Pinus pinea L. (Mediterranean stone pine) is an important forest species not only for its economically relevant kernel production, but also for environmental protection. The detection of genetic variability is an essential issue for Mediterranean forest species for conservation and improvement programs. Based on data collected for several years at three field sites from a P. pinea L. provenance trial established in Portugal in 1993, the present study aimed to evaluate the genetic variability of the adaptive traits of P. pinea L. and to identify a group of provenances with high performance for growth traits to be used in future plantings. Several mixed models were fitted to survival and growth trait data to estimate provenance and provenance × site interaction variability. The empirical best linear unbiased predictors of provenance genotypic effects were used to select a superior group of provenances. The provenance genetic variability of P. pinea L. was successfully detected for survival and height of different planting ages and also for diameter at breast height at age 13 years after planting. For growth traits, the most successful methods used for evaluating provenance genetic variability were based on the linear mixed spatial models. When multi-environmental analysis was performed, provenance × site interaction variability was detected for survival, but not for height at age 6 years. The existence of provenance variability in P. pinea L. permitted the identification of a seed lot composed of a mixture of the best provenances for height and diameter.
