Intercropping with legumes alleviates soil N limitation but aggravates P limitation in a degraded agroecosystem as shown by ecoenzymatic stoichiometry

Agroforestry is a commonly applied restoration practice in nutrient-poor or degraded soils, and knowledge of soil nutrient limitations is critical for its successful implementation. Ecoenzymatic stoichiometric theory (EEST) models have been proposed as easy and fast methods to estimate soil nutrient limitations in terrestrial ecosystems; however, a comprehensive application of different models in agricultural studies is still lacking. To bridge these gaps, in this study, we analysed the nutrient limitation of three plantation types, Camellia oleifera monoculture (CK), C. oleifera–Arachis hypogaea (peanut) intercropping (CP), and C. oleifera–Senna tora intercropping (CS), within a fully controlled C. oleifera agroforestry system in nutrient-poor purple soils. We compared the estimates of the nutritional limitations of each plantation type using three different EEST models: vector, vector-threshold element ratio (V-T), and threshold. Estimates of nutrient limitation calculated using the threshold model recapitulated those calculated using the other models. Further, N and P limitations existed in all three plantation types, C limitation was not detected in any plantation type, and the CP plantation type alleviated N limitation and aggravated P limitation. These results were consistent with the actual nutrient characteristics of the purple soil, indicating that the microbial metabolic limitation profile estimated by the EEST model was reliable. The results of this study suggest that EEST models are valuable tools for estimating soil nutrient limitation in the context of agricultural ecosystem restoration, such as in agroforestry.