Analysis of Potential Supply of Ecosystem Services in Forest Remnants through Neural Networks

Abstract

Analyzing the landscape configuration factors where they are located can ensure a more accurate spatial assessment of the supply of ecosystem services. It can also show if the benefits promoted by ecosystems depend not only on the supply of these services but also on the demand, the cultural values, and the interest of the society where they are located. The present study aims to demonstrate the provision potential of regulating ecosystem services by forest remnants in the municipality of Campinas/SP, Brazil, from the analysis and weighting of geospatial indicators, considering the assumptions of supply of and demand for these ecosystem services. The potential supply of regulating ecosystem services was evaluated through the application of an artificial neural network using landscape indicators previously surveyed for the 2319 forest remnants identified in six watersheds. The findings show that the classified remnants have a medium to very high regulating potential for the provision of ecosystem services. The use of artificial intelligence fundamentals, based on artificial neural networks, proved to be quite effective, as it enables combined analysis of various indicators, analysis of spatial patterns, and the prediction of results, which could be informative guides for environmental planning and management in urban spaces.