Classifications based on functionality improve estimates of monitoring in forests of Tierra del Fuego, Argentina

MARTÍNEZ PASTUR, GJ; RODRÍGUEZ SOUILLA, J; LENCINAS, MV; POLITI, N; RIVERA, LO; SILVEIRA, EMO; RADELOFF, VC; PIDGEON, AM; PERI, PL

Conferencia; INTERNATIONAL SCIENTIFIC CONFERENCE Forest science for people and societal challenges. The 90th "Marin Drăcea" INCDS Anniversary; 2023

Forest classifications are mainly based on tree species, and are not effective to characterize ecological processes related to management and conservation. One alternative is using functional forest types (phenoclusters, [1]), which were also used to determine forest resilience [2]. Besides, monitoring design needs novel approaches that consider the long-term effects of climate change [3]. The objective was to classify and map the forests of Tierra del Fuego (Argentina) according to their phenology characteristics (EVI time series from Sentinel2 and Landsat8) and regional climate (LST Band 10 TIRS Landsat8 and BIO12 from Wordclim). 27 categories were determined, which were grouped using cluster analysis into 6 types. These groups were characterized through ecosystem services (ES, provisioning, regulating, supporting, cultural) and biodiversity (proxy: understory species), and tested with ground-truth data (forest structure, soils) according to different forest types using uni- and multivariate analyses. Results showed significant differences for all the variables, which were not detected through the traditional forest-type classification. Multivariate analyzes showed gradients among functional forest groups in relation to the landscape. Differences were also found also for different ES, including phenoclusters of the different forest types. The results showed the advantage of including these complementary classifications to differentiate natural forest types according to their functionality. Besides, functional groups detected for each forest type highlight their potential uses or their intrinsic characteristics, allowing the development of new management and conservation strategies, e.g. specific stocking rate, timber yield, or conservation values. Finally, these phenocluters also allowed differentiating forests with differential responses to management (e.g. natural regeneration) or climate change resilience.