Testing Observers of Plant Compositional Diversity in Forest Monitoring

This thesis deals with forest monitoring and with the assessment of the inter-observer error in vegetation survey. I participated in the life project “LIFE MODERn NEC” which is part of the National Emission Reduction Commitments Directive (NEC) and it’s focused on the monitoring and reduction of air pollutants. The main aims of the project were the enhancement of the ICP FORESTS PROGRAM for the forest ecosystem survey with the development of new indicators and study protocols. Thesis work includes the ‘Quality Assurance Program’ for vegetation monitoring, which contributes to standardise the monitoring protocols for the warranty of data quality and comparability. I was involved within a series of activities: preliminary writing and elaboration of the official field manuals, training and intercalibration course for the surveyors, field check activity, MODERnNEC official monitoring campaign in 2023, and data analysis. The inter-observer error was assessed through the estimation of the data variability and the comparison among the means for each surveyor team for different variables. Furthermore, the variability of the “species richness” variable among five different training and intercalibration courses held in different years was assessed through the percentage coefficient of variation (CV%). The analysis proved that the training of observers was effective in the reduction of the observer error for some variables bringing to a convergence of the surveyor’s estimations and to reduce the misunderstandings of the monitoring procedures after the training and intercalibration session. Furthermore, the analysis of the CV% of the species richness year by year defined a decreasing trend of the observer estimations variability, suggesting that the improvement of the field manual together with the reiteration of training sessions are a winning combination in the reduction of the inter-observer error.