Résumé
One definition of horticulture is "the art of cultivating garden plants" and pruning is a horticultural practice that is traditionally approached as more of an art than a science. This is largely because of the complexity of tree growth and development and a lack of general understanding and appreciation about the processes involved in governing shoot and tree growth and development. However recent tree architectural studies have provided systematic analyses of the shoot growth and statistical and dynamic simulation models have been developed that predict tree development and responses to pruning based on scientific concepts. These concepts include apical dominance (and its subcomponents; correlative inhibition, apical control and shoot epinasty); prolepsis and syllepsis; preformation and neoformation; epicormic shoot formation and plastochron (leaf emergence rates). In this paper we will discuss how many of these concepts can be combined with hidden semi-Markov chain models of shoot bud fates and a simulation model of source-sink interactions in peach trees (L-PEACH) to understand and predict natural development of peach trees and their responses to pruning. The results of these modeling efforts help explain the architectural and physiological basis of several common, empirically-based pruning systems used in California. These concepts also provide an understanding of the limitations of relying primarily on the use of pruning to control size of trees growing on commonly used invigorating rootstocks. This research demonstrates how computer simulation modeling can be used to test and analyze interactions between environmental factors and management practices in determining patterns of tree growth and development.