with a row spacing ranging from 4.45 to 6.7 m. The growing season of
the Gansu Poplars was from May to October, whereas those of the
spring wheat and maize were from March to July and April to September,
respectively.
During the growing seasons, the cropland was irrigated by conventional
flood irrigation once every 7 to 14 days with an irrigation amount
of approximately 100 mm per irrigation event. A similar irrigation
amount was applied at the beginning of the winter season after sowing
both crops. The amount of irrigation was intended not only to meet the
crop water requirements but also to leach the soil profile of salts in
order to avoid problems related to increased salinity. As shown in
Fig. 3, 21 irrigation events occurred in the spring wheat subplot during
the studied period (nine events in 2012 and 12 events in 2013), while
24 irrigation events occurred in the maize subplot (12 events in both
2012 and 2013). Daily rainfall and other climatic variables were recorded
by an AG1000 automated weather station at a distance of about
500 m from the study site. During the experimental period, there
were 21 rainfall events in 2012 and 25 rainfall events in 2013, which
produced total rainfall amounts of 84.4 and 73.2 mm, respectively. For
the 46 rainfall events, the mean rainfall depth, storm duration, rainfall
intensity, and interval time between two successive rainfall events
were 4.3 mm, 4.3 h, 0.84 mm/h, and 287.4 h, respectively.
Three parallel transects were laid out along the length of the spring
wheat-shelterbelt-maize plot, with a spacing of 4 m (Fig. 2). Twelve
Trime-Time Domain Reflectometry (TDR) access tubes (4-cm diameter,
polycarbonate) were installed to a depth of approximately 280 cm in
each subplot (Fig. 2). Four access tubes were located along each of the
three transects passing through a given subplot in order to facilitate
the measurement of volumetric SWC. The distance between adjacent