Is the spatial correlation between the tree basal area and its increment a good indicator for competition?

Is a negative spatial correlation between basal area and its increment indicative of tree competition, in a forest stand considered as a realisation of a point process? For a bivariate point process, the spatial correlation results from two components: the value of the marks and tree positions, and a negative spatial correlation can occur without any correlation between the marks knowing their type. In a forest stand in French Guiana, a negative spatial correlation is observed. Moreover, the big trees exhibit a regular pattern, independently from the small trees that are clustered. This configuration is sufficient to explain the negative correlation without involving competition. Competition however remains as a process that explains the shift from clustering to regularity with an increasing tree size.