Geostatistics, spatial dependences in a tree: a new approach in fruit tree studies
1990
Monestiez, P. | Habib, R. | Audergon, J.M.
Different models can be used to describe spatial correlations in trees. In this paper, the spatial variations of quantitative mesurements are considered. The tree is considered as a spatial set of points (i.e. branches and shoots) on which a random function can be defined (i.e. measurements made on fruits or leaves). We have define this as "tree space". The theory of "regionalized variables" and the geostatistic methods, firstly developed in mining prospect, are now classical methods in automatic cartography or spatial data analysis. The aim of the present paper is to adapt this theoretical background to the structure of a tree and to study the particularities that it induces. The second step is to apply the geostatistical methods to an actual fruit tree. The results are in agreement with expectations and the different hypotheses seem to be acceptable. Spatial dependences are observed on fruit characteristics and on leaves content analysis. These first results allow us to extend the concepts of "spatial structure analysis" and "cartography" to a tree, with useful consequences in sampling techniques and total amount estimations.
显示更多 [+] 显示较少 [-]