Soil fertility matters! A new conceptual model for carbon stewardship in neotropical croplands taking climate-smart agricultural practices into account

Mismanagement of agroecosystems in Neotropical regions threatens global security, accelerating the transgression of planetary boundaries. Therefore, understanding carbon (C) stewardship and how climate-smart agriculture (CSA) practices change nutrient availability plays a central role. Here, we analyzed nutrient availability, nitrogen (N) inputs, climate, and soil texture influence C flow into particulate (POC) and mineral-associated organic carbon (MAOC) pools to support sustainable C management in neotropical agroecosystems. To test our hypothesis data were collected from three field experimental agroecosystem sites and a literature overview. Our machine learning models estimated that nutrient availability, notably zinc (Zn), and soil texture, regulate C flow into POC and MAOC pools in agroecosystems. The climate variables exhibited minimal effects. There was no MAOC C saturation in neotropical agroecosystems, with an upper boundary of 36 g C kg-1. This demonstrates the potential of nature-based solutions for C storage in tropical soils. Synthetic N fertilization was not a key driver of C flow into POC and MAOC pools in these agroecosystems; however, organic N inputs, such as those from legumes, showed significant potential in increasing soil C and reducing carbon-to‑nitrogen ratio. Our main finding reveals soil fertility as a key regulator of C flow into POC and MAOC pools in Neotropical agroecosystems. Additionally, nature-based solutions from CSA are viable for atmospheric carbon removal strategies in Neotropical areas. Thus, by integrating experimental and simulated insights, we propose a new conceptual model linking nutrient availability to C stewardship in neotropical agroecosystems, outlining existing knowledge gaps and suggesting directions for future research toward climate-smart agriculture.

Palavras-chave

Carbon saturation, Intercropped agroecosystems, Machine learning models, Nitrogen, No-tillage farming, Tropical croplands, Agro-ecosystems, Intercropped agroecosystem, No tillage, Nutrient availability, Organic carbon pools, Smart agricultures, Tropical cropland