Predicting flowering of 130 plants at 8 locations with temperature and daylength

Abstract

An improved plant phenological method is needed to accurately predict flowering of a large array of plant species at locations with a wide range of latitude. Degree days or degree days times daylength cannot be used to accurately predict flowering of both early and late flowering species when grown at locations with wide range of latitude. Published flowering dates of 130 plant species from among 8 locations in central North America ranging in latitude from 39 to 50 degrees N and longitude 84 to 108 degrees W were used to develop a degree days times daylength factor to predict flowering dates. Plants flowering in late June flowered at the same time at all 8 locations regardless of latitude. Species flowering earlier than late June flowered earlier at southern locations than those at Treesbank, Manitoba. Species flowering after late June flowered later at southern locations than those at Treesbank. Flowering of 124 species divided among 8 locations was most accurately predicted by the accumulation of degree days (threshold = 2 degrees C) times daylength factor (1/(0.259-0.0140*daylength) from the first of December. This method slightly discounts daylength below 13 hours and greatly increased its weight for every hour over 13 hours. This method predicted flowering dates with a standard deviation of 0.1, 0.5, -1.7, 2.4, -0.1, 6.0, -1.8, and -1.1 days for Swift Current, Saskatchewan; Treesbank, Manitoba; Sidney, Mont.; Fargo, N.D.; Sauk and Dane Co., Wisc.; Wauseon, Ohio; and Manhattan, Kans.; respectively. Degree days or degree days times daylength had a standard deviation of 10 and 18 days in predicting flowering dates at Manhattan, Kans.