The experiment was proceed in Long’an, a township of Guang’an City during 2015-2016. The average elevation is 320 m, the annual average temperature is 17.58˚C. The average temperature is 4˚C in January, it is the coldest in a year. The average temperature is 34˚C in July, it is the hottest in a year. Frost-free has 306 - 328 d, the average annual rainfall of 1240 mm, the average annual sunshine hours of 1213 h. The testing area which is a typical southwest hilly region of high temperature and rainy climate, and soil fertility is low.

The experiments were conducted in a 5-year-old plantation, with a 5.0 m × 5.3 m planting spacing. In the experiment, 5 fertilizer schemes were set up. The treatments were as follows: treatment 1 (T1: nitrogen 315 kg, phosphorus 173.25 kg, potassium 306 kg and organic matter 1575 kg were used per hectare), treatment 2 (T2: nitrogen 450 kg, phosphorus 247.5 kg, potassium 438 kg and organic matter 2250 kg were used in per hectare), treatment 3 (T3: all applied organic fertilizer, 3000 kg in per hectare), treatment 4 (T4: total soluble fertilizer was reduced with fertilization times increasing, nitrogen 202.8 kg, phosphorus 252.3 kg and potassium 268.95 kg were used in per hectare), and control (CK: the trees that abandoned by farmers in the production). The applying fertilizer method: open hole (0.4 m × 0.4 m). The fertilization time is divided into: Basal fertilizer (November 10-15, 2015), Sprout fertilizer (February 25-30, 2016), and strong fruit fertilizer (May 25-30, 2016). One tree as a plot, each treatment has 15 trees.

Data sorting with Excel, SPSS13.0 software was used for analysis of variance and correlation analysis.

The effects of different fertilization treatments on the nutrient contents in different soil layers were inconsistent. Among them, the trend of nutrient content of 0 - 20 cm soil layers were as follows: T2 > T3 > T4 > T1 > CK. The treatments of 20 - 40 cm soil layers were not significant, but T2, T4 and T3 were higher than those of T1 and CK. The different soil layers in same treatment showed: 0 - 20 cm > 20 - 40 cm soil layer (Table 1). But the contents of alkaline N and available Fe were the highest in 0 - 20 cm soil layer of T3, the contents were: 138.00 mg・kg⁻¹, 0.77 g・kg⁻¹, 0.82 mg・kg⁻¹, respectively. And the organic matter content of T3 (17.2 g・kg⁻¹) was significantly higher than that of T4, T1, CK and 20 - 40 cm soil layer of T2. Compared with the control, the organic matter, total P, alkaline N, available P, available K, available B, available Cu, available Zn and available Mn were significantly improved in 0 - 20 cm soil layer. Then T2 and T3 were beneficial to promote soil fertility levels.

The different fertilization treatments had significant differences in the leaves' nutrient content of “Longanyou” (Table 2). Compared with the control, the

Note: Values within a column followed by the different lowercase letter are significantly different at 0.05 level.

Note: Values within a column followed by the different lowercase letter are significantly different at 0.05 level. The same as followed.

content of K and B increased significantly with the increase of organic manure application, but had little effect on the content of P in leaf. The Zn content in the vegetative branches and bearing branches of T3 were higher. The content of P (1.60 g・kg⁻¹), B (26.00 mg・kg⁻¹) and Mg (1.18 mg・kg⁻¹) were the highest, and N, K, Zn and Fe content were relatively higher than others. While the nutrient contents of control leaves’ were the lowest. The nutrient contents of different branches were different, the trend was vegetative branches higher than bearing branches, but the Fe content was the opposite. Compared comprehensively, the nutrient contents of leaves in T2 were higher than others.

According to Figure 1(a), the yield per plant of T2 was the highest, 95.40 kg plant⁻¹, followed by T4, and the lowest was control, only 14.13 kg plant⁻¹. The fruit number per tree of T3 was the highest, followed by T2, and the lowest was the control (Figure 1(b)).

According to the comparison of the appearance quality of the fruit (Table 3), There were significant difference in the average fruit weight among the treatments. The fruit weight of the upper layer was the largest (1736.70 g) in T2, and it was 2.01 and 2.58 times of the upper and lower in control, respectively. Then followed by T4 upper layer (1662.2 g). Moreover, the fruit peel of T2 and T4 were thicker than T1, T3 and CK, which indicating that T2 and T4 can improve the ability of anti-fungal invasion and enhance the storability, combining with the fruit weight, we suggested that T2 and T4 were benefit for increasing the production and commodity value of Longanyou.

It was observed in Table 4 that the internal quality including the converting sugar, the total sugar, the ratio of sugar to acid, and the content of Vc were the highest in the upper layer fruits of the T2, followed by the upper layer fruits of T3. Titratable acid of the upper layer fruits were lower in T2 and T3, it was 0.66% and 067% respectively. Soluble solids (12.70%), solid acid ratio (19.01),

sucrose (5.11%) were the highest in the upper fruits of T3, followed by the upper layer fruits of T2. However, Titratable acid were the highest and the contents of the soluble solids, TSS/acid ratio, reducing sugar, sugar, sucrose, total sugar, sugar/acid ratio and Vc content were the lowest in control. Therefore, it was explained that T2 and T3 were beneficial to improve the internal quality of fruit.

From Table 5, the contents of B, Zn and Fe in leaves were significantly correlated with soil nutrient. However, the contents of N, P, K, Mg were no significant correlations with soil nutrient contents. There was a significant negative correlation between B content in leaf and pH value in soil, and positively correlated with available B, organic matter, total N, total P, available N, available P, available K, available Cu, available Zn, available Fe in soil. The Mg content in leaves was positively correlated with available N and available Fe, and negatively correlated with other indexes, especially with available Cu, available Zn and pH. While the Fe content of leaves was negatively correlated with pH value and exchangeable Ca. It was positively correlated with other indexes and had a significant positive correlation with total K, available K, available P, available B, available Cu and available Zn. It shows that leaf nutrient and soil nutrient were closely related.

According to Table 6, the pH value was negatively correlated with the fruit yield

Note: Correlation is significant at the 0.01 level. Correlation is significant at the 0.05 level, The same as followed.

and quality index, but only the correlation between pH value and yield per plant reached significant level. There was a significant positive correlation between total K and fruit weight, and the other indexes did not reach significant level. The contents of organic matter, total N, total P, available N, available P, available K, available B, available Cu, available Zn, available Fe, available Mn and exchangeable Mg were negative correlations with the juice yield, edible rate and titratable acid. There was a significant positive correlation with fruit weight, fruit thickness, total sugar, solid acid ratio, sugar and acid ratio, Vc content and fruit yield. There was a significant positive correlation between available B content, fruit weight, fruit thickness, total sugar, sugar and acid ratio and Vc content, and negatively correlated with juice yield, edible rate and titratable acid. The effective Fe and Zn content was negatively correlated with the juice yield, edible rate and titratable acid, which was significantly correlated with other fruit yield and quality indexes. Therefore, soil nutrients have played a very important role in improving fruit quality.

According to Table 7, the correlation between K, B, Zn, Fe and fruit yield and quality index was significant, while the N, P, Mg were not significant correlations with the yield and quality of fruit. There was a significant positive correlation between the K content, fruit weight, fruit thickness, fruit number, fruit yield, total sugar, the ratio of sugar to acid, and the content of Vc, and negatively correlated with juice yield and edible rate. Negative correlations were shown between the B content and juice rate, edible rate and titratable acidity. The content of B was a significant positive correlation with the yield and quality of other indexes. There was a significant negative correlation between Fe content and juice yield and edible rate, and negatively correlated with titratable acid. There was a significant positive correlation with single fruit weight, single yield, fruit thickness, total sugar, and the solid acid ratio was significantly positive correlation. The content of Zn was significant negatively with the juice rate, and was negatively correlated with the edible rate, and was positively correlated with the soluble solids, fruit shape index, fruit number per plant and yield per plant. It was found that the leaf nutrient contents were negatively correlated with juice rate and edible rate.

Different fertilization treatments had a significant effect on soil nutrient content,

and different fertilization treatments had significant difference, but the evaluation results of fertilization treatments were different in the existing reports. Some researchers suggested that organic fertilizer alone significantly promote the absorption of calcium, and promote the transformation of nitrate nitrogen, adjust the soil nitrate content changes [18]. It was also suggested that organic manure could increase soil organic matter, total N, available P, available K, exchangeable Ca, exchangeable Mg, pH value and soil fertility [19]. As well as the researchers suggested that the orchard increased organic fertilizer can significantly improve the nutrient contents of leaf [20] [21]. More studies have found that the rational application of chemical fertilizers and organic fertilizers can significantly improve the soil fertility and fruit yield [22] [23].

In this experiment, the results showed that the nutrient contents in soil increased with the increase of organic manure application. The nutrient contents in T2 and T3 were significantly increased, and the soil nutrient contents of 0 - 20 cm depth in T2 were the highest, which indicated that the increase of organic fertilizers was beneficial to improve soil available nutrient, and organic fertilizers with chemical fertilizers were better (T2). Furthermore, the nutrient contents of vegetative leaves in T2 were also the highest. It indicated that most of the nutrients in the tree came from the soil, and the leaf nutrient was closely related to the soil nutrient [24]. In addition, the results showed that the Mg content in the leaves was positively correlated with the available N and available Fe, and negatively correlated with other parameters, which indicated that the Mg deficiency in leaves was not only caused by the corresponding elements in the soil, and applying single fertilizer could not meet the tree's requirement for nutrient [25]. The results also showed that there was a positive correlation between Fe, B content in leaves and soil nutrient content, it indicated that soil nutrients directly affect the Fe and B nutrient content in leaves, and fertilization could significantly improve to the nutrient level in leaves. There were many factors that affect the relationship between soil and leaf nutrient, but there has not yet reached a uniform conclusion about the correlation in the existing reports. Thus the specific relationship and the reasons need to be further studied.

The nutrient of leaf and soil had significant effect on fruit yield and quality, especially the index of leaves’ nutrient content can respond to the nutritional requirements more accurately for optimum fruit yield [4]. Zang et al. suggested that the application of organic fertilizer can increase the soluble solids, Vc content and solid acid ratio, reduce the content of titratable acid [26]. Wang et al. suggested that the application of different ratio of compound fertilizer is enable to improve the yield and quality of citrus [27]. Marathe reported that the application of FYM (to supply 50% N) + 50% RDF (1000 g N + 400 g P₂O₅ + 400 g K₂O tree⁻¹ year⁻¹) was optimum for sustaining higher fruit yield and quality of sweet orange [28]. In this study, the fruit and yield were the highest, the fruit peel was thicker, and the content of sugar, total sugar, Vc and rate of sugar and acid were the highest in T2, it indicated that T2 can improve the ability of anti-fungal invasion and enhance the storability [29].

The relationship between soil or leaf nutrient content and fruit yield and quality was complex. Zhang et al. found that available P in soil had the greatest effects on fruit quality, followed by K, N, and P in leaf could significantly increase the content of Vc and sugar, but reduce acidity, while Mg and B increase the acidity of fruit [30]. In addition, researchers have suggested that the content of soluble solids in fruit was negatively correlated with soil pH value, and there was a significant positive correlation with soil organic matter, available N, Fe, Mn and Zn. The content of Vc in fruit was positively correlated with soil pH and available K, and negatively correlated with soil organic matter and other nutrient contents [4]. It has also been suggested that most of the nutrients in citrus leaves were negatively correlated with the soluble solids of fruit quality, and positively correlated with Vc content [31]. In this study, we found that the content of available K, B, Fe and Zn in soil increased the content of mineral elements in leaves, which increased fruit peel thickness, fruit weight, total sugar and Vc content etc. The similar results have been reported by Aula et al. [32], Khayyat et al. [33], Patel et al. [34] and Asadi et al. [35]. Meanwhile, B, Fe, Zn content and titratable acid were significantly negatively correlated, it indicated that it could increase sugar and reduce acid content with increasing B, Fe, Zn content, Furthermore, the content of B and Zn was positively correlated with the soluble solid, and the content of available B and Zn was helpful to increase the soluble solids content of fruit. Therefore, foliage spraying K, B, Fe, Zn appropriately is the effective way for improving fruit yield and quality in the orchard fertilization practice.