Mean leaf-date gaps fell across 80 Swiss stations from 1961–1990 to 1996–2025, but correlation changes were uncertain for four species

Did annual spring leaf emergence become more correlated between Swiss stations? We tested beech, hazel, larch and spruce with unchanged MeteoSwiss open data. The same 5339 species-specific station pairs at 80 stations had at least 20 shared years in both 1961–1990 and 1996–2025. We removed linear trends on the shared years and gave equal weight to each species. The mean Fisher z change was −0.0777, with a nominal 95% year-block interval of −0.2135 to +0.1139. The transformed mean correlation changed from 0.480 to 0.418. Its change interval was −0.170 to +0.084. Thus the data did not establish an increase or decrease. In the same pairs, the mean absolute calendar-date gap fell from 15.48 to 13.44 days. This change was −2.04 days, with an interval of −2.89 to −1.07 days. Stricter coverage checks gave the same distinction. The correlation bootstrap rejected 180 of 2000 draws because a fixed pair had too little data; a stricter panel and a block-deletion check also left correlation change uncertain. Smaller calendar gaps can occur without a clear change in annual spatial correlation. The result applies to this station network and these four species. It does not measure frost damage or identify a cause. The package preserves source files, provenance, code, all estimates, figures and an exact offline reproduction check.

phenology · spatial synchrony · leaf emergence · Switzerland · MeteoSwiss · reproducibility