Local spring slopes were lower than displaced slopes in three of four US shrub samples in 1961-2004

We tested whether first bloom had a more negative link to local spring temperature than to temperature 10 degrees of longitude away. Four plant and year models used 8,456 paired leaf and bloom records in the US East and West in 1961-1982 and 1983-2004. Models controlled local and displaced January-February temperature. Local minus displaced March-April slopes were -3.36, -3.40, -2.64 and -3.91 days per C. Three four-test adjusted intervals were below zero. The later western interval included zero. All planned sensitivity estimates were negative, but some intervals included zero. This supports geographic detail in the association in three samples under the model assumptions. It does not show a causal temperature effect. Shared weather, coarse grids, missing visits, selected paired records and different plant types limit inference. The package gives source hashes, paired weather data, code, all results, a figure and numerical checks.

phenology · spatial specificity · temperature · observation bias · lilac · honeysuckle