Corrected intervals were below zero for one of four prior-minus-future ozone and warmth contrasts in US bird surveys in 1995-2023

We tested whether prior-summer and next-summer ozone and warmth product slopes differ for recorded bird richness. We used a published US bird survey panel and four region-period samples in 1995-2023. Models controlled route-observer pairs, region-year changes, climate and survey conditions. Only the early western contrast had a four-test corrected interval below zero: -0.946 species per 10 ppb per 2 C, with interval -1.564 to -0.329. The other three intervals included zero. Removing each state kept this contrast negative but could greatly widen uncertainty. A common time-specific pattern did not repeat. Future exposure cannot cause an earlier count, but shared causes and serial exposure correlation can affect both slopes. Monitor error, future-record selection, detection bias and few western clusters limit inference. This is not a causal or equivalence test. The bundle includes code, aggregate estimates, a figure, exact input hashes and a verified retrieval script. External record-level data remain in cited signed publications.

ozone · seasonal warmth · bird richness · negative control · time order · observation bias · regional replication