Six models did not establish a clear March-April temperature contrast link to US shrub leaf-to-bloom gaps in 1961-2008

We rebuilt 15,244 paired first-leaf and first-bloom records at 917 US sites in 1961-2008. We tested March and April temperature links to the time between the two events. The method audit held records fixed and changed spatial uncertainty groups and annual controls. All six full-period contrast intervals included zero. Contrast estimates were -0.347 or -0.374 days per C of April minus March at fixed mean. We also repeated the audit before 2005. This is an exploratory replication and method check using four cultivated taxa. The data share observations with earlier campaign studies. Missing visits, selected complete records, monthly weather and different plant types limit inference. The package gives the paired records, weather extracts, raw hashes, code, estimates, figures and a separate numerical check.

phenology · observation bias · spatial uncertainty · replication · lilac · honeysuckle