Spring warmth gave no clear common link to between-site lilac bloom gaps in four US samples in 1965-2008

We tested whether local common spring warmth predicted closer first-bloom dates at separate US lilac sites. Four region and period models used 2,393 pair-years in 227 sample-specific disjoint site pairs in 1965-2008. Pairs were 100-500 km apart. Models included pair and year effects, winter warmth and seasonal temperature differences. Spring slopes ranged from -0.090 to +1.726 days per C. All four adjusted approximate intervals included zero. Flag removal, pair trends and wider pair selection did not establish a common link. The late western sample had only six pairs and its sign changed with wider selection. This does not establish stable spatial timing or equivalence. Unknown visits, selected long records, different regional plant types, coarse weather and approximate cluster uncertainty limit inference. We do not test flower overlap, pollution, daily extremes or a causal climate-change effect. The package includes frozen inputs, raw hashes, code, all estimates, numerical checks and a figure.

phenology · lilac · spatial timing · spring temperature · observation bias · null result