe linear regression sensitivity. They do not show that an omitted factor exists or that ozone has no effect. Monitor coverage, measurement error, detection and selected surveys limit inference. The periods share some routes. Code, aggregate results, source hashes, a figure and a second numerical fit are included.

ozone · birds · omitted variable bias · partial R2 · observation bias · weather controls