A model of genetic variation for Pinus ponderosa in the Inland Northwest (U.S.A.): applications in gene resource management

Abstract

Models were developed to describe genetic variation among 201 seedling populations ofPinusponderosa var. ponderosa in the Inland Northwest of the United States. Common-garden studies provided three variables that reflected growth and development in field environments and three principal components of six variables that reflected patterns of shoot elongation. Regression models were developed for describing genetic variation across the landscape. Using functions of latitude, longitude, and elevation as descriptors, these models produced values of R 2 that were as large as 0.66, while averaging 0.39. The models described genetic variation as occurring along relatively steep elevational clines and gentle geographic (i.e., latitudinal and longitudinal) clines. An exercise at validating the models with independent data supported their veracity. Predictions made by the models are applied to limiting seed transfer, designing breeding zones, planning gene conservation programs, interpreting phenotypic variation, and predicting the effects of environmental change on the adaptedness of populations.