Survey date changes altered annual ozone contrasts in matched four-state bird surveys in 2005-2019

We split annual ozone changes into year and survey-date components in 1,341 matched bird survey pairs from California, Oregon, New York, and Pennsylvania. Each pair had the same route, observer, and ozone site. We held each observed month-day fixed in turn and used daily ozone data from both years. Across two regions and two periods, the date component had 29-36% of the root-mean-square size of the actual annual change. Ratios fell to 16-21% in samples with dates at most seven days apart. Actual and date-standardized annual changes remained strongly correlated. Primary adjusted bird taxon-count slopes all had intervals that included zero. A strict complete-data check gave a negative standardized slope in the later western sample. This result did not repeat across all samples. Date movement can change short-window pollution contrasts even when route, observer, and site are held fixed. It need not be exposure error: birds can experience different seasonal windows. Selected monitor coverage, missing daily data, survey detection, and unmeasured seasonal weather limit interpretation. We do not estimate climate damage, phenology shifts, or ozone toxicity. The package has daily evidence, code, hashes, all results, and offline checks.

ozone · birds · survey timing · exposure alignment · observation bias · Breeding Bird Survey · date standardization