Site-specific soil fertility management requires a fundamental understanding of factors that modulate soil fertility variability in the local context. To verify this assumption, this study hypothesized that soil fertility variability across two regions in Central and Western Ethiopia is determined by inter-related effects of agro-ecological zones and farmers’ resource endowment (‘wealthy’ versus ‘poor’ farmers). Mid-infrared spectroscopy coupled to partial least squares regression (midDRIFTS-PLSR) and wet-laboratory analyses were used to assess the soil fertility (soil pH, total soil carbon [TC] and nitrogen [TN], plant-available phosphorous [P_av] and potassium [K_av]) across four agro-ecological zones: ‘High-Dega’ (HD), ‘Dega’ (D), ‘Weina-Dega’ (WD) and ‘Kola’ (K). MidDRIFTS peak area analysis of spectral frequencies (2,930 [aliphatic C-H], 1,620 [aromatic C = C], 1,159 [C-O poly-alcoholic and ether groups] cm^-1) was applied to characterize soil organic carbon (SOC) quality and to calculate the SOC stability index (1,620:2,930). Higher TC in HD, as well as higher TN and K_av contents in K were found in fields of wealthy compared with poor farmers. Resource endowment dependent soil fertility management options revealed SOC of higher quality in wealthy compared with poor farms in D. Agro-ecological zones distinctions contributed to these soil fertility differences. Farmers distinguished visually fertile and less fertile fields based on soil colour. Higher pH in K and WD as well as P_av in K and HD were found in fertile (brown/black) than less fertile (red) soils. To conclude, tailor-made soil fertility management in the local context must consider agro-ecological zones and resource endowment interactions along with farmers’ indigenous knowledge.

中文翻译：

---

农业生态，资源end赋和土著知识互动调节埃塞俄比亚中部和西部混合农业系统中的土壤肥力

特定地点的土壤肥力管理要求对当地条件下调节土壤肥力变异性的因素有基本的了解。为了验证这一假设，本研究假设埃塞俄比亚中部和西部两个地区的土壤肥力变异性是由农业生态区和农民资源end赋（“富裕”与“贫困”农民）之间的相互影响决定的。使用中红外光谱结合偏最小二乘回归（midDRIFTS-PLSR）和湿实验室分析来评估土壤肥力（土壤pH，土壤总碳[TC]和氮[TN]，植物有效磷[P _av ]和钾[K _av]）跨越四个农业生态区：“高地加”（HD），“地加”（D），“维纳-地加”（WD）和“可乐”（​​K）。使用MidDRIFTS光谱频率的峰面积分析（2,930 [脂族CH]，1,620 [芳族C = C]，1,159 [CO多元醇和醚基] cm ^-1）表征土壤有机碳（SOC）的质量并计算SOC稳定性指标（1,620：2,930）。HD的TC较高，TN和K _av较高与贫穷的农民相比，在富裕的农田中发现了钾的含量。与资源end赋相关的土壤肥力管理选项表明，富裕地区的SOC比D州的贫困农场要高。农业生态区的差异是造成这些土壤肥力差异的原因。农民根据土壤的颜色区分了视觉肥沃的土地和肥沃的土地。在肥沃（棕色/黑色）土壤中，发现钾肥和WD的pH值较高，钾肥和HD的P _av值要高于肥沃的土壤（红色）。总而言之，根据当地情况进行量身定制的土壤肥力管理必须考虑农业生态区和资源end赋的相互作用以及农民的本土知识。