Thirteen year pairs did not establish lower winter annual plant turnover on two Portal control plots in 2003–2024

We tested whether winter annual plants had lower year-to-year species turnover than summer annual plants at Portal, Arizona. We used PortalData 6.117.0 and two plots that have always been unmanipulated controls. We held the sampled area at 8 square metres. Thirteen consecutive year pairs had complete sampling in both seasons from 2003 to 2024. Mean Sørensen dissimilarity was 0.536 in winter and 0.598 in summer. The winter minus summer difference was −0.063. A nominal 95% interval from 20,000 paired time-block draws was −0.311 to +0.200. The evidence did not establish lower winter turnover or equal turnover. Separate plots, other species rules, and block lengths one to four also left the contrast unresolved. Six of 32 seasonal samples had no named annual plants. A check restricted to seven pairs with no empty samples gave a more negative difference of −0.153, but its interval still included zero. The difference between earlier and later seasonal contrasts was also uncertain. The intervals condition on the selected plots and assume approximately stable short-range variation. Limits include one site, a small sample area, gaps, census timing, and unmeasured detection. The result describes observed plants above ground and does not identify a climate cause. The package contains unchanged source tables, provenance, code, derived data, resamples, figures, and offline analysis instructions. An independent CSV rebuild matched all 6,402 count cells and all 52 seasonal transition rows.

annual plants · temporal turnover · seasonal stability · Portal Project · permanent quadrats · empty census · reproducibility