In a large area of the Loess Plateau of China, surface and groundwater resources are often either unavailable or too saline and brackish for irrigation. Precipitation is the major water source for plant growth. Annual precipitation ranges from 200 to 750 mm, with 70% of rainfall falling between June and September, often in the form of heavy thunderstorms. This not only causes tremendous amounts of erosion, but also problems of water loss (Li et al., 2000a). Revegetation is a pressing issue for eco-environmental improvement in this region; however, the establishment of vegetation is quite difficult due to inadequate availability of moisture.

Water harvesting, based on the collection of precipitation runoff from a prepared catchment surface and the storage in the adjacent crop area, has been used successfully for crop yield improvement and tree establishment in the arid and semiarid regions of the world for thousands of years (Bruins et al., 1986 and Reij et al., 1988). Microcatchment water harvesting (MCWH) which collects runoff from short slopes is especially useful in arid and semi-arid regions (Boers et al., 1986 and National Academy of science, 1974). The major advantages of MCWH are simple, cheap, replicable, efficient and adaptable (Reij et al., 1988). MCWH can be successful in years of normal or above normal rainfall and are best suited for situations in which drought-resistant trees or other drought-hardy perennial species are grown (Brooks et al., 1991). Particularly, the microcatchment procedure can be used in complex terrain or on steep slopes, where other water-harvesting techniques may be difficult to install. The effective use of MCWH systems in growing trees and shrubs was reported as early as in the nineteenth century for growing olive trees in Tunisia (Pacey and Cullis, 1986). In MCWH systems, the runoff area to basin area ratios may range from 1:1 to 20:1 (Ojasvi et al., 1999). The microcatchment size has varied from 0.5 m2 (Aldon and Springfield, 1975) to 1000 m2 (Evenari et al., 1968) for trees, shrubs and row crops; average annual rainfall has ranged from 87.5 mm (Karnieli et al., 1988 A. Karnieli, J. Ben-Asher, A. Dodi, A. Issar and G. Oron, An empirical approach for predicting runoff yield under desert conditions, Agric. Water Manage. 14 (1988), pp. 243–252. Abstract | View Record in Scopus | Cited By in Scopus (6)Karnieli et al., 1988) to 650 mm (Anaya and Tovar, 1975).

Caragana korshinskii is a drought resistant shrub, which is widely used for vegetation rehabilitation in arid and semiarid regions of China, for its high ecological and economic values, including: (i) it plays a key role in vegetation succession from shifting dune to sandy grassland, (ii) helps restore degraded land by fixing atmospheric nitrogen, (iii) can form shrub shelterbelt for crops or artificial grassland, and (iv) can serve as a supplemental livestock forage (Hansson et al., 1995, Ren et al., 2002 and Zheng et al., 2004). However, the establishment and growth of C. korshinskii is often limited by water shortage. This study was intended to investigate the effect of the MCWH technique on plant growth of C. korshinskii taking into account of microcatchment parameters (size, slope length), rainfall, runoff, soil moisture storage and water use efficiency.

A field study was conducted to explore the effectiveness of using microcatchment water harvesting to grow Caragana korshinskii in the semiarid loess region of China between 2002 and 2004. The experiment involved four different size microcatchments (5, 15, 30 and 50 m2) and the control (0 m2) with six replications to supply runoff water for C. korshinskii. For the microcatchments, runoff volume (m3) increased with increasing catchment size, following a positive linear function, whereas runoff depth (mm) decreased with increasing catchment size, following a negative power function. Runoff efficiency was 3.09–13.69%, 1.05–6.25%, 0.39–4.49% and 0.30–3.25% for the 5, 15, 30 and 50 m2 microcatchments, respectively, and varied between years. Microcatchment rainwater harvesting treatments supplied additional runoff water to the infiltration basin occupied by C. korshinskii tree. Total rainfall and runoff water collected in the plant area was significantly higher for the rainwater harvesting treatments than the control, and total runoff to the planted area increased with the size of the microcatchment, thus resulting in significant higher soil water storage and evapotranspiration in the water harvesting treatments. Water harvesting treatments had a pronounced effect on the growth of C. korshinskii. Tree height, crown diameter, collar girth, above-ground biomass and water use efficiency were significantly higher for the water harvesting treatments than the controls. This demonstrates that C. korshinskii can be grown successfully using microcatchments on rainfed lands. From the perspective of soil water storage efficiency and the minimum acceptable slope length, catchment/planted area ratio of 19–38 for the catchment area between 15 and 30 m2 was most suitable to grow C. korshinskii under semiarid condition.