Modelling scenarios of soil properties and managements in olive groves at the micro-catchment scale with the AnnAGNPS model to quantify organic carbon

• Published in: Catena (Giessen) 2021-08, Vol.203, p.105333, Article 105333
• Main Authors: Taguas, E.V., Bingner, R.L., Momm, H.G., Wells, R., Locke, M.A.
• Format: Article
• Language: English
• Subjects: AnnAGNPS; Olive groves; Organic carbon; Sensitivity analysis; Soil losses
• ISSN: 0341-8162; 1872-6887
• DOI: 10.1016/j.catena.2021.105333

[Display omitted] •108 scenarios were built with AnnAGNPS to model SOC in an olive grove catchment.•The main variables for SOC modeling were identified with a sensitivity analysis.•Sediment OC had a maximum with the combination NT-Clo whereas SOC pool with C-SC-F3.•CN, soil and fertilization type were the most significant features for calibration.•Annual OC exported and SOC pool simulations were also compared with field studies. Soil organic carbon (SOC) stock changes are crucial in identifying risk of desertification in fragile areas In this work, 108 scenarios were prepared with the model AnnAGNPS in a small catchment of extensive olive groves by considering 6 different soil types (with textures sandy, S; sandy loam, Slo; loam, L; clay loam, Clo; silty loam clay, SiLoC; clay, C), 3 different managements (no till, NT; conventional tillage, CT, and cover crop, SC), 3 types of fertilization (two organic at rates of 40 and 80 kg.ha−1, F2 and F3, and another inorganic F1 with 100 kg.ha−1), and 2 contrasting reach organic carbon half-life time (0.1 day–730 days). The consistency of simulated annual OC attached to sediment and variations in ground SOC (h = 200 mm) were evaluated and compared in the context of the region of Andalusia, Spain. In addition, the most significant parameters to consider in the subsequent calibration process were also identified. There were significant differences in annual exported OC for the scenarios of soil and management, ranging between 0.0 kg.ha−1 and 368.9 kg.ha−1. In addition, S and SC showed the lowest variability intervals, while Clo and NT had the highest OC values and variation ranges. For the SOC pools, the effects of soil and fertilization were more evident than that of management. The combination C-SC-F3 presented the maximum increase of SOC (0.150 mg OC.g-1soil.y-1) as a result of the maximum OC inputs while Slo-NT-F1 presented the minimum (0.080 mg OC.g-1soil.y-1). Finally, the most significant parameters derived from the sensitivity analysis were soil texture, saturated hydraulic conductivity and Curve Numbers for annual exported OC and ground OC, fertilization parameters and rate of decomposition of residues. Model parameterization based on a sensitivity analysis and the abundance of previous experimental studies supports the AnnAGNPS applications for simulating the OC loss in agricultural catchments.