Spatial analysis of phenotypic variables in a clonal orchard of Pinus arizonica Engelm. In northern Mexico

Abstract

Introduction: Seed orchards provide germplasm genetically suitable for use in forest restoration. Knowledge of the spatial distribution of attributes is crucial for their management. Objective: To model cone production and tree size variables in a clonal orchard of Pinus arizonica Engelm. from a geospatial perspective in order to determine their behavior and distribution. Materials and methods: The spatial pattern of tree size variables and cone production of 126 ramets were determined through a geospatial analysis, using the Getis-Ord G statistic. A Pearson correlation analysis (P ≤ 0.05) determined the variables best associated with cone production and these were examined with stepwise regression. In terms of cone production, the best combination was modeled through a geographically weighted regression. Results and discussion: Statistically significant (P < 0.01) clustering values were found in the orchard. Correlation analysis showed that all tree size variables, including the moisture index, were statistically related to cone production. Stepwise regression identified a model that presented crown diameter as the variable that best explained cone production. Geographically weighted regression showed that crown diameter moderately influenced cone production. Conclusion: Tree size variables and cone production presented a tendency towards clustering. The use of a geospatial perspective allowed a better understanding of the spatial dynamics of tree size variables.