Low-Cost Monitoring of Spatiotemporal Changes in Forest Cover: Implications for Conservation in a Case in Quindío (Colombia)

Globally, changes in forest cover and fragmentation are a conservation concern. Updated information on their state is required, but it is often costly and difficult to obtain. This study evaluated the feasibility of using low-cost technologies, such as freely available remote sensing data and software, to monitor spatiotemporal changes in forest cover and their relationship to landscape dynamics in terms of structure, shape, and configuration. The latter, using a tropical ecosystem in Quindío (Colombia), between 2002 and 2019, as an example. This is particularly relevant given the complexities of land cover classification in tropical ecosystems and the presence of red howler monkeys (Alouatta seniculus), whose populations are threatened by changes in landscape dynamics. Satellite imagery (Landsat 7, Landsat 8, and Sentinel-2) and a 30-m Digital Elevation Model were processed using analysis algorithms: CART, Random Forest, Naive Bayes, and Support Vector Machine using six spectral indices (NDVI;EVI;SAVI;CIgreen;NDBI;NDWI). The classifications with the highest Kappa index for the Fragstats analysis were CART for 2002 under spectral bands with topographic variables (C4), and Random Forest for 2014 and 2019 under spectral indices with topographic variables (C5). The combination of spectral indices and topographic variables (C5) with Random Forest achieved the highest performance, with Kappa values up to 0.882 and detection accuracy above 95%. The latter, indicating superior capability for detecting land cover changes. Landscape metrics were also calculated using six indicators (NP,LPI, AREA_MN,SHAPE_MN,FRAC_MN,ENN_MN). Between 2002 and 2019, forest cover decreased by 4.05%, from 77.42% to 73.37%, reflecting quantifiable landscape transformation during the study period. The results demonstrate that open-source remote sensing data and software are useful and reliable to study land cover dynamics, while reducing processing costs, time, and resource demands. This cost-efficiency is particularly relevant in the Global South, where technological and budgetary constraints often limit effective monitoring of landscape change and informed conservation decision-making.