Intensification of rice-based farming systems in Central Luzon, Philippines: Constraints at field, farm and regional levels

Understanding the opportunities for sustainable intensification requires an integrated assessment at field, farm and regional levels of past developments. Two hypotheses regarding current rice production in Central Luzon (Philippines) were developed for this purpose. First, we hypothesize that there are trade-offs between rice yields, labour productivity, gross margin and N use efficiency and, second, that farm(er) characteristics and socio-economic conditions at farm and regional level affect the management practices used by farmers. These hypotheses were tested using two household surveys characterizing rice-based farming systems in Central Luzon in terms of changes over time (1966–2012) and spatial variability. Over the past half-century there was an increase in the proportion of irrigated fields and adoption of improved varieties, which allowed the cultivation of a dry season rice crop in Central Luzon. Moreover, transplanting has been replaced by direct-seeding and herbicides substituted hand-weeding. These resulted in greater rice yields and labour productivity, and contributed to gradual transition from subsistence to commercial farming systems, as observed in the increasing proportion of hired labour and rice sold. Our results indicate the existence of a trade-off between rice yields, labour productivity and N use efficiency as yield levels maximising labour productivity and N use efficiency were ca. 25% and 35% lower than climatic potential yield in the wet and dry season, respectively. At field level, this can be explained by 1) the use of transplanting as crop establishment method, which resulted into higher yields but lower labour productivity as compared to direct-seeding, and 2) the high N application levels, which led to higher yields but lower N use efficiency. In contrast, yield levels which maximised gross margin were ca. 80% of the climatic potential in both wet and dry seasons, so there was little trade-off between rice yields and economic performance. Regarding the second hypothesis results were not always conclusive. As an example, N application per ha was negatively associated with farm size and the timing of the first fertiliser application positively associated with household size and with the number of parcels. More intensive practices, and better farm performance, were recorded in the province at the heart of the irrigation system. We thus conclude that closing rice yield gaps in the production systems of Central Luzon incurs trade-offs with environmental and social objectives at field and farm levels but less with economic objectives. However, we could not clearly show whether, and to what extent, management practices used by farmers are influenced by farm or regional level constraints.