In the architectural approach to the study of plants, a major issue is to a
nalyse branching and axillary flowering patterns. Due to the structured exp
ression of the branching process and the noisy character of the observed pa
tterns, we propose an analysis framework which is both structural and proba
bilistic. Data take the form of sequences which naturally represent the und
erlying structural information of branching and axillary flowering patterns
and allow the application of a large number of methods ranging from explor
atory analysis to stochastic modeling. The primary aim of the proposed anal
ysis methods is to reveal patterns not directly apparent in the data, and t
hus to deepen our biological understanding of the underlying mechanisms tha
t control the branching and the axillary flowering of plants over time and
space. The proposed approach is illustrated using a set of examples corresp
onding to different plant species and different biological or agronomic obj
ectives., (C) 2001 Academic Press.