Rural organic wastes are spatially dispersed across agricultural production areas and settlements, creating challenges for collection, treatment, nutrient recovery, and return to agricultural land. This study developed a township-scale spatial planning framework integrating waste quantification, centralized and decentralized composting, nutrient recovery, and techno-economic assessment in Beiyang Township, eastern China. Agricultural residues were allocated to five centralized composting reactors, whereas household kitchen waste was treated using decentralized bins. Machine-learning models estimated total carbon (TC) and total nitrogen (TN) retention under three composting-adoption scenarios. The defined study area was estimated to generate approximately 8745 t yr⁻¹ of rural organic waste, 82.8% of which originated from agricultural production. Under the high-adoption scenario, with 70% of agricultural waste and 75% of kitchen waste composted, retained TN was equivalent to approximately 27.9% of annual local crop N fertilizer demand on a total-N basis. Economic performance improved with increasing waste participation. Under the baseline scenario, four centralized reactors and the decentralized system were financially viable, whereas one reactor was non-viable. Under the moderate- and high-adoption scenarios, all configurations achieved positive NPVs and IRRs above the 10% benchmark rate. The framework supports coordination of rural organic-waste treatment, nutrient recycling, and sustainable agricultural land management at the township scale.

To evaluate the nutritional value of fermented cottonseed meal (FCSM) for growing pigs, ten representative FCSM samples were collected and assessed in two experiments for the standardized ileal digestibility (SID) of amino acids (AAs) and for digestible energy (DE) and metabolizable energy (ME). In Experiment 1, 11 growing pigs (Duroc × Landrace × Yorkshire, 25 ± 1.8 kg) fitted with a T-cannula were assigned to an 11 × 6 incomplete Latin square design, with 11 diets (one nitrogen-free diet and 10 diets containing FCSM as the sole nitrogen source) and 6 periods. All diets contained 0.30% Cr₂O₃ as an indigestible marker. Experiment 2 used 22 healthy crossbred barrows with an initial body weight of 45 ± 2.3 kg. A replicated 11 × 3 incomplete Latin square design was adopted, comprising 11 dietary treatments and 3 periods. The dietary treatments included a corn–soybean meal basal diet (control) and 10 FCSM-containing diets. The results demonstrated marked variation among FCSM samples in terms of chemical composition and energy content. The mean Ca/P ratio was 0.32. The average DE and ME were 11.20 MJ/kg (range 9.69–12.65 MJ/kg) and 10.68 MJ/kg (range 8.55–11.84 MJ/kg), respectively. The SID values for Lys, Met, Thr, Trp and Val were 65.90% (60.01 to 68.92%), 77.34% (74.48 to 81.78%), 68.10% (59.97 to 72.02%), 72.12% (70.25 to 73.62%), and 75.50% (68.60 to 78.65%), respectively. Stepwise regression analysis generated prediction equations for estimating DE and ME from chemical composition. These data fill a gap in energy and amino acid digestibility parameters for FCSM, provide a basis for precision diet formulation, and indicate the need to supplement limiting amino acids and optimize the Ca/P ratio.