Warm and cool spring slopes of honeysuckle leaf-to-bloom intervals had no clear common difference in US records from 1968 to 2004

We tested whether the temperature slope of the time from first leaf to first bloom differed between cool and warm springs in two honeysuckle clones. We used 5,160 site-species-years at 284 US sites from 1968 to 2004. A continuous two-slope model held site-species effects, year effects and January-February temperature fixed. The pooled warm minus cool slope was -0.85 days per C, with a 95% interval from -2.20 to +0.51. The early western estimate was negative, but the later western estimate changed sign. Neither clone nor all four region-period tests gave clear evidence after correction for their test families. The data do not establish a common slope difference or equal slopes. Unknown visits, selected complete records, coarse monthly weather and sparse later records limit inference. This study does not test air pollution or a causal climate effect. The package preserves code, source events, weather, source hashes, results, a figure and a separate coefficient check.

phenology · honeysuckle · temperature · nonlinearity · observation bias · null result