Seasonal harebell width contrasts had opposite signs at two ozone targets in a North Wales warming experiment in 2021

We tested whether an ozone by warming contrast in harebell flower width changed in the same direction from early to late summer at two ozone targets. We used four monthly visits in one North Wales experiment in 2021. The changes were -0.1224 log units for 80 versus 30 ppb and +0.0659 for 110 versus 30 ppb. The latter became negative when September was omitted. Each treatment had one dome. Treatment columns added no rank to chamber indicators. Fixed chamber offsets cancel in the seasonal contrast, but errors that change across months do not. Exact bounds include zero when allowed cell errors reach 1.54% and 0.83%, respectively. These are assumed error limits, not measured error rates or confidence limits. Missing flowers, selected width measurements and unequal achieved exposures limit inference. The results do not identify a causal warming effect or replicate across regions or years. Code, source hashes, aggregate results, checks and a figure are included. External source records remain outside the publication.

ozone · warming · flower width · harebell · chamber effects · observation bias · sensitivity analysis · seasonal contrast