Distinguishing Seedlings of Pines of Piedmont Upland Grassland Systems


Alexander Krings

Additional Authors:


January 2020


Bayesian, juvenile trees, prairies, probabilistic identification, seedling identification

Despite the ecological and economic importance of pines (Pinus) in the southeastern United States, there exist no taxonomic keys or resources that facilitate the comprehensive identification of pine seedlings. The objectives of this study were to analyze select aspects of seedling morphology for the three primary pines associated with Piedmont upland grassland systems of conservation concern (Pinus echinata, P. taeda, and P. virginiana) and subsequently to develop a resource, traditional or probabilistic, depending on the outcome, to facilitate species-level determination. Two characters— the number of primary needles in the first whorl post-germination and needle apical curvature relative to the stem—were assessed through analysis of 3,200 seedlings under field conditions. The mean number of primary needles in the first whorl was 5.9 (s.d.=0.56) for P. echinata, 7.3 (s.d.= 0.79) for P. taeda, and 5.3 (s.d.=0.63) for P. virginiana. Predominant needle apical curvature was away from the stem in P. echinata (90.5%) and P. virginiana (91.5%), but toward the stem in P. taeda (97.5%). Based on these data, a probabilistic diagnostic framework that takes into consideration both morphology and stand composition is presented, including supplemental appendices of calculated posterior probabilities.

Corrigendum: The calculation provided in the formula on p. 37 should result in 0.05201, rather than the printed 0.5201.