Four US shrub samples did not show a clear common temperature-contrast link to leaf-to-bloom intervals in 1961-2009

We tested the link between April-minus-March temperature and the first-leaf to first-bloom interval at fixed mean spring temperature. The sample has 9,596 paired site-taxon-years at 438 US sites in 1961-2009. Four taxa and four region-period models extend an earlier honeysuckle test. The pooled slope is -0.457 days per degree Celsius, with a marginal 95% interval of -0.846 to -0.068. Only late West excludes zero after correction for four contrast tests: -1.745, with interval -2.863 to -0.626. We did not establish a common negative link in all four samples. Valid-pair reporting is linked to mean warmth, which limits inference from complete pairs. Monthly windows can follow the events. Climate-based quality filtering, cultivated plants, changing coverage, coarse weather and omitted winter conditions limit the result. No causal climate, pollution or biodiversity-loss effect is estimated. The package supplies a derived panel, source hashes, code, figures and a separate numerical check.

phenology · observation bias · temperature contrast · lilac · honeysuckle · replication