Resumen: 

The Andean region of southern Ecuador with its climatic and geographical heterogeneity has given rise to different production systems: from intensive agriculture in the lowlands to subsistence and agroforestry systems in the highlands. With a marked structural division between intensive agriculture and biodiverse agriculture. Social dynamics such as migration have an impact, aggravating rural depopulation and food insecurity. In this context, Andean agrobiodiversity represents a key strategy to cope with climate change and promote sustainable agricultural systems. The objectives of this study were: i) To analyse the agro-ecological characteristics (slope, elevation, area, floristic diversity and chlorophyll concentration) of different agricultural production systems in the southern Andes of Ecuador. ii) To evaluate the consumption, efficiency and sustainability of energy use in different Andean agricultural production systems. iii)Analyse the link between agricultural energy inputs and vegetation condition through the NDVI index, using Landsat satellite images in Andean production scenarios in Ecuador between 2012 and 2023. iiii) Develop a theoretical model to estimate the carbon stock in biomass and soil, as well as the CO emissions derived from agricultural practices in different Andean production scenarios.

High mountain agroforestry systems (HAFS), short-cycle cropping (SHC) and greenhouse systems (GHSE) were found on sloping land, which makes them vulnerable to erosion and requires the implementation of specialised agricultural practices. The low mountain agroforestry systems (LAFS) are in flat areas, which favours mechanisation and extensive crop management. The values obtained for the Margalef, Simpson and Shannon-Wiener indices confirmed the greater structural and functional heterogeneity of the HAFS. Despite the high species similarity between the systems, abundance was significantly lower in LAFS. Chlorophyll content (Chl), varied significantly between species and systems, with higher values in LAFS, due to both its favourable climate and high fertiliser use. In terms of total energy consumption, the GHSE and SHC systems were the most demanding, with values of 46.38 GJ ha- and 44.40 GJ ha-, respectively. In contrast, HAFS and LAFS reported significantly lower consumptions (11.98 and 10.49 GJ ha-). Fertiliser was the main source of energy consumption (31.8%), followed by labour (30%) and plant management (25.7%). All systems had a high dependence on non-renewable energy (mean 0.62) and low energy autonomy (0.12), with GHSE and SHC being the most dependent and HAFS and LAFS the most sustainable.

NDVI analysis using Landsat images (TOA, SR and C) proved to be effective for our scenarios. Landsat-SR images yielded the highest Max_NDVI values: up to 0.92 in LAFS, 0.80 in SHC and lower and non-significant values in HAFS. A moderate positive correlation (r = 0.36) was observed between energy efficiency and NDVI. C stock estimation showed that both HAFS and LAFS accumulate similar amounts of carbon in aboveground and root biomass (20.34 and 17.51 Mg C ha-, respectively), with no significant statistical differences. However, in soil carbon, SHC stood out with the highest values (16.54 Mg ha-), followed by HAFS (15.37 Mg ha-) and LAFS (13.95 Mg ha-). GHSE recorded the lowest value (9.21 Mg ha-). In terms of CO emissions, the GHSE system had the highest total emissions (17.41 Mg CO ha-), followed by SHC (7.63 Mg CO ha-). In contrast, the HAFS and LAFS systems had the lowest emissions (1.37 and 0.30 Mg CO ha-).