1. Introduction
‘Bingtang’ sweet orange [
C. sinensis (L.) Osbeck], a high-quality variety grown in Hunan Province, China, is a bud mutation of the common sweet orange and is characterized by a low acid content. Yongxing County is the main producing area of ‘Bingtang’ in Hunan. The existing 4330 ha cultivation area in Yongxing is considered the major production area of high-quality sweet oranges in China [
1]. However, nutrient imbalances or deficiencies have been identified in citrus orchards in recent years due to poor soil fertility and long-term unreasonable fertilization [
2,
3]. In addition, excessive use of chemical fertilizers not only causes soil nutrient imbalances and rapid decline in fertility but also increases the loss of soil nutrients and the risk of environmental pollution [
4,
5,
6]. Therefore, it is urgent to improve the efficient utilization of crop nutrients and reduce fertilizer loss in citrus orchards.
In fruit crops, soil analysis alone is not a satisfactory guide for fertilization recommendation. A citrus grower cannot rely on soil analysis alone to formulate a fertilizer program or diagnose a nutritional problem in a grove, mainly because of the difficulty in determining accurately enough the nutrient availability in root zones, where deep-rooted plants take up most nutrients [
7,
8,
9]. The mineral-element content in leaves reflects the nutritional status of the fruit trees. Leaf mineral nutrient analysis has been widely used as an early diagnostic method to avoid irreversible losses of nutrients [
10,
11]. Much research has been carried out in the past to develop and improve leaf analysis for identifying nutritional constraints and, subsequently, the fertilizer recommendation in fruit trees [
12,
13].
The delayed response of fruit trees to fertilizer applications compared with annual crops makes it difficult to determine their nutritional status through soil analysis immediately after application. The plant’s nutritional status is, oftentimes, better reflected by the element concentration in the leaves than the other organs. Foliar analysis shows actual uptake. It is considered to indicate the current nutritional status of fruit trees as an alternative to soil analysis [
14,
15]. The critical values or optimum nutrient concentrations in foliar analysis play an important role in citrus cultivation. In particular, earlier analysis in the growing season could allow for sufficient time to correct any deficiencies before harvest. In earlier studies, Cao et al. [
1] used the partial least squares regression analysis method to determine the precise relationship between soil nutrients and ‘Bingtang’ sweet orange quality, and then, the recommended values of soil nutrient content for the optimal fruit quality were obtained. In addition, Li et al. [
16], Guo et al. [
17], and Zhang et al. [
18] applied the canonical correlation analysis to screen for the main soil nutrient factors affecting the fruit quality index of grape, kiwifruit, and apple, respectively, and established the regression equation between them. These studies found that different tree species have different absorption and utilization abilities for mineral nutrients, and the effects of mineral nutrient elements on the quality indexes of different fruits also varied. Thus, it is important to study the relationship between the mineral-element content in leaves and fruit quality. However, the critical values of the mineral composition of sweet orange leaves have not been published in detail. For scientific fertilizer recommendation, the correct diagnosis of plant nutrient deficiency is important and should be an integrative approach to crop production [
19,
20,
21,
22].
Herein, in this work, the ‘Bingtang’ sweet orange was used as the testing material in order to monitor the status of leaf-nutrient abundance and deficiency in citrus orchards and to identify the leaf-nutrient limiting factors affecting the improvement of fruit quality. Through the determination of related parameters based on these samples, multivariate statistical analysis methods were used to quantitatively formulate the leaf nutritional content requirement scheme of high-quality fruit. The findings are expected to improve the understanding of citrus-orchard nutrition in Hunan Province and provide important implications for future orchard fertilization management.
2. Materials and Methods
2.1. Experimental Sites
In this study, 120 representative citrus orchards were involved in Yongxing (25°58′–26°29′ N, 112°43′–113°35′ E) during the 2019–2022 growing season, which is in the hilly region of Southern Hunan, China. The orchard has a subtropical continental humid monsoon climate, characterized by an average total annual precipitation of 1370.6 mm and mean annual sunshine hours of 1625.2 h, as well as an average annual temperature of 17.6 °C (
Table 1). The dominant soil type is red soil, which is primarily derived from Quaternary red clay, and it has a silty loam texture and is characterized as an Ultisol based on the USDA soil taxonomy [
23]. All studied citrus orchards were managed according to the ordinary practices used by growers in the region. Chemical fertilizers were mainly applied during citrus cultivation. In mid-July of every year, complex fertilizer (N:P
2O
5:K
2O = 15:15:15, m/m/m, 750 kg/ha) was spread along vertical edges beneath the tree canopy. The basic physicochemical properties of citrus-orchard soil are described in
Table S1. The geographical locations of the sampling sites are shown in
Figure 1.
2.2. Plant Materials
The cultivar of citrus studied was ‘Bingtang’ sweet orange grafted onto trifoliate orange [Poncirus trifoliata (L.) Raf.], with an average plant age of about 15–20 years old. The cultivation density of citrus (broad row 350 ± 15 cm; narrow row 250 ± 15 cm; the distance between two plants 300 ± 15 cm) ranged from 950 to 1350 plants per hectare. Leaf and fruit samples were collected during the harvest season in late November to early December. Overall, a total of 120 leaf samples and 120 fruit samples were obtained from 2019 to 2022. To reduce experimental errors, the two trees at the beginning and the end of each row were avoided. Leaves and fruits were randomly sampled from five similar trees at each citrus orchard (600 trees in total). Each composite leaf sample consisted of about 100 leaves taken from the north, east, south, and west directions and the same height as the tree’s periphery. Leaves were collected from non-fruiting spring vegetative shoots (nine months old) of the same variety and rootstock. After being washed with running tap water and three times with deionized water, the leaves were oven-dried at 70 °C until constant weight was achieved and then ground into a fine powder using a centrifugal mill (0.425 mm sieve). During the fruit sampling period, the ripening period was about the same in the various orchards. Fruit samples (one composite sample per orchard) were collected from the same plants chosen for collecting leaf samples and each sample consisted of at least 10 fruits. The harvested fruits were immediately sent to the laboratory for fruit quality parameter analysis.
2.3. Leaf Minerals and Fruit Quality Parameter Analysis
Analysis of the leaf mineral elements was performed according to the method described by Bao [
24]. Leaf N, P, and K concentrations were measured after the samples were digested with H
2SO
4-H
2O
2. The leaf N concentration was determined by the Kjeldahl method, whereas leaf P and K concentrations were assayed by vanadomolybdate and flame spectrophotometry, respectively. After digestion with HNO
3-HClO
4, the leaf Ca, Mg, Fe, Mn, Cu, and Zn concentrations were determined by atomic absorption spectrophotometry, and the leaf S concentration was measured using X-ray fluorescence spectrometry (Zetium, Malvern Panalytical, Almelo, The Netherlands). The leaf B concentration was extracted with 0.1 M HCl and determined using the curcumin colorimetric method. The contents of N, P, K, Ca, and Mg were expressed as %, and the contents of S, Fe, Mn, Cu, Zn, and B were expressed as mg/kg. The nutrient concentrations of the leaves were expressed as dry weights (DW). Three technical replicates were analyzed for each sample.
Single fruit weight (SFW) was determined using an electronic balance with an accuracy of 0.01 (Yongzhou Weighing Apparatus, Jinhua, China) and expressed in grams. The length and diameter of the fruits were assessed using a digital Vernier caliper (Harbin Measuring & Cutting Tool Group Co., Ltd., Harbin, China, precision ± 0.02 mm), and the length/diameter ratio was used as the fruit shape index (FSI). Total soluble solids content (TSS, °Brix) was determined using a digital refractometer (model: Pocket PAL-1, Atago Inc., Tokyo, Japan) and expressed as a percent (%). Titratable acid (TA) was measured using NaOH (0.10 N) until a pH of 8.20, using an automatic potentiometric titrator (TitroLine 6000, SI Analytics GmbH, Mainz, Germany), and was expressed as % (
w/
v) citric acid. The concentration of vitamin C (Vc) was determined by 2,6-dichlorindophenol sodium titration and expressed in mg/100 g [
25]. The maturation index (MI) was calculated using the TSS/TA ratio. All analyses were performed in triplicate.
2.4. Statistical Analysis
The results were the means ± standard deviation (SD) of three replicates. The leaf-nutrient index (U) consists of the following total nutrient contents: N (x1), P (x2), K (x3), Ca (x4), Mg (x5), S (x6), Fe (x7), Mn (x8), Cu (x9), Zn (x10), and B (x11). The fruit quality index (V) consists of SFW (y1), FSI (y2), TSS (y3), TA (y4), Vc (y5), and MI (y6). Statistical analysis was performed using SPSS (version 22.0; SPSS, Inc., Chicago, IL, USA) and R (version 3.6; University of Auckland, New Zealand) for the Pearson test, canonical correlation analysis (CCA), partial least-squares regression (PLSR), and model establishment of the relationship between fruit quality attributes and leaf factors. LINGO software (version 11.0; Lindo System Inc., Washington, DC, USA) was used to calculate the optimization scheme for the leaf nutrients.
4. Discussion
In fruit trees, leaves are the most sensitive vegetative organ to mineral elements, which can reflect nutrient accumulation and redistribution throughout the growing plant. Leaf analysis integrates all the factors that might influence nutrient availability and uptake. Dominguez et al. [
27] concluded that the most adequate moment for nutritional diagnosis would correspond to the period in which nutrient concentrations remain highly stable for some time in the tissue analyzed, allowing for a reliable comparison with their references. Fortunately, previous research provides a guide [
1,
9,
16]. Samples were taken during the fruit maturation stage, when ‘Bingtang’ sweet orange quality tends to be stable, and the leaf nutrients do not fluctuate much. Based on the field investigation, most of the ‘Bingtang’ sweet orange orchards appeared to have nutrient deficiency symptoms. Through the determination and analysis of the mineral nutrient content in the leaves of spring vegetative shoots, the status of nutrient deficiency in orchard leaves was clarified, and the nutrient-limiting factors in the tree body were also identified [
7,
28]. In this study, most of the average nutrient concentrations of the leaves from the 120 citrus orchards differed significantly among the three grades. According to the leaf-nutrient grade standards established by Lu [
26], more than 65% of ‘Bingtang’ sweet orange orchards were found to have excess levels of total P, possibly because orchard farmers usually have overused complex fertilizers (N-P-K = 15-15-15) for many years. The phenomenon of Ca and Mg deficiency was common in the acidic soil of Southern China, which may be related to strong soil weathering and leaching in subtropical climates. When the soil pH decreased, the positive charge in the soil increased. And the adsorption of Ca
2+ and Mg
2+ decreased significantly, further affecting the nutrient uptake of fruit-tree roots [
29,
30]. The concentrations of Mn, Cu, and Zn in the leaves had a high-level spatial variation and may be related to the extensive use of Bordeaux mixture and mancozeb in the control of citrus diseases and insect pests in some orchards. In addition, little attention has been paid to Zn fertilization by local farmers in orchard production. According to our investigation, ignoring the foliar application of B fertilizer in citrus orchards was also crucial for B deficiency in trees. In addition, we found that the coefficients of variation (CVs) for SFW, FSI, and TSS were less than 12%, meaning that fruit quality from different orchards was relatively identical [
9,
31]. In contrast, the CVs of the TA showed a high level of variation, possibly because of high soil acidity and nutrient imbalances (
Table S1). In fact, soil acidification in citrus orchards was widespread in the main citrus-producing areas of Southern China [
12,
32,
33].
Recently, many researchers have used the mineral nutrition status of leaves to study the relationship between different elements and the quality and yield of various fruits [
2,
12,
34]. We could use these research results to establish criteria for diagnosing leaf nutrition constraints for high-yield and high-quality citrus fertilization and, thus, further improve the level of citrus cultivation. Previous studies have indicated that the fruit quality index is affected by many leaf-nutrient factors [
9]. As well, there are not only synergistic and antagonistic effects among different elements but also supplementary and substitution relationships between leaf-nutrient elements and fruit quality factors [
35]. Thus, it is necessary to combine fruit quality diagnosis with leaf-nutrient diagnosis to accurately identify the production issue in the fertilization management of orchards. In this study, a correlation analysis of 11 leaf-nutrient indexes and 6 fruit quality parameters in the ‘Bingtang’ sweet orange orchards was carried out. Simple bivariate correlation analysis results showed that the correlation between the two indicator groups is not strong. When calculating the canonical correlation coefficients, the canonical variable was 0.814. At the same time, we observed that leaf N, K, S, Fe, Mn, Cu, Zn, and B displayed positive relations with SFW, FSI, TSS, and Vc but were negatively related to the TA content in fruit quality. Obviously, there was a big difference between the leaf-nutrition factors that affect fruit quality factors screened by canonical correlation analysis and the leaf-nutrition factors selected according to the correlation coefficient of a single factor, which indicates that it is not comprehensive enough to use simple correlation analysis in the study of the relationship between leaf nutrition and fruit quality. There may be multicollinearity among leaf nutrients. Since canonical correlation analysis can better address the problem of multicollinearity, this method was chosen for critical leaf-nutrient factors that affect fruit quality [
1,
16,
17].
Macronutrients and micronutrients are indispensable nutrient elements for fruit growth and quality improvement [
36,
37]. Notably, the lack or excess of any mineral element may cause an imbalance in fruit nutrients and physiological diseases affecting the growth of citrus trees [
38,
39,
40]. Fertilization imbalances cause waste, pollute the environment, and cause toxicity to fruit trees, resulting in a decline in fruit quality [
3,
4]. The knowledge of the desirable level for each nutrient in the leaf allows one to define its excess or deficiency in the plant and design a correct fertilization program. In recent studies, fertilizer recommendations for sweet oranges are based on calibrated leaf-nutrient concentrations and fruit quality parameter tests [
41]. Li et al. [
16] screened for major soil-nutrient deficiencies affecting berry quality and defined the corresponding soil-nutrient indicator content. Gui et al. [
11] investigated the floral nutrients analysis in Satsuma mandarin and obtained the critical concentrations of K, Ca, Cu, Fe, Mn, and B in flowers, which are 2.08–2.10%, 0.53–0.56%, 13.01–13.80 mg/kg, 65.80–69.90 mg/kg, 16.55 mg/kg, and 30.64–33.20 mg/kg, respectively. In the present study, the optimum leaf mineral nutrient content on fruit quality was determined by partial least-squares regression and linear programming in ‘Bingtang’ sweet orange during fruit maturation. The results showed that the optimum concentrations of N, P, K, Ca, Mg, S, Fe, Mn, Cu, Zn, and B in leaves are 2.41–4.92%, 0.10–0.28%, 1.30–2.11%, 2.99%, 0.26–0.41%, 340–640 mg/kg, 89.65–127.46 mg/kg, 13.48–51.93 mg/kg, 2.60–13.84 mg/kg, 15.59–51.48 mg/kg, and 53.95 mg/kg, respectively. The leaf-nutrient optimization results (
Table 6) calculated by the theory were compared with the measured value of leaf nutrients in the orchard investigated at present (
Table 2), and the problem of nutrient management in the orchard was analyzed. It is noteworthy that, when the quality of the citrus fruit is optimal, the leaf Ca and B contents all took the maximum value, indicating that increasing the content of these nutrient indicators could directly affect the improvement of fruit quality. Interestingly, Hu et al. [
7] established the diagnosis standards and norms of citrus-leaf macronutrient and micronutrient analysis. There are some variations in standards. The macronutrient (N, P, K, and Ca) concentration range was relatively wider than the recommended standards or norms, whereas the micronutrients (Fe, Mn, Cu, Zn, and B) were relatively narrower, possibly because of differences in soil fertility and fruit-tree nutritional properties. Therefore, further studies of field verification and adjustment should be carried out based on the nutritional management level of orchards in different areas.
5. Conclusions
We monitored the ‘Bingtang’ sweet orange fruit tree vegetative growth and fruit quality parameters during fruit maturation. In general, the contents of N, Ca, Mg, B, and Zn in the leaves were especially lacking. The CVs of TA, Vc, and MI in the fruits were particularly elevated. The results of the present study demonstrate that fruit quality attributes could be influenced by leaf mineral nutrients. In terms of the Pearson correlation coefficient and canonical correlation coefficients, it was found that SFW, FSI, TSS, Vc, and MI were positively correlated with most leaf-nutrient indicators, while TA was inversely correlated. Multivariate statistical analysis results showed that, when the leaf-nutrient indicators were N, 2.41–4.92%; P, 0.10–0.28%; K, 1.30–2.11%; Ca, 2.99%; Mg, 0.26–0.41%; S, 340–640 mg/kg; Fe, 89.65–127.46 mg/kg; Mn, 13.48–51.93 mg/kg; Cu, 2.60–13.84 mg/kg; Zn, 15.59–51.48 mg/kg; B, 53.95 mg/kg, the ‘Bingtang’ sweet orange quality indicators could reach the optimal values of SFW, 206.12 g; FSI, 1.01; TSS, 15.45%; TA, 0.40%; Vc, 92.76 mg/100 g; MI, 37.81.