Microlevel

Methods applied to determine microlevel nutrient balances differ considerably from those applied at the macrolevel and mesolevel as they are generally based on on-site inventories and monitoring studies that provide firsthand data. However, they all utilize regression models and other proxies to estimate difficult flows, such as leaching and gaseous losses. A recently developed research tool known as NUTMON (www.nutmon.org) allows co-calculation of the effects of INM and combines the calculation of nutrient flows and balances with farm economic performance under different nutrient management scenarios (Vlaming et al., 2001). The methodology differentiates between ‘easy to measure’ and ‘difficult to measure’ nutrient flows. The ‘easy to measure’ flows, IN1, IN2, OUT1 and OUT2, can be quantified from farm survey data and expressed in monetary and labour units (Figure 1). These flows form the partial nutrient balance, as often determined in farm-level studies. The ‘difficult to measure’ flows, IN3, IN4, IN5, OUT3, OUT4 and OUT5, are normally estimated with transfer functions. The transfer functions are simple relations that explain ‘difficult to quantify’ variables as a function of easily obtainable parameters. These ‘difficult to measure’ nutrient flows are difficult or impossible to express in monetary terms (Van den Bosch et al., 1998a). The EC-funded VARINUTS project in Kenya helped to build the above toolbox, as data was collected and nutrient flows were calculated according to the NUTMON procedures. Fifteen farms were monitored on a monthly basis for a two-year period. Further information about the NUTMON approach is available in: De Jager et al. (1998), Van den Bosch et al. (1998a) and Van den Bosch et al. (1998b).

One study (Kanté, 2001) compared two villages in southern Mali with similar farming systems but different land pressures as a result of higher population density and a higher ratio between cultivated land and total land. This study relied partly on CMDT farm-survey data. The nutrient-flows calculations were as per Stoorvogel and Smaling (1990), with slightly modified regression models for OUT3 and OUT4. The ‘difficult flows’ were estimated, while the nutrient flows of the partial balance were measured. They were quantified on the basis of a general diagram of how farmers perceive nutrient flows to occur (Figure 13). Partial balances (IN1, IN2, OUT1 and OUT2) were calculated per crop, per farm, per wealth class and per village. Finally, the complete balance with all inflows and outflows was calculated. Nutrient losses resulting from deep latrines (OUT6) were introduced at the microlevel, because the whole farm was considered part of the system.