Perennial Crops Can Compensate for Low Soil Carbon Inputs from Maize in Ley-Arable Systems

(1) Background: Soil organic carbon (SOC) in agricultural soils plays a crucial role in mitigating global climate change but also, and maybe more importantly, in soil fertility and thus food security. Therefore, the influence of contrasting cropping systems on SOC not only in the topsoil, but also in the subsoil, needs to be understood. (2) Methods: In this study, we analyzed SOC content and δ(13)C values from a crop rotation experiment for biogas production, established in southern Germany in 2004. We compared two crop rotations, differing in their proportions of maize (0 vs. 50%) and perennial legume–grass leys as main crops (75 vs. 25%). Maize was cultivated with an undersown white clover. Both rotations had an unfertilized variant and a variant that was fertilized with biogas digestate according to the nutrient demand of crops. Sixteen years after the experiment was established, the effects of crop rotation, fertilization, and soil depth on SOC were analyzed. Furthermore, we defined a simple carbon balance model to estimate the dynamics of δ(13)C in soil. Simulations were compared to topsoil data (0–30 cm) from 2009, 2017, and 2020, and to subsoil data (30–60 cm) from 2020. (3) Results: Crop rotation and soil depth had significant effects, but fertilization had no effect on SOC content and δ(13)C. SOC significantly differed between the two crop rotations regarding δ(13)C in both depths but not regarding content. Annual enrichment in C(4) (maize) carbon was 290, 34, 353, and 70 kg C ha(−1) per maize year in the topsoil and subsoil of the unfertilized and fertilized treatments, respectively. These amounts corresponded to carbon turnover rates of 0.8, 0.3, 0.9, and 0.5% per maize year. Despite there being 50% maize in the rotation, maize carbon only accounted for 20% of the observed carbon sequestration in the topsoil. Even with pre-defined parameter values, the simple carbon model reproduced observed δ(13)C well. The optimization of model parameters decreased the carbon use efficiency of digestate carbon in the soil, as well as the response of belowground carbon allocation to increased aboveground productivity of maize. (4) Conclusions: Two main findings resulted from this combination of measurement and modelling: (i) the retention of digestate carbon in soil was low and its effect on δ(13)C was negligible, and (ii) soil carbon inputs from maize only responded slightly to increased above-ground productivity. We conclude that SOC stocks in silage maize rotations can be preserved or enhanced if leys with perennial crops are included that compensate for the comparably low maize carbon inputs.