3.4. Yield, evapotranspiration (ET) and water use
efficiency (WUE)
Data showed a significance test by ANOVA analysis for Yield, ET
and WUE (Table 3). During the three experimental years, watermelon
yield was lowest for conventional ridge planting (treatment
7), which was significantly lower than that for flat gravel–sand
mulched field and RFRHS with plastic covered ridge. Yield was highest
for the RFRHS with 1:1 ridge furrow ratio, gravel–sand mulched
furrow and plastic mulch on the entire surface (treatment 2). This
indicated that gravel–sand mulch, assisted with plastic film, certainly
resulted in yield improvement. However, this effect varied
with different ratio of ridge:furrow. The ratio of 1:1 was the best,
followed by 5:3 ratio and 4:3 ratio in decreasing order. In addition,
the yield for RFRHS with plastic mulched ridge was significantly
higher than that without plastic mulch (treatment 2, or 3, or 4, or 6
vs. 5). The yield effect of RFRHS with plastic mulch was greater than
conventional ridge planting (treatment 6 vs. 7), but it was smaller
than RFRHS with gravel–sand mulched furrow and plastic mulched
ridge (treatment 6 vs. 2, or 3, or 4).
Rainfall varied during the three trial years and resulted in different
evapotranspiration (ET) amounts (Table 3). Compared with
each pair of same treatments between 2007 and 2009, reductions
in ET ranged from 37 mm to 107 mm, with a minimum ET value in
all treatments in 2009 due to lower rainfall. Among the treatments,
ET was the lowest for the conventional ridge planting in all three
years, and it was second lowest for RFRHS with plastic mulched
furrow and ridge. The three-year average ET decreased as the ratio
of ridge:furrow increased. In the three years, ET was highest for flat
gravel–sand mulched field plus irrigation, especially in 2007 and
2008, when it was significantly higher than all other treatments.