Warm spring pairs gave no clear increase in annual lilac bloom shifts in four US samples in 1965-2008

We tested whether warmer pairs of springs had larger absolute year-to-year changes in recorded first bloom. Four US lilac samples used 6,556 consecutive-year pairs in eastern and western regions in 1965-1986 and 1987-2008. Models included plant and ending-year effects, absolute annual spring and winter temperature changes, and pair mean winter temperature. Spring pair-mean slopes ranged from +0.301 to +1.649 days per C. All four adjusted intervals included zero. Flag exclusion, signed temperature changes, longer plant records, and curved temperature terms gave positive but uncertain slopes. These records do not establish increased timing instability in warmer spring pairs. They also do not establish equivalence. Unknown visit schedules, selected consecutive reports, different lilac types, coarse seasonal weather, and few clusters limit inference. This is not a test of pollution, daily extremes, or causal climate change. The package includes the raw-data recipe, source hashes, derived panel, pairs, code, all results, numerical checks, and a figure.

phenology · lilac · annual variability · spring temperature · observation bias · regional replication · null result