The semi-arid zone of West Africa is characterized by an agro-pastoral land management, where different ethnic groups are engaged in agricultural and pastoral activities and interact with each other. To achieve harmonious co-existence of different ethnic groups and sustainable ecosystem management, we should link “field realities” with their scientific explanation. Particularly in the Sahel, where sandy soils prevail and commercial mineral fertilizer is not always expected to be applied, organic matter management would play an significant role in soil conservation and crop production. In this study, we characterize inter-relations among different ethnic groups in the Fakara region, southwestern Niger in the context of organic matter management, set up the experiments to explain the organic matter dynamics under the prevailing management through monitoring CO₂ emission, soil moisture and temperature, and finally propose the optimal organic management for the multi-stakeholders. We found that most households of pastoralists (Fulani people) corralled their livestock at their rented fields, and not at the fields of cultivators (Zarma people), suggesting that the cultivators' fields could unlikely expect to maintain their fertility by the livestock manure under the current corralling practice. The close relationship between the cultivators and the pastoralists through corralling we had assumed might be transformed by the field renting practice. Field experiments confirmed that the corralling with the cow dung of 6 Mg DM ha^-1, or the actual application rate in the study area, could significantly increase grain, ear, and total biomass of pearl millet. From the monitoring data at the experimental plots, sharp decrease of soil temperature immediately after rainfall events was observed through the rainy season especially at the corralling plot and the fallow plot where the much soil coverage with plants prevented soil surface from direct exposure to sunlight. The most striking difference in soil moisture was found between the cultivated plots and the fallow plot. The fallow plot could hold the water at most 50 mm down to the depth of 85 cm, much less than the cultivated plots, due to the much water uptake in the preceding dry season at the fallow plot. Plants at the fallow plot would absorb water until soil water diminished, whereas plants were harvested at the cultivated plots before soil water depleted. Such cooler and drier conditions under fallow could contribute to build-up soil organic matter by restricting the microbial decomposition. The corralling treatment increased soil moisture down to the depth of 30 cm, confirming that input of organic materials such as cow dung and crop residues attracted termites that could facilitate water infiltration. This role of organic materials is highly crucial for efficient water use in the study area where the amount of rainfall is limited and fluctuating. CO₂ emission measured at the corralling and the fallow plots was significantly correlated with soil surface moisture but not with soil temperature. Apparently unusual finding of insignificant correlation of CO₂ emission with soil temperature reflected the decrease of soil temperature upon rainfall. When soil temperature was high, low soil water content might inhibit microbial activity. On the other hand, CO₂ emission at the control plot showed no significant correlation with soil moisture or temperature. Thus, the soils without application of livestock manure or input of plant materials like the fallow practice would not increase CO₂ emission even under the moist condition, and thus could release little nutrients. The relationship between CO₂ emission and soil moisture enabled to estimate the seasonal variation in CO₂ emission with the monitored soil moisture and the total amount of carbon mineralized in any given period. This estimate implied that about half of applied cow dung was decomposed in the first rainy season. In conclusion, the corralling could improve crop production through rapid decomposition of manure and better infiltration mainly at the fields rented by the pastoralists. Thus, the possible way of fertility maintenance for the cultivators would be the more efficient use of fallow where build-up of soil organic matter could be expected. In addition, when we considered the risk for pastoralists that field renting can be cancelled by the landlord cultivators at any time, the establishment of new social relationship between the two stakeholders that the cultivators request pastoralists to corral at the fields of cultivators at the expense of security of field renting would deserve further examination.