The National Institute for Agricultural and Food Research and Technology (INIA) has carried out this study within the framework of the SustForest Plus project on the modelling of resin production per tree in the study area, using dasometric and environmental variables, obtaining models with a mean square error between 25 and 28% referred to the mean.

Of the covariates used in the cokriging, the resin obtained by drilling followed by the number of faces were the two variables that explained most of the variance of the resin variable mt.

From the environmental variables analysed, orientation was the terrain variable with the greatest statistical significance. The results show a maximum of production on east-facing slopes and another on west-facing slopes, although the latter is less clear. The statistical significance is weaker when the 1998 and 1999 data are included and the East-West effect is not clear.

Factors influencing resin production

The use of Landsat bands as auxiliary variables provides information on what factors may be influencing resin production. A significant correlation has been found with the red and green bands, positive and negative respectively, which seems to indicate higher production per tree at lower densities. The inclusion of these auxiliary variables is not improving the model in terms of prediction, but it is providing useful information.

Prediction of mt resin by cokriging using resin obtained by drilling in 2019 and 2020 as a covariate. Orientation auxiliary variable. The last three images correspond to the variance of the prediction in the years 2010, 2015 and 2020.