Recorded visit intervals support leaf before bloom in 629 of 634 selected US shrub records in 2009-2014

We tested whether time between visits can explain the recorded order of first leaf and first bloom in the same shrub. We used US lilac and honeysuckle records from 2009 to 2014. Correct Yes and No reports and one onset per event define a date interval. Sharp bounds on bloom date minus leaf date require leaf before bloom in 629 of 634 selected plant-years at 182 sites. Five orders remain unresolved. The pattern holds in eastern and western samples in two separate periods. All 460 records with visit gaps of at most seven days require leaf before bloom. Another 483 paired records lack usable brackets and are outside this conclusion. The bounds test an observation explanation; they do not estimate a climate or pollution effect. Selected sites, planted shrubs, missed events, and report errors limit wider use. The package gives source rows, date bounds, code, hashes, tests, and a figure.

phenology · interval censoring · observation bias · leaf emergence · flowering · lilac · honeysuckle · data audit