The Efficiency of Trenches As Runoff Water Harvesting Systems and the Role of Their Design in Minimizing Water Losses.

Water is a primary limiting factor for agricultural development in many arid and semi-arid regions in which a large fraction of the annual rainfall is usually the result of a few intensive convective storms. The intense rainfalls frequently induce the formation of thin surface crusts that reduce markedly the infiltration rate, as a result of which only a small fraction of the rain is absorbed by the soil. The fraction that is not absorbed by the soil can be conveyed to dyke surrounded plots and ponded, and thereafter used for tree production. The runoff generating areas may not be necessary located close by the receiving areas, but one of the most promising configurations is that of Microcatchments in which runoff water is collected close to the area in which it was generated and stored in adjacent shallow pits. The main objective of the work present herein was to assess the effect the geometry of the runoff water collection area (shallow pit or trench) has on direct evaporative water losses to the atmosphere and on the water use efficiency of the olive trees grown in them.

The study was carried out using regular micro-catchments with basins of 9 m² (3 x 3 m) by 0.1 m deep and trenches one meter deep and one meter wide. Each configuration was replicated three times. One tree was planted in each shallow basin and the distance between trees in the trench was four meters. Access tubes for neutron probes were installed in micro-catchments and trenches (four and seven, respectively) to depths of 2.5 m. Soil water content in the soil profile was monitored periodically throughout drying out periods in between simulated runoff events. Transpiration of trees was estimated from Sap flow measurements using a Granier system every half hour. Total transpiration fluxes were computed for time intervals that correspond to consecutive soil water measurements.

The study was carried out during two years. During the first year a large runoff event was simulated by applying four cubic meters to each plot and in the second year the same volume was split into four applications, thus simulating a series of small runoff events.

In both treatments trees received the same amount of water per tree. Total water loss and transpiration were monitored during both seasons as described above. Evaporation from trenches and micro-catchments was estimated as the difference between evapotranspiration obtained computing the differences in total by soil water content between two consecutive measurements and transpiration for this interval estimated from sap flow measurements.

In both years the evaporation from micro-catchments was significantly larger than that of trenches. The fractional loss due to evaporation from the total applied water for the second year for example, was 53 and 22% for micro-catchments and trenches, respectively.