The area in our study is located in Ansai district, Yan’an city, and it is a site of loess accumulation situated deeply in the Loess Plateau in northwest China, where there are loess ridges, hills and gullies [3]. Agricultural development in this area depends mainly on fruit trees. The area is in the temperate zone that has the climate of continental semi-arid monsoon and distinct seasonal changes. In this area, the mean annual temperature is 8.8˚C, the mean annual precipitation is 505.3 mm and there are 157 frost-free days. Our experiment is carried out in Fangta orchard in Ansai district. For the 0 - 40 cm soil layer in this orchard, the volume-weight is 1.21 - 1.66 g.cm⁻³, the soil porosity is 0.39% - 0.64% and the pH value is 6.02 - 7.36. The age of the trees is 18a and the spacing between most tree rows is 4 m × 5 m.

The design of the 3-factor and 3-level orthogonal experiment [4] :

1) The three factors:

A: the degree of irrigation; B: the time of irrigation; C: the amount of fertilization.

2) The three levels:

The degree of irrigation (A). A1: 40%, A2: 60% field capacity, and A3: 80% field capacity, in which, field capacity represents the water-holding capacity in the experimental field.

The time of irrigation (B). B1: 5 times of irrigation in winter (middle November) + pre-flowering period (from late March to early April) + the first fruit expanding stage (late June) + the second fruit expanding stage (late July) + fruit transition (coloring) period (late August); B2: 3 times of irrigation in ? winter + the first fruit expanding stage + the second fruit expanding stage; B3: 2 times of irrigation in ? winter + the second fruit expanding stage.

The amount of fertilization (C): C1: 0.25 kg per tree; C2: 0.5 kg per tree; C3: 1 kg per tree.

Each treatment is repeated 3 times and a control group contains 3 trees that are normally managed. Trees in orchards A, AF and E are treated in the same way. The irrigation water is taken from the rainwater harvesting cistern. The fertilizers are calcium superphosphate (containing 15% P₂O₅) and potassium sulphate (containing 51% K₂O) which have been commonly used by farmers in this area. Mix the two fertilizers together and control the mixture of P₂O₅ and K₂O at the pure mass ratio of 1:2.

The scheme of different treatments in the 3-factor and 3-level orthogonal experiment is shown in Table 1.

A number of outer leaves of each tree in the experiment were collected from all directions after the fruits had been picked (in (middle October) in 2016. For each tree, 7 - 9 leaves were put into a plastic bag and taken back to the lab. Use deionized water to clean the leaves and put them into envelopes after they have been spin-dried. Put the envelops in an environment at the temperature of 105˚C for 15 min to deactivate enzymes, them dry them at 75˚C till the weight keeps constant. After that, use a 100-mesh sieve to screen them. Seal them in a plastic bag.

Detection items include: the contents of total nitrogen, total phosphorus and total potassium in leaves. The leaves are boiled in H₂O₂-H₂SO₄ and detected by using the continuous flow analyzer FLOWSYS.

Use EXCEL and SPSS to process and analyze the data.

Figure 1 is a description of the contents of total nitrogen in leaves under different treatments of irrigation and fertilization. It can be concluded from the analysis of Figure 1 that: there is no big difference between the contents of total nitrogen in leaves in orchard A and AF, but the content of total nitrogen in leaves in orchard E is relatively low. With regard to Table 2, it can be known that 2.15% - 2.30% is a low level for the content of leaf nitrogen while 2.31% - 2.50% is a normal level. For orchard A, the contents of total nitrogen in leave maintain at a normal level under treatment 4, 6 and 7, but are low under other treatments. For orchard AF, the contents of total nitrogen in leaves maintain at a normal

level under treatment 3, 4, 6, 8 and 9, but are low under other treatments. For orchard E, the contents of total nitrogen in leaves are highest under treatment 3, 7 and 8, but in general, the contents are lower than the normal level. Compared with the control group, the contents of total nitrogen in leaves that had been treated with the irrigation and fertilization are generally higher, which proves that the irrigation and fertilization can promote apple trees to adsorb the element of nitrogen.

Total phosphorus can promote the growth of fruits and a proper amount of phosphori is beneficial to the yield, but excess phosphori will generate a negative effect on the synthesis of sugar and, as a result, affect the taste of the fruit. Figure 2 is a description of the contents of total phosphorus in leaves under different treatments of irrigation and fertilization. It can be seen that the contents of total phosphorus in leaves in all the orchards maintain at a normal or higher level, which indicates that the irrigation and fertilization can promote apple trees to adsorb the element of phosphorus. However, the promotion effect is unstable in different fields. A possible reason is the relative small amount of the phosphatic fertilizer designed in the experiment (calcium superphosphate 0.083 kg - 0.33 kg/tree each time), which is not enough to generate a distinct effect on the adsorption of phosphorus.

The element of potassium is a participant and a promoter of the synthesis of both sugar and acid in fruit. Figure 3 is a describe the contents of total potassium in leaves under different treatments of irrigation and fertilization. It can be known that there is a big difference between the contents of total potassium in leaves under different treatments, and the highest content occurs under treatment

6. The contents of total potassium in leaves in orchard A and AF show that: under treatment 7, 8 and 9 in which the degree of irrigation is the highest, the contents of total potassium in leaves are higher than those under other treatments, which reflect that irrigation can promote apple trees to adsorb the element of potassium.