Monitoraggio e valutazione del danno degli insetti defogliatori su diverse specie di Quercus in Sicilia

Most of the insects feeding on oaks are usually ‘in balance’ with their host trees, causing occasional damage. When severe defoliations regularly occur, oak trees vulnerability to the complex of factors causing the overall decline syndrome can increase. Assessing canopy damage due to defoliating insects through the usually adopted visual methods is quite difficult, requiring highly experienced operators and a long time spent to collect field data. The present research has been carried out to investigate the different susceptibility of Sicilian oak species to defoliating insect attacks, in particular the oak moth Tortrix viridana (L.) (Lepidoptera Tortricidae), with the aim to assess the relationship among infestation, crown transparency and the health vegetation status (oak decline). The study was carried out in two of the main oak woodlands of western Sicily in the years 2015-2017: “Bosco di Ficuzza” (Palermo province), involving oak species of the Quercus section (Quercus leptobalanos Guss. and Quercus virgiliana (Ten.) Ten.) and of the Cerris section (Quercus gussonei (Borzì) Brullo and Quercus trojana subsp. trojana Webb) and “Monte Cammarata” (Agrigento province), involving Q. virgiliana. The research was developed following some different main topics. In the two study areas 15 sampling plots were considered (12 at Ficuzza and 3 at Mt. Cammarata). In each plot 20-30 oak trees were randomly selected and used for samplings of egg masses and damaged leaves and for transparency and health status recording. The first research topic was focused on monitoring the population of T. viridana through recording the egg masses number on the four Quercus species and correlating them to the characteristics of the oak population. The sampling plan of T. viridana egg masses in the study areas involved branches collected from different oak species and their examination in the laboratory. The two Sicilian species belonging to the Quercus section resulted more susceptible to T. viridana infestation, with a mean number of egg masses significantly higher than the other two species. Moreover, the effect of habitat structure in terms of coexistence of different oak species resulted significant too, as both Q. leptobalanos and Q. virgiliana were less infested when Q. gussonei was present in the same plot. Furthermore, different visual methods to record in the field the defoliation damage have been considered: crown transparency and vegetation state of the studied oak woods. In the first case, the transparency was recorded using a standardized method, whereas health status of examined trees was assessed recording crown transparency, cancer and injury occurrence in the tree trunk. The two survey methods allowed calculating a transparency index and a damage index, respectively. The two indexes showed to be significantly correlated between them and linked to oviposition. Moreover, the leaf damage was investigated, through the collection of leaf samples, which were subsequently examined in the laboratory. Leaves were scanned and digital analysis of leaf images was carried out to measure the total leaf surface and the portion of the surface damaged by the trophic activity of the insect. No differences among oak species and sampling plots were found in the number and surface of damaged leaves and in the distribution of leaves in size classes, probably due to the low infestation levels recorded in the study years. To explore the potential of the medium resolution data to evaluate the crown damage caused by defoliating insects and to classify the infested areas, a spectral vegetation index, namely Normalized Difference Vegetation Index (NDVI), was applied using the Landsat 8 satellite images. In general, NDVI values showed low differences among the years, probably due to the low infestation level, to the potential interference of lower part of trees crown and of brushwood vegetation or to an insufficient sensibility of the used satellite images in detecting low differences in the vegetation conditions. The trend of the NDVI values in the three years resulted consistent with oviposition, transparency and damage indexes in most of sampling plot. Finally, the T. viridana population density was compared with the damage / transparency indexes obtained with different methods, in order to identify simplified and efficient procedures to investigate the effects of infestations on the oaks populations. Results show that mapping crown transparency using satellite imagery analysis may represent a useful method to evaluate crown transparency, but standardized visual monitoring still remains essential to validate the results obtained by remote sensing, especially in heterogeneous woods like the Mediterranean deciduous oak ones.