Accepted Manuscript

Interactions between N, P and K fertilizers affect the environment and the yield and
quality of satsumas

Zhiguo Li, Runhua Zhang, Shujie Xia, Li Wang, Chuang Liu, Runqin Zhang, Zhanhui
Fan, Fang Chen, Yi Liu

PII: S2351-9894(19)30162-3
DOI: [Link]
Article Number: e00663
Reference: GECCO 663

To appear in: Global Ecology and Conservation

Received Date: 22 March 2019

Revised Date: 11 May 2019
Accepted Date: 11 May 2019

Please cite this article as: Li, Z., Zhang, R., Xia, S., Wang, L., Liu, C., Zhang, R., Fan, Z., Chen, F.,
Liu, Y., Interactions between N, P and K fertilizers affect the environment and the yield and quality of
satsumas, Global Ecology and Conservation (2019), doi: [Link]

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

1 Interactions between N, P and K Fertilizers Affect the Environment and the

2 Yield and Quality of Satsumas

4 Zhiguo Li1, Runhua Zhang2, Shujie Xia1, Li Wang1, Chuang Liu1, Runqin Zhang1,

PT
5 Zhanhui Fan3, Fang Chen1, Yi Liu1*

RI
6

7 1. Key Laboratory of Aquatic Botany and Watershed Ecology, Wuhan Botanical Garden, Chinese

SC
8 Academic of Sciences, Wuhan 430074, China;

U
9 2. Wuhan Academy of Agriculture Science and Technology, Vegetable Research Institute, Wuhan
AN
10 430345, China

11 3. Chengdu Academy of Agriculture and Forestry Sciences, Chengdu, 611130, China

M

12
D

13
TE

14 * Corresponding author: Yi Liu

15 Phone: +86 27 87700864

EP

16 Email: liuyi@[Link]
C

17
AC

1
 ACCEPTED MANUSCRIPT

18 Abstract：
：The balance between crop yield, quality and soil nutrient status can be

19 manipulated by exploiting nutrient interactions. Field experiments were conducted in

20 citrus orchards in Danjiangkou, Hubei Province to assess the effects of N (0, 0.3, 0.6

21 and 0.9 kg plant-1), P (0, 0.1, 0.2, 0.3 kg plant-1), K (0, 0.15, 0.30, 0.45 kg plant-1) and

PT
22 their interactions on residual nutrient content in soils, and the yield and quality of

RI
23 satsumas. The results showed that the effects of N and K fertilizers on the yield and

24 quality of satsumas were greater than those of P fertilizers. Appropriate application rates

SC
25 of N (0.6 kg plant-1), P (0.3 kg plant-1) and K (0.15 kg plant-1) significantly improved

U
26 yield and quality. Lack or excess or inappropriate ratios of N, P and K fertilizer can
AN
27 affect the absorption and utilization of nutrients and reduce yield and quality. The best

28 fertilizer combination, determined by comprehensively considering yield, quality and

M

29 environment, was N0.3P0.2K0.3, which delivered moderate yield, high fruit quality and
D

30 low potential risk of pollution. The amount of fertilizer in this case was 0.3 kg N, 0.2 kg
TE

31 P, 0.3 kg K per plant, which is lower than the current recommended application rates

32 (0.5-0.8 kg N, 0.2-0.5 kg P, 0.25-0.64 kg K, respectively). The economic benefit did not

EP

33 decrease because of a decline of yield, but increased slightly because of the improved
C

34 fruit quality and the reduction in fertilizer input, compared with the treatment with the
AC

35 highest yield. Therefore, we suggest that it is necessary to consider the quality of

36 agricultural products and environmental protection when applying fertilizer and to

37 reduce N and P fertilizer input and appropriately increase the proportion of K.

38 Key words: Interaction of NPK fertilizer; citrus; balanced fertilization; optimum

39 fertilization rate; yield; quality

2
 ACCEPTED MANUSCRIPT

40 1. Introduction

41 Balanced fertilization ensures application of the appropriate amount and

42 proportions of nitrogen (N), phosphorus (P), potassium (K) and trace elements based on

43 crop requirements and soil fertility performance. Many countries in the world have been

PT
44 actively promoting and applying this approach. China has carried out experimental

RI
45 research and set up demonstrations of effective application of K fertilizer and trace

46 element fertilizer in some areas. Examples include the combination of P and zinc

SC
47 applied to wheat seedlings (Yang et al., 2009), the combination of boron and calcium

U
48 applied to rape (Du, 2003), and the combination of N and K applied to potato (Xian et
AN
49 al., 1993). After decades of experiments and promotion, both the theory and practice of

50 appropriate fertilizer application have been well-established in China. However, due to

M

51 the great differences in climate, soil fertility and soil texture, as well as the cultivation
D

52 of new crop varieties and climate change in China, it is difficult for farmers to be certain
TE

53 about amounts and proportions, and inappropriate use of fertilizer still happens. In

54 addition, the aim of balanced fertilization in China is to achieve high yield in order to
EP

55 supply the large population and the high demand for agricultural products; therefore, the
C

56 optimum amount of fertilizer was generally determined on the basis of maximizing

AC

57 yield or economic benefits (Hui et al., 2015). Crop quality, agricultural sustainability

58 and environment issues associated with fertilization have not been fully considered (Zhu

59 and Chen, 2002). Many studies have shown that the fertilizer treatment producing the

60 highest yields tends to produce lower crop quality and have greater nutrient losses.

61 Nitrate leaching increases significantly when the nitrogen application rate exceeds

3
 ACCEPTED MANUSCRIPT

62 150-180 kg hm-2 (Zhu and Chen, 2002). Excessive application of N fertilizer could lead

63 to a significant increase in nitrate nitrogen content in groundwater (Akbariyeh et al.,

64 2018), and N2O production increases significantly when the amount of nitrogen applied

65 exceeds 180 kg hm-2 (Nan et al., 2016). In recent years, with the improvement in

PT
66 China's economy and living standards and the enhancement of environmental protection,

RI
67 people not only want to consider the yield and economic benefits delivered by fertilizer

68 application, but also pay more attention to its relationship with crop quality and

SC
69 environment. Thus, the determination of appropriate fertilizer application rates for each

U
70 region and each crop should take account of the relationship between crop yield, quality
AN
71 and environment, and should not trade quality and environment for high yield.

72 In studies of balanced fertilization, potassium plays a more important role than

M

73 other elements in improving crop quality and reducing environmental pollution caused
D

74 by loss of N and P fertilizer. Many studies have shown that, when there is sufficient K,
TE

75 K ions significantly promote the absorption and utilization of N and P by crops in the

76 form of compensation charges (Barnes et al., 1976), improve the utilization rate of N
EP

77 and P fertilizers (Reid et al., 2016) and thus reduce pollution. The application of K
C

78 fertilizer also promotes carbohydrate and N metabolism and thus improves the quality
AC

79 of crop products, for example by increasing the protein content of grain crops (Gaj et al.,

80 2013), crude fat and palmitic acid content of oil crops (Gao et al., 2010), starch and

81 sugar content of potatoes and sugar crops (Westermann et al., 1994), and the fiber

82 length, strength and fineness of fiber crops and cotton (Pervez et al., 2004). Despite the

83 above advantages of K fertilizer, the proportion of N, P and K nutrients applied is still

4
 ACCEPTED MANUSCRIPT

84 often inappropriate due to the habits of farmers and a lack of guidance about K fertilizer.

85 The average ratio of N, P and K fertilizer used in China is 1: 0.43: 0.17, which is

86 significantly lower than the world average of 1: 0.47: 0.37 and 1: 0.57: 0.55 in

87 developed countries (Gu et al., 2014).

PT
88 Satsuma (Citrus unshiu Marc) is a subtropical evergreen fruit tree and is the largest

RI
89 fruit planted in southern China. It has played an important role in China’s economic

90 development. In 2010, the citrus cultivation area in China reached 2 211 000 hm2 and

SC
91 the output increased to 26 645 million tons, making the country the world's largest citrus

U
92 producer (Kong et al., 2018). However, with the significant increase in citrus yield, the
AN
93 quality of fruit is declining, and the problem of citrus sales is becoming increasingly

94 serious. Many studies have shown that fertilization is an important factor affecting the
M

95 yield and quality of citrus. Proper application of N, P and K fertilizers can significantly
D

96 increase the yield and increase the soluble solids content of fruit by 10 - 20 %, sugar by
TE

97 15 - 30 % and Vc (Vitamin C) content by 13 - 57 % (Quaggio et al., 2002; Han et al.,

98 2008; Obreza et al., 2008). However, excessive or insufficient N, P, and K fertilizer will
EP

99 significantly reduce the yield and quality. For example, Alva, Paramasivam et al. (2006b)
C

100 reported that excessive N fertilizer will lead to overgrowth of branches and leaves of
AC

101 citrus and thus an obvious reduction of flowering and fruit. Obreza et al. (2008)

102 reported that the plants receiving excessive P fertilizer have more wrinkled fruits and a

103 less sweet taste, which significantly reduces the fruits’ ability to survive storage and

104 transportation. Kawada (1984) also reported that lack of K affects citrus fruit

105 enlargement, with wide gaps forming between fruit epidermis cells, which increases the

5
 ACCEPTED MANUSCRIPT

106 likelihood of dehydration and rot. Despite many studies being carried out on citrus

107 nutrition at home and abroad, climate, site condition, variety and age of trees still

108 greatly influence the amount and ratios of N, P, K applied to citrus, so inappropriate

109 levels of fertilizer are still being applied to citrus plants. In addition, current research

PT
110 focuses on the effect of single fertilizers, and the optimal fertilizer amount is determined

RI
111 on the basis of yield response to each fertilizer or the nutrient removal by fruit; there has

112 been little research into the interaction between N, P and K and how this affects the

SC
113 yield and quality of citrus.

U
114 Therefore, we examined the effects of N, P, K and their interaction on yield, quality
AN
115 and soil nutrients in citrus orchards in Danjiangkou, China. The purpose of this study is:

116 (1) to reveal the effects of N, P and K fertilizers and their interactions on citrus yield,
M

117 fruit quality and soil nutrients; (2) to determine the optimum amount and proportion of
D

118 N, P and K fertilizer by comprehensive evaluation of the effect of fertilizer on yield,

TE

119 fruit quality and environment; (3) to compare the differences between the optimal

120 fertilization rates based on yield response and on a comprehensive evaluation of yield,
EP

121 quality and environment.

C

122 2. Material and methods

AC

123 2.1 Site description

124 The experiment was conducted in a citrus orchard located in the Danjiangkou

125 Reservoir area in Hubei Province, central China (45.9' N, 110 9.9' E). This region is the

126 largest citrus production base in China and the water source for the Middle Route of the

127 South-to-North Water Transfer Project. Water quality and the ecology and environment

6
 ACCEPTED MANUSCRIPT

128 around the reservoir area are important. However, in order to pursue economic benefits,

129 farmers have vigorously developed cultivation, and more and more resources have been

130 invested in agricultural production. Combined with traditional extensive cultivation,

131 inappropriate fertilizer application and the farming terrain dominated by hills and

PT
132 mountains has resulted in serious loss of soil, water and nutrients in the reservoir area,

RI
133 and the water quality in the area has declined (Liu et al., 2012). Many soil and

134 ecological scientists believe that fertilization in agriculture is the most urgent issue to be

SC
135 solved, because inappropriate fertilizer application rates and proportions, on the one

U
136 hand, and reduced quality and yield of fruit, on the other, increases the risk of nutrient
AN
137 loss and environmental pollution (Liu et al., 2014).

138 The Danjiangkou Reservoir area experiences a typical subtropical monsoon for the
M

139 north subtropical zone, with an average annual temperature of 15.9 ℃ and an average
D

140 annual precipitation of 833 mm. The soil type is mainly yellow cinnamon soils as
TE

141 described by the Chinese Soil Classification System, with heavy texture and poor

142 permeability. The soil layer is thin, generally less than 20 cm. The citrus orchard soils
EP

143 contain 12.22 g kg-1 of organic matter, 56 mg kg-1 of alkaline N, 36.8 mg kg-1 of
C

144 available P, 213.6 mg kg-1 of available K, and have a pH of 6.53.

AC

145 2.2 Experimental design

146 Fertilizer studies were initiated at the citrus demonstration farm of Danjiangkou in

147 central China, beginning in October 2015. Fourteen N-P-K combinations were applied

148 to investigate the simple effects of N, P, and K applied alone as well as their interactive

149 effects (Fig. 1). The treatments are listed in Table 1 and included four levels of N (0, 0.3,

7
 ACCEPTED MANUSCRIPT

150 0.6 and 0.9 kg plant-1), four levels of P (0, 0.1, 0.2 and 0.3 kg plant-1) and four levels of

151 K (0, 0.15, 0.30 and 0.45 kg plant-1). An incompletely randomized block design with

152 three replicates was used. The plot areas were 5×18 m. Each plot had a row of six

153 satsuma plants, with a row spacing of 3 m. Trees on either side were not sampled to

PT
154 prevent the impact of other fertilizer plots. The satsuma variety in the orchard was

RI
155 Citrus unshiu Marc. cv. Weizhang, aged over 20 years. Urea (N 46%) was used as the

156 sole N source, superphosphate (P2O5 12%) was used as the sole source of P, and

SC
157 potassium sulfate (K2O 50%) was used as the sole source of K. The fertilizer was

U
158 applied in furrows and then covered with soil. The fertilization ditch was 40-50 cm
AN
159 away from the main stem and 30 cm depth. Fertilizers were applied three times each

160 year, about 35% of the total N was applied during the fall period in November, 45% in
M

161 early spring in March, and 20% in mid-July. P fertilizer was applied once in the fall
D

162 period in November, while about 40% of the total K fertilizer was applied in early
TE

163 spring in March and 60% in mid-July. The monthly fertilizer application schedule is

164 shown in Table 2. The plots were not irrigated. Citrus trees were pruned in late winter,
EP

165 in early-February.
C

166 2.3 Sampling and Measurement

AC

167 Before the start of the fertilization experiment, a mixed soil sample was collected

168 at a depth of 0-20 cm from randomly chosen positions in the orchard test area to provide

169 baseline soil data. After harvesting, soil samples from the 0-20 cm soil layer, with a

170 composite of eight subsamples, were taken using a 2 cm diameter stainless steel sample

171 auger in each plot. Each soil sample was air-dried in the laboratory then ground to pass

8
 ACCEPTED MANUSCRIPT

172 through a 0.25 mm sieve and analyzed for soil organic matter, pH, available N, available

173 P, and available K content. The soil properties were determined using routine analytical

174 methods (Rukun, 1999) : soil organic matter content was measured using the

175 dichromate-sulfuric acid (K2CrO7-H2SO4) oxidation method, soil pH was measured in a

PT
176 soil-to-water suspension ratio of 1:1, available N (AN) was determined by the Alkali N

RI
177 proliferation method, available P (AP) and available K (AK) were determined by the

178 molybdate blue and flame photometry methods, respectively.

SC
179 Samples of leaves and fruits for nutrient analysis were collected from four

U
180 directions (east, south, west and north) around trees in each plot after harvesting. The
AN
181 fifth to seventh leaves from top to bottom of the new branches of trees were collected.

182 About 30 fruits from each plot were picked to determine single fruit weight (SFW) and
M

183 six fruits were used for nutrient and fruit juice analysis. Yields were calculated by
D

184 weighing all fruits from each plot. Total N, P, and K content of fruit and leaves was
TE

185 determined by indigo-phenol blue colorimetry, vanadium-molybdenum yellow

186 colorimetry and flame photometry, respectively (Rukun, 1999). Total acid (TA) of the
EP

187 fruit juice was determined by acid base titration. Total soluble solids (TSS) of the fruit
C

188 juice were determined using an Abbe refractometer (NAR- IT, Japan).
AC

189 2.4 Data analysis

190 The optimal fertilization treatment was determined by means of the comprehensive

191 evaluation index (Wang et al., 2011a). In this study, there were three categories of

192 factors upon which the fertilization ratios and proportions were evaluated: yield, fruit

193 quality and environment. The yield factor mainly considered the average yield of each

9
 ACCEPTED MANUSCRIPT

194 fruit tree, the quality factor mainly considered SFW, TA and TSS, and the environmental

195 factor mainly included AN, AP and AK in soil after harvesting.

196 The comprehensive evaluation index was calculated as follows:

197 = + + (1)

PT
198 Where Ey is the evaluation index of the yield factor, Eq is the evaluation index of

RI
199 the fruit quality factor, and Ee is the evaluation index of the environmental factor. Wy,

200 Wq and We are the weightings for yield, fruit quality and environment factors, which

SC
201 were set to 0.4, 0.3 and 0.3, respectively.

U
202 =∑ + + ⋯+ (2)
AN
203 =∑ + +⋯+ (3)

204 =∑ + + ⋯+ (4)
M

205 where Y is the membership value of the yield indicator, Q is the membership value
D

206 of the fruit quality indicator, E is the membership value of the environmental indicator,
TE

207 and W is the weighting factor. In the fruit quality index, the weightings of TSS, TA and

208 SFW were 0.4, 0.3, and 0.3, respectively. In the environmental index, the weightings of
EP

209 AN and AP in soil were 0.5 and 0.5, respectively. The membership values were
C

210 calculated as follows:

AC

211 = − ⁄ − (5)

212 =1− − ⁄ − (6)

213 Where M(Xi) is the membership value, and ∈ [0, 1]; Xi is the measured

214 value of each index; Xmin and Xmax are the minimum and maximum values of each index.

215 The indicators that positively correlated with fertilizers used formula (5), while the

10
 ACCEPTED MANUSCRIPT

216 indicators negatively correlated with fertilizers used formula (6).

217 All data were analyzed using SPSS version 16.0 (SPSS Ine., Chicago, IL, USA).

218 The General Linear Model (GLM) procedure was used to analyze the significance (at P

219 ≤0.05 level) of N, P, K fertilizers and their interactions in explaining the response

PT
220 variable. Least significant difference (LSD) at the 0.05 level was calculated to allow

RI
221 comparisons between the N-P, N-K, P-K treatments for the main effects.

222 3 Results and Discussion

SC
223 3.1 Nitrogen, Phosphorus, and Potassium Applied individually

U
224 The effects of nutrient application rates on satsuma yield, quality and soil nutrients
AN
225 are shown in Table 3 and Fig.1. It can be seen that application of a single nutrient

226 fertilizer, N, P or K, had a significant effect on yield. Among them, N fertilizer had the
M

227 greatest impact, increasing yield by 77-144%, followed by K fertilizer, increasing yield
D

228 by 33-83%, and P fertilizer produced the lowest yield increase of 3-41% (Table 3).
TE

229 Quaggio, et al. (2002) reported that K had greater effects on citrus yield than N and P.

230 Our somewhat different results suggest that the response of citrus yield to fertilizer may
EP

231 differ depending on environmental and cultural conditions (Han et al., 2008). In this
C

232 study, the soil had very low N (56 mg kg-1) and very high P (36 mg kg-1), resulting in
AC

233 the strong yield response to N and a weaker yield response to P. In addition, the weak

234 fruit response to P may be related to high P content in soils and the relatively lower

235 requirement for P than N and K in citrus trees (Köseoglu et al., 1993).

236 3.1.1 Effect of N fertilizer alone

237 Increasing rates of N fertilizer significantly increased satsuma yield but this

11
 ACCEPTED MANUSCRIPT

238 declined at the higher application rate of 0.9 kg N plant-1 (Table 3). Similar results for

239 the response to increasing N rates have also been reported by Quaggio et al. (2002) and

240 Jones et al. (1970), who found that N rates higher than 200 kg N ha-1 decreased yield

241 and quality. High leaf N concentration in the trees supplied with a high N rate (Table 3)

PT
242 may imply a delay in leaf senescence and the growth of other vegetative organs, which

RI
243 led to the consumption of more carbon skeletons and energy and low carbohydrate

244 accumulation in the fruits or storage organs (Choi et al., 2012).

SC
245 Fruit quality parameters such as SFW, TA and TSS were also significantly affected

U
246 by N rates, and exhibited a quadratic trend, as for yield. However, the quality response
AN
247 to N rates was more sensitive than yield, e.g., generally the highest values were reached

248 at 0.3 kg N while yield was maximized at 0.6 kg N. Similar results were reported by
M

249 Alva et al. (2006b), who found that the optimum fertilizer N rate for yield was about
D

250 260 kg N ha-1 yr-1, while that for fruit quality was 224 kg N ha-1 yr-1. However, some
TE

251 reports, such as Jones (1970) and Koo (1988) showed that fruit quality such as TSS and

252 juice percentage decreased with increasing N. These variations can probably be
EP

253 attributed to soil N conditions. In low N soils, appropriate N supply increased fruit
C

254 quality, but in high N soils, a small increase in N fertilizer can reduce fruit quality. The
AC

255 decrease in fruit quality as N supply increases may be due to increased consumption of

256 carbohydrates in N metabolism (Choi et al., 2012). Under conditions of limited

257 carbohydrate in fruits, plants will preferentially use the carbohydrate transported into

258 fruits for the formation of carbon skeletons (Choi et al., 2012), probably resulting in the

259 declining quality index being observed earlier than declining yield.

12
 ACCEPTED MANUSCRIPT

260 As for residual AN in soil, there was not as much in the soil as we expected; in fact,

261 conversely, AN decreased linearly with increasing N application (y = -24.51x + 67.199，

262 R² = 0.59, Fig. 3 and Table 5). This indicated that satsuma trees exhibited the

263 characteristics of luxury N uptake. Similar results were found by Alva et al. (2006b),

PT
264 namely that citrus can absorb and utilize soil N by improving plant nutritional growth in

RI
265 the case of sufficient soil N supply. Soil AP content was the lowest at 0.3 kg N and

266 increased with the increasing N application rates (Table 5), indicating that the addition

SC
267 of low N rates increased P uptake by plants but high N had a somewhat inhibitory effect

U
268 on P absorption. In a study with Arabidopsis thaliana, Yan et al. (2015) also found
AN
269 addition of N decreased P absorption efficiency. The enhancement of P uptake after N

270 application can probably be attributed to acidification of the rhizosphere (Thomson et

M

271 al., 1993) or the effects of mobilizing soil P or increasing P uptake efficiency by roots
D

272 (Ruan et al., 2000), while the limitation of P uptake after high N application probably
TE

273 lies in the competitive inhibition of phosphate absorption through the enhancement of

274 nitrate transport under high nitrogen conditions, due to the absorption of nitrate ions and
EP

275 phosphate ions which are both associated with proton-driven coupled active transport
C

276 (Rausch and Bucher, 2002).

AC

277 3.1.2 Effect of K fertilizer alone

278 K supply significantly increased citrus yield when 0.15 kg plant-1 was applied, but

279 beyond this application rate, yield began to decline markedly (Table 5). Alva et al.

280 (2006a) also demonstrated the positive effect of K fertilization on fruit yield of citrus

281 trees grown in soils with low K levels, but fruit yield significantly reduced with excess

13
 ACCEPTED MANUSCRIPT

282 K supply. The possible reason for a decline in yield at high K rates is salt damage

283 caused by excessive K, which affects the absorption of water and other nutrients such as

284 calcium and magnesium and induces severe defoliation (Alva et al., 2006a). However,

285 in this study we did not observe the phenomenon of branches and leaves wilting under

PT
286 high K rates. On the contrary, fruit trees sprouted many new branches, with growth just

RI
287 like at high N rates. Also, we found leaf N content sharply increased but fruit N content

288 significantly decreased at K application rates of 0.45 kg plant-1 (Table 3). The decrease

SC
289 in the N distribution ratio in fruit and leaves may be the main reason for the decrease in

U
290 yield at high K rates.
AN
291 Fruit size and fruit weight appear to be the most affected fruit traits with varying K

292 availability in citrus groves (Köseoglu et al., 1993; Alva et al., 2006a; Han et al., 2008).
M

293 Alva, et al. (2006a) demonstrated an increase in fruit size and grade with applications of
D

294 fertilizer from 25 to 225 kg ha-1. Embleton et al. (1975) and Koo (1988) found a 15%
TE

295 increase in acid content of lemon with the application of high K compared with low K

296 fertilization. In this study, K fertilizer at less than 0.30 kg plant-1 resulted in a rather
EP

297 marginal increase in fruit size (Table 5), a further increase in K rates to 0.45 kg plant-1
C

298 then a significantly decreased fruit weight due to the limitation of N in fruit. TA content
AC

299 declined sharply with an increase in K rate from 0.15 to 0.45 kg ha-1, which was related

300 to the activity of decomposing organic acid enzymes, such as isocitrate dehydrogenase

301 and aconitase induced by K fertilizer (Walker and Famiani, 2018). TSS content

302 increased with K fertilizer application compared to no K application, there was no

303 significant difference between K fertilizers.

14
 ACCEPTED MANUSCRIPT

304 The addition of K fertilizer reduced the content of residual AN in soil after

305 harvesting, but there was no significant difference, with an increase in K rate from 0.3

306 to 0.9 mg plant-1 (Table 5), which indicated that K had a positive effect on N uptake by

307 plants. The soil AP content decreased with the increase in K rates less than 0.2 kg plant-1.

PT
308 However, further increases in K rates to 0.3 kg plant-1 increased soil AP content, which

RI
309 may be due to the increase in nitrate nitrogen absorption inhibiting the absorption of

310 phosphorus by plants exposed to high potassium levels (Rausch and Bucher, 2002). The

SC
311 high N content and low P content in leaves under excessive K supply (Table 3) may

U
312 support the above speculation. The addition of K fertilizer from 0.15 to 0.45 mg kg-1
AN
313 had no significant effect on soil AK and just showed a little increase compared with no

314 K fertilizer treatment, which was probably due to high K content in this soil.
M

315 3.1.3 Effect of P fertilizer alone

D

316 The effect of P on citrus yield was very pronounced in this study (Table 5), yield
TE

317 increased with the increasing P rates from 0.1 to 0.3 kg P plant-1 (Table 5). Quaggio et.

318 Al. (2002) and Obreza et al. (2008) also found that P supply can significantly increase
EP

319 citrus yield, but when soil P exceeded 30 mg dm-3, a linear and negative effect of this
C

320 nutrient on yield was observed. These results suggest that a small increase in P fertilizer
AC

321 application promotes yield, but an excessive supply reduces yield. Generally, calcium is

322 present in P fertilizer, because it is often applied in the form of superphosphate; this may

323 lead to a decrease in yield by reducing K uptake because of competition between ions.

324 In this study, the K content of fruits and leaves was significantly reduced during

325 excessive P supply (see Table 3), indicating that the process causing calcium ions to

15
 ACCEPTED MANUSCRIPT

326 inhibit K uptake by competition had already taken place.

327 As for citrus quality indicators, the change in SFW was consistent with the yield,

328 indicating that P fertilizer mainly affects the yield by affecting the fruit size (Quaggio et

329 al., 2002). However, both TSS and TA both exhibited a quadratic trend in relation to P

PT
330 application rates: TSS and TA reached their maximum when 0.1 kg P plant-1 was applied,

and further increases of P up to 0.2 and 0.3 kg P plant-1 caused significant reductions.

RI
331

332 Similar results were reported by Quaggio et al. (2006) with low soil P, the increase in

SC
333 TSS per area varied from 2411 to 2738 kg·ha–1 with rates of P varying from 9 to 63

U
334 kg·ha–1. In contrast, in high soil P, these same rates promoted a decrease in total yield
AN
335 from 2324 to 2282 kg·ha–1. Excessive P fertilization has been found to reduce orange

336 juice soluble solids and delay external color development (Obreza et al., 2008). The
M

337 decrease of K ions in trees (Table 3) under high P rates was the main reason for the
D

338 decline in quality.

TE

339 The residual AN, AP and AK content in soil decreased with the application of 0.1

340 to 0.2 kg P plant-1, however, when P fertilizer was added at a rate of 0.3 kg P plant-1, the
EP

341 content of these available nutrients in soil increased significantly (Table 5). These
C

342 results once again show that the optimum amount of P application is 0.1-0.2 kg plant-1,
AC

343 and exceeding this would reduce the absorption of other nutrients by plants.

344 3.2 Nitrogen, Phosphorus, and Potassium interactions

345 3.2.1 The interaction between N and P

346 N and P nutrition are closely related, and there is often a significant interaction

347 between N and P fertilizers in relation to the yield and quality of crops (Ruan et al.,

16
 ACCEPTED MANUSCRIPT

348 2000; Han et al., 2008). Wallace (1990) reported that when N and P were applied

349 together without K, severe antagonism resulted with a large yield decrease. In the

350 current study, a positive effect on yield of the interaction between N and P was observed.

351 The combination of 0.6 kg N and 0.3 kg P has the highest yield: 160% and 80% higher

PT
352 than that of N fertilizer alone and P fertilizer alone respectively. However, an

RI
353 inappropriate N-P ratio reduced the positive interaction effect. For example, under

354 moderate N rates, the addition of 0.3 kg P plant-1 significantly reduced yield, with a

SC
355 decline of 50% compared with the 120 kg plant-1 yield when 0.2 kg P plant-1 had been

U
356 added.
AN
357 For fruit quality factors, a positive significant N×P interaction was also evident in

358 the combination of low and medium levels of N and P, but under the combined
M

359 treatment of high amounts of N and P, a negative interaction was observed for TA and
D

360 TSS (Table 6). Bouma and Mcentyre (1963) also reported a negative interaction
TE

361 between N and P in a factorial field experiment with citrus, which was attributed to the

362 addition of N reducing availability of soil P due to the increase in soil acidity when
EP

363 there were higher levels of ammonium sulphate. In the study, the reason for the negative
C

364 interaction between N and P affecting fruit quality under high P may be due to the
AC

365 increased absorption of N by leaves when there is high P (Table 3), which affects the

366 carbon transfer in fruits (Choi et al., 2012) and thus reduces the quality.

367 In this study, we did not observe a statistically significant interaction between N

368 and P on soil available nutrients; this was due to the large difference between soil

369 nutrients in different samples caused by the large spatial variation in the soil.

17
 ACCEPTED MANUSCRIPT

370 Nevertheless, the effect of the N and P interaction on soil residual nutrients was obvious

371 (Table 6). The combined application of medium and low amounts of N and P fertilizer

372 could promote the absorption of N, P and K nutrients by plants, and decreased the

373 nutrient content in the soil, but excessive N or P increased soil nutrient content.

PT
374 3.2.2 The interaction between N and K

RI
375 Positive effects on yield of the interaction between N and K have been reported in

376 various crops (Barnes et al., 1976), while negative interactions are rarely found (Zhang

SC
377 et al., 2010). In this study, a significant positive interaction was observed, but

U
378 insufficient or excessive amounts of either N or K decreased this interaction effect
AN
379 (Table 7). In all N and K treatments, the combination of 0.6 kg N and 0.3 kg K produced

380 the strongest interaction, resulting in a 175 per cent yield increase compared with the no
M

381 fertilizer treatment. The increase in yield can be mainly attributed to the increase in
D

382 SFW (Table 7), with a significantly positive correlation with SFW (y = 0.26x + 54.12,
TE

383 R² = 0.533).

384 Positive significant N×K interactions on fruit SFW, TA and TSS content were
EP

385 observed (Table 7). Similar results were reported by Köseoglu et al. (1993) with
C

386 Satsuma Mandarins (Citrus unshiu marc.) and Quaggio et al. (2006) with Sicilian lemon
AC

387 in Brazil. The rates of N and K application for the maximum SFW, TA and TSS were

388 different: 0.6 kg N with 0.15 kg K, 0.6 kg N with 0.30 kg K and 0.3 kg N with 0.15 kg

389 K, respectively. This indicates that the ratio of N to K needs to be adjusted according to

390 the quality objectives. In the current study, 0.3kg N and 0.15 kg K per tree was optimum

391 based on the grade standard of fruit quality for satsumas (Citrus unshiu Marc) (Jun et al.,

18
 ACCEPTED MANUSCRIPT

392 2012).

393 N and K fertilizers have significant interaction when soil available nutrients are

394 considered (Bouma and Mcentyre, 1963; Alva et al., 2006a; Alva et al., 2006b; Han et

395 al., 2008). N, through its interaction with K, decreased the content of soil AN, AP and

PT
396 AK. For example, with N at 0.6 kg plant-1 increasing rates of K significantly decreased

RI
397 soil AN by 32-41%, AP by 26-67% and AK by 6% compared with no K (Table 7). The

398 reason is an increase in plant growth resulting from the positive interaction effect of N

SC
399 and K with nitrate ions accompanying anions during K ion uptake and/or transport

U
400 (Zhang et al., 2010). In addition, we found with N application at 0.6 kg plant-1 the
AN
401 addition of high rates of K increased the content of soil AN by 31-54%, AP by 433-441%

402 and AK by 11-12%, compared with a low K application rate. This is probably related to
M

403 antagonism between high K and other nutrient ions, such as magnesium (Zhang et al.,
D

404 2010), limiting the growth of plants and then reducing he uptake of these nutrients.
TE

405 3.2.3 The interaction between P and K

406 The effect on yield of the interaction between P and K was significant (Table 8). K
EP

407 fertilizer obviously increased the effects of P on yield, e.g., under moderate P rates, the
C

408 yield increased from 65.5 kg plant-1 with no K applied to 103.6 kg plant-1 with 0.15 kg
AC

409 plant-1. In contrast, P fertilizer also increased the effects of K on yield. Under moderate

410 K rates, the yield increased from 87.3 kg plant-1 with no K applied to 120 kg plant-1.

411 However, yield reductions were found when at high rates of P or K or a combination of

412 a high rate of one nutrient with a low rate of the other. This reveals that excessive

413 application of individual elements was affected by a lack of other elements, thus

19
 ACCEPTED MANUSCRIPT

414 limiting the yield (Quaggio et al., 2002; Han et al., 2008). Generally, high rates of P can

415 induce zinc or molybdenum deficiency in soil (Agbenin, 1998), and high rates of K can

416 affect the absorption of Ca and Mg by plants (Zhang et al., 2010). In this study, we did

417 not further investigate the limiting elements that reduced yield with high potassium and

PT
418 high phosphorus applications. Overall, the ratios of P and K fertilizer for the highest

RI
419 yield were 0.20:0.15-0.30 with low and moderate N rates.

420 P and K fertilizer interacted to significantly affect fruit quality. (Köseoglu et al.,

SC
421 1993; Xian et al., 1993). Like previous reports, we also observed significant positive

U
422 interactions on fruit quality of citrus. At low and moderate P application rates, the
AN
423 combination of K and P fertilizer significantly increased the fruit SFW, TA and TSS

424 content by 21-160 %, 62-127% and 9-26%, respectively, compared with no P or K

M

425 fertilizer (Table 8). However, the combinations with high P or high K both reduced
D

426 these fruit quality parameters. The combination of P and K at 0.1 and 0.3 kg,
TE

427 respectively was the most effective treatment for optimizing TSS content and TSS/TA.

428 Soil available nutrients were also affected by the P × K interaction, and the
EP

429 combined addition of K and P fertilizer increased uptake of N, P and K by plants and
C

430 decreased soil AN, AP and AK content (Quaggio et al., 2002; Hussain et al., 2015). In
AC

431 this study, with 0.2 kg P plant-1, the increasing rates of K resulted in a significant

432 decrease in AN, AP and AK contents, which were 44-51%, 52-78% and 2-11%,

433 respectively, compared with treatments without P or K fertilizer (Table 8). In contrast,

434 the addition of P increased the effects of K on soil AN, AP and AK, but the application

435 of excessive P at a rate of 0.30 kg P plant-1 increased the contents of AN, AP and AK in

20
 ACCEPTED MANUSCRIPT

436 soils. Similar results were reported by Wang et al. (2011b), namely that interactions

437 between P and K affecting rice quality were synergistic when fertilizer application rates

438 were at lower or moderate levels, otherwise they were antagonistic. In this study, it was

439 possible that the antagonistic effect of phosphate and sulfate ions associated with high

PT
440 rates of application of potassium sulfate fertilizer increased available nutrients in the

RI
441 soil (Wang et al., 2011b). Shankaralingappa et al. (2000) reported that the interaction

442 effect of phosphorus and sulfur depressed uptake of N, P and K by pigeon pea (Cajanus

SC
443 cajan). The combination of P and K at 0.2 and 0.3 kg, respectively was the most

U
444 effective treatment to reduce residual available nutrients in the soil.
AN
445 3.3 Optimum Fertilization Rate

446 The amount of N, P and K fertilizer applied to citrus varies depending on which
M

447 response variable needs to be maximized (Bouma and Mcentyre, 1963; Han et al.,
D

448 2008). For example, Quaggio et al. (2006) in Brazil found that nutrient rates for
TE

449 optimizing fruit yield were 189 kg·ha-1, P rates of 9 kg·ha-1 and K rates of 124 kg ha-1,

450 respectively, greater than that for maximum fruit yield, which required 174 kg·ha-1 N, 9
EP

451 kg·ha-1 P and 97 kg ha-1 K. Du Plessis and Koen (1988) found maximum fruit yield of
C

452 orange trees was attained at N 225 kg·ha-1 and K 310 kg·ha-1, but superior fruit size for
AC

453 the fresh market was recorded at a lower N rate of 45 kg·ha-1. As in previous studies, we

454 also found that higher nutrient application rates were required for maximum fruit yield

455 than for optimizing fruit quality. Yield was maximized at 0.6 kg N plant-1, 0.2 kg P

456 plant-1, and 0.3 kg K plant-1, whereas optimum fruit quality was recorded at the lower

457 rates of 0.3 kg N, 0.2kg P, and 0.15kg K, due to the slower yield response to fertilizer

21
 ACCEPTED MANUSCRIPT

458 than fruit quality.

459 If considering the yield, fruit quality and environment, the best fertilizer rates were

460 0.3 kg N, 0.2 kg P, and 0.3kg K (Table 9), in which the proportion of K fertilizer was

461 higher than the rate for optimum fruit quality. This indicates that K fertilizer could play

PT
462 a good regulatory role in reducing soil residual available nutrients and improving fruit

RI
463 quality. At present, the local recommended application rates are 0.5-0.8 kg N for each

464 adult citrus tree in orchards with a yield of about 30 t ha-1, and the ratio of N to P to K is

SC
465 1:0.4-0.6: 0.5-0.8 (Wang Chengqiu et al., 2005), which is higher than that indicated by

U
466 our comprehensive consideration of environment, quality and yield. This coincides with
AN
467 our expectations, because the amount of fertilizer required for maximum yields usually

468 reduces the citrus quality indicators and increases the risk of soil nitrogen and
M

469 phosphorus pollution (Quaggio et al., 2006). Therefore, we suggest that the
D

470 recommended amount fertilizer applied to citrus crops should be reduced in order to
TE

471 protect the environment and improve fruit quality.

472 In addition, we compared the best fertilization treatments on the basis of an

EP

473 economic benefit analysis (Fig. 2). Using the current local unit price of high-quality
C

474 fruit of RMB 2.0 yuan kg-1 and standard fruit of RMB 1.4 yuan kg-1, the income
AC

475 resulting from the recommended fertilizer application level according to our

476 comprehensive evaluation was not reduced because of the decline in yield, but increased

477 a little. The net income using our recommendation is about RMB 111,825 yuan ha-1,

478 which is higher than that of RMB 105,510 yuan ha-1 using the current optimum

479 fertilization treatment to maximize yield (Fig. 2). Therefore, with the demand for high

22
 ACCEPTED MANUSCRIPT

480 quality fruit and protection of the environment, we advocate that quality and

481 environmental indicators should be included when evaluating fertilizer

482 recommendations in the future.

483 4. Conclusion

PT
484 N, P and K fertilizers and their interaction significantly influence the yield and

RI
485 quality of citrus and residual soil available nutrients. A deficiency or excess of N, P or K,

486 as well as a relative imbalance, such as high levels of one nutrient combined with low

SC
487 levels of the other nutrients, will lead to a decline in yield and fruit quality. Thus, fruit

U
488 production can be optimized by exploiting these nutrient interactions. Citrus yield was
AN
489 less responsive to fertilizer application than the fruit quality index, which resulted in

490 lower fertilizer application rates in the treatment with better fruit quality than those
M

491 delivering the highest yield. Considering the effects of fertilizer on yield, quality and
D

492 potential environmental pollution, the optimum fertilizer application for each adult
TE

493 citrus tree in orchards is 0.3kg N, 0.2kg P, and 0.3 kg K plant-1 year-1.

494 Acknowledgements
EP

495 This study was financially supported by the projects of National Key Research and
C

496 Development Plan (2016YFD0200108), National Natural Science Foundation of China

AC

497 (41501313), and the Knowledge Innovative Program of the Chinese Academy of

498 Sciences (Y455434I02).

499

500 References

501 Agbenin, J., 1998. Phosphate‐induced zinc retention in a tropical semi‐arid soil.

23
 ACCEPTED MANUSCRIPT

502 European Journal of Soil Science 49, 693-700.

503 Akbariyeh, S., Bartelthunt, S., Snow, D., Li, X., Tang, Z., Li, Y., 2018.

504 Three-dimensional modeling of nitrate-N transport in vadose zone: Roles of soil

505 heterogeneity and groundwater flux. Journal of Contaminant Hydrology, 15-25.

PT
506 Alva, A.K., Mattos Jr, D., Paramasivam, S., Patil, B., Dou, H., Sajwan, K.S., 2006a.

RI
507 Potassium management for optimizing citrus production and quality. International

508 Journal of Fruit Science 6, 3-43.

SC
509 Alva, A.K., Paramasivam, S., Obreza, T.A., Schumann, A.W., 2006b. Nitrogen best

U
510 management practice for citrus trees : I. Fruit yield, quality, and leaf nutritional status.
AN
511 Scientia Horticulturae 109, 223-233.

512 Barnes, A., Greenwood, D.J., Cleaver, T.J., 1976. A dynamic model for the effects of
M

513 potassium and nitrogen fertilizers on the growth and nutrient uptake of crops. Journal of
D

514 Agricultural Science 86, 225-244.

TE

515 Bouma, D., Mcentyre, G., 1963. A factorial field experiment with citrus. Journal of

516 Horticultural Science 38, 175-198.

EP

517 Choi, S.T., Park, D.S., Kang, S.M., Kang, S.K., 2012. Influence of leaf-fruit ratio and
C

518 nitrogen rate on fruit characteristics, nitrogenous compounds, and nonstructural

AC

519 carbohydrates in young persimmon trees. HortScience 47, 410-413.

520 Du, C.W., 2003. Effects of boron and calcium supply on calcium fractionation in plants

521 and suspension cells of rape cultivars with different boron efficiency. Journal of Plant

522 Nutrition 26, 789-806.

523 Du Plessis, S., Koen, T., 1988. The effect of N and K fertilization on yield and fruit size

24
 ACCEPTED MANUSCRIPT

524 of Valencia. Proceeding of sixth international citrus congress, Rehovot, pp. 663-672.

525 Embleton, T., Jones, W., Platt, R., 1975. Plant nutrition and citrus fruit crop [oranges,

526 lemons] quality and yield. HortScience (USA) 10, 48-50.

527 Gaj, R., Górski, D., Przybył, J., 2013. Effect of differentiated phosphorus and potassium

PT
528 fertilization on winter wheat yield and quality. Journal of Elementology 18, 55-67.

RI
529 Gao, J., Thelen, K.D., Min, D.-H., Smith, S., Hao, X., Gehl, R., 2010. Effects of manure

530 and fertilizer applications on canola oil content and fatty acid composition. Agronomy

SC
531 Journal 102, 790-797.

U
532 Gu, N., Zhao, L.P., Zhao, X.M., 2014. A review and perspective on slow and controlled
AN
533 release fertilizer in China. Applied Mechanics & Materials 535, 222-225.

534 Han, S.G., Hancheol, L., Jaeho, J., Kyunghwan, M., Kang, T.W., Song, S.J., 2008.
M

535 Effects of long-term application of N, P, K fertilizers on fruit quality and yield of citrus
D

536 tree (Citrus unshiu Marc.). Wonye kwahak kisulchi = Korean journal of horticultural
TE

537 science and technology / 26, 203-208.

538 Hui, L., Cong, R., Tao, R., Li, X., Ma, C., Lei, Z., Zhi, Z., Lu, J., 2015. Yield response
EP

539 to N fertilizer and optimum N rate of winter oilseed rape under different soil indigenous
C

540 N supplies. Field Crops Research 181, 52-59.

AC

541 Hussain, Z., Khattak, R., Fareed, I., Irshad, M., Mahmood, Q., 2015. Interaction of

542 phosphorus and potassium on maize (Zea mays L.) in saline-sodic soil. Journal of

543 Agricultural Science 7, 66.

544 Jones, W., 1970. Nitrogen rate effects on lemon production, quality and leaf nitrogen.

545 Journal of the American Society of Horticultural Science 95, 46-49.

25
 ACCEPTED MANUSCRIPT

546 Jun, L.U., Ting, L.I., Huang, Q.R., Zhang, D.C., 2012. Analysis and evaluation on the

547 grading factors of citrus. Hubei Agricultural Sciences, 4631-4633.

548 Kawada, K., 1984. Effect of sorbic acid and potassium sorbate on the control of sour rot

549 of citrus fruits. Proc. Fla. State Hort. Soc, pp. 133-135.

PT
550 Kong, W., Zeng, Z., Xiong, W., Zhengliang, W.U., Xia, R., 2018. Production forecast of

RI
551 citrus in China and production and marketing situation of citrus in Chongqing in 2016

552 production season. Asian Agricultural Research, 16-31.

SC
553 Koo, R., 1988. Fertilization and irrigation effects on fruit quality. Factors Affecting Fruit

U
554 Quality. Citrus Short Course Proc, 35-42.
AN
555 Köseoglu, A., Eryüce, N., Colakoglu, H., 1993. The effects of N, P, K fertilizers on fruit

556 yield and quality of satsuma mandarines (Citrus unshiu Marc.). International
M

557 Symposium on Quality of Fruit and Vegetables: Influence of Pre-and Post-Harvest

D

558 Factors and Technology 379, pp. 89-96.

TE

559 Liu, R., Kang, Y., Zhang, C., Pei, L., Wan, S., Jiang, S., Liu, S., Ren, Z., Yang, Y., 2014.

560 Chemical fertilizer pollution control using drip fertigation for conservation of water
EP

561 quality in Danjiangkou Reservoir. Nutrient cycling in agroecosystems 98, 295-307.

C

562 Liu, Y., Tao, Y., Wan, K., Zhang, G., Liu, D., Xiong, G., Chen, F., 2012. Runoff and
AC

563 nutrient losses in citrus orchards on sloping land subjected to different surface mulching

564 practices in the Danjiangkou Reservoir area of China. Agricultural Water Management

565 110, 34-40.

566 Nan, W., Yue, S., Li, S., Huang, H., Shen, Y., 2016. Characteristics of N2O production

567 and transport within soil profiles subjected to different nitrogen application rates in

26
 ACCEPTED MANUSCRIPT

568 China. Science of the Total Environment 542, 864-875.

569 Obreza, T.A., Rouse, R.E., Morgan, K.T., 2008. Managing phosphorus for citrus yield

570 and fruit quality in developing orchards. HortScience 43, 2162-2166.

571 Pervez, H., Ashraf, M., Makhdum, M.I., 2004. Effects of potassium rates and sources on

PT
572 fiber quality parameters in four cultivars of cotton grown in aridisols. Journal of Plant

RI
573 Nutrition 27, 2235-2257.

574 Quaggio, J., Mattos Jr, D., Cantarella, H., Almeida, E., Cardoso, S., 2002. Lemon yield

SC
575 and fruit quality affected by NPK fertilization. Scientia Horticulturae 96, 151-162.

U
576 Quaggio, J.A., Mattos, D., Cantarella, H., 2006. Fruit yield and quality of sweet oranges
AN
577 affected by nitrogen, phosphorus and potassium fertilization in tropical soils. Fruits 61,

578 293-302.
M

579 Rausch, C., Bucher, M., 2002. Molecular mechanisms of phosphate transport in plants.
D

580 Planta 216, 23-37.

TE

581 Reid, J.B., Trolove, S.N., Tan, Y., Johnstone, P.R., 2016. Nitrogen or potassium

582 preconditioning affects uptake of both nitrate and potassium in young wheat (Triticum
EP

583 aestivum). Annals of Applied Biology 168, 66-80.

C

584 Ruan, J., Zhang, F., Wong, M.H., 2000. Effect of nitrogen form and phosphorus source
AC

585 on the growth, nutrient uptake and rhizosphere soil property of Camellia sinensis L.

586 Plant and Soil 223, 65-73.

587 Rukun, L., 1999. Analysis method of soil agro-chemistry. . China Agricultural Science

588 and Technology Press, Beijing.

589 Shankaralingappa, B., Shivaraj, B., Sudhir, K., Viswanatha, K., 2000. Interaction effect

27
 ACCEPTED MANUSCRIPT

590 of phosphorus and sulphur on uptake of nitrogen, phosphorus, potassium and sulphur by

591 pigeonpea (Cajanus cajan). Indian Journal of Agronomy 45, 348-352.

592 Thomson, C., Marschner, H., Römheld, V., 1993. Effect of nitrogen fertilizer form on

593 pH of the bulk soil and rhizosphere, and on the growth, phosphorus, and micronutrient

PT
594 uptake of bean. Journal of Plant Nutrition 16, 493-506.

RI
595 Walker, R.P., Famiani, F., 2018. Organic acids in fruits: metabolism, functions and

596 contents. Hortic Rev, 371-430.

SC
597 Wallace, A., 1990. Nitrogen, phosphorus, potassium interactions on Valencia orange

U
598 yields. Journal of plant nutrition 13, 357-365.
AN
599 Wang Chengqiu, Jiao Bining, Tang Zhonghai, Zhao Qiyang, 2005. Study on the quality

600 and safety of citrus and Its Countermeasures. Fruit Trees of Southern China 34, 19-21.
M

601 Wang, F., Kang, S., Du, T., Li, F., Qiu, R., 2011a. Determination of comprehensive
D

602 quality index for tomato and its response to different irrigation treatments. Agricultural
TE

603 Water Management 98, 1228-1238.

604 Wang, W.N., Lu, J.W., He, Y.Q., Li, X.K., Li, H., 2011b. Effects of N, P, K fertilizer
EP

605 application on grain yield, quality, nutrient uptake and utilization of rice. Chinese
C

606 Journal of Rice Science 25, 645-653.

AC

607 Westermann, D.T., James, D.W., Tindall, T.A., Hurst, R.L., 1994. Nitrogen and

608 potassium fertilization of potatoes: Sugars and starch. American Potato Journal 71,

609 433-453.

610 Xian, Y., Guan, P., Li, B., 1993. Effects of NK interaction on yield, quality and uptake

611 of nitrogen, phosphorus and potassium in potato (Solanum tuberosum L.). Journal of

28
 ACCEPTED MANUSCRIPT

612 South China Agricultural University 1, 28-32.

613 Yan, Z., Kim, N., Han, W., Guo, Y., Han, T., Du, E., Fang, J., 2015. Effects of nitrogen

614 and phosphorus supply on growth rate, leaf stoichiometry, and nutrient resorption of

615 Arabidopsis thaliana. Plant and soil 388, 147-155.

PT
616 Yang, X.W., Tian, X.H., Lu, X.C., Cao, Y.X., Mai, W.X., 2009. Effect of varying

RI
617 combination of P and Zn in chelater-buffered solution on P-Zn interaction in wheat

618 seedling. Chinese Journal of Eco-Agriculture 17, 1055-1062.

SC
619 Zhang, F., Niu, J., Zhang, W., Chen, X., Li, C., Yuan, L., Xie, J., 2010. Potassium

U
620 nutrition of crops under varied regimes of nitrogen supply. Plant and soil 335, 21-34.
AN
621 Zhu, Z.L., Chen, D.L., 2002. Nitrogen fertilizer use in China – Contributions to food

622 production, impacts on the environment and best management strategies. Nutrient
M

623 Cycling in Agroecosystems 63, 117-127.

D

624
TE

625
EP

626
C
AC

29
 ACCEPTED MANUSCRIPT

627 Table 1 Experimental design of the citrus fertilizer trials

Treatments Nutrient rates

N P2O5 K2O
kg N plant-1 year-1 kg P2O5 plant-1 year-1 kg K2O plant-1 year-1
N0P0.2K0.3 0.0 0.2 0.30

PT
N0.3P0.1K0.3 0.3 0.1 0.30
N0.3P0.2K0.15 0.3 0.2 0.15
N0.3P0.2K0.3 0.3 0.2 0.30

RI
N0.6P0K0.3 0.6 0.0 0.30
N0.6P0.1K0.15 0.6 0.1 0.15
N0.6P0.1K0.3 0.6 0.1 0.30

SC
N0.6P0.2K0 0.6 0.2 0.00
N0.6P0.2K0.15 0.6 0.2 0.15
N0.6P0.2K0.3 0.6 0.2 0.30

U
N0.6P0.2K0.45 0.6 0.2 0.45
AN
N0.6P0.3K0.3 0.6 0.3 0.30
N0.9P0.2K0.3 0.9 0.2 0.30
N0P0K0 0.0 0.0 0.00
M

628 Note: N0, N0.3, N0.6, N0.9 indicate four N application rates, 0, 0.3, 0.6 and 0.9 kg plant-1,

629 P0, P0.1, P0.2, P0.3 indicate four P application rates, 0, 0.1, 0.2 and 0.3 kg plant-1, K0, K0.15,
D

630 K0.30, K0.45 indicate four K application rates, 0, 0.15, 0.3 and 0.45 kg plant-1.
TE
C EP
AC

30
 ACCEPTED MANUSCRIPT
631 Table 2 Monthly N (urea), P (superphosphate), K (potassium sulfate) application

632 schedule

Proportions of annual total fertilization (%)

Treatments
N P2O5 K2O
2015, December 35 100 40

PT
2016, March 45 - -
2016, July 20 - 60
2016, November 35 100 40

RI
2017, March 45 - -
2017, July 20 - 60
2017, November 35 100 40

U SC
AN
M
D
TE
C EP
AC

31
 ACCEPTED MANUSCRIPT

633 Table 3 Yield, fruit quality and soil available nutrient content of satsumas under different fertilizer combinations

PT
Fruit Quality Soil Available Nutrient content Leaf Fruit
Yield

RI
Treatment SFW TA TSS AN AP AK N P K N P K
kg plant-1
g % % mg kg-1 mg kg-1 mg kg-1 % % % % % %

SC
N0P0.2K0.3 49.1 61.7 0.26 12.66 60.9 29.6 221.5 1.83 0.119 0.66 0.13 0.021 0.13
N0.3P0.1K0.3 56.7 81.7 0.41 12.00 64.5 19.3 205.5 1.82 0.120 1.09 0.14 0.013 0.12
N0.3P0.2K0.15 98.2 78.3 0.26 14.26 60.9 47.2 221.5 1.86 0.110 0.60 0.14 0.014 0.09

U
N0.3P0.2K0.3 87.3 62.9 0.27 14.20 51.5 18.6 221.5 1.72 0.120 0.74 0.19 0.020 0.10

AN
N0.6P0K0.3 85.1 34.3 0.15 12.86 45.2 24.8 206.4 1.96 0.121 0.98 0.23 0.019 0.09
N0.6P0.1K0.15 109.1 61.0 0.26 12.68 55.4 54.2 218.8 2.01 0.102 0.45 0.22 0.017 0.10
N0.6P0.1K0.3 87.3 64.4 0.27 13.98 49.0 13.0 197.5 1.94 0.126 0.99 0.19 0.020 0.12

M
N0.6P0.2K0 65.5 73.3 0.21 11.07 65.0 33.8 202.8 1.71 0.118 0.79 0.16 0.016 0.11
N0.6P0.2K0.15 103.6 89.4 0.31 12.52 44.1 25.0 209.9 1.67 0.114 0.93 0.19 0.020 0.12

D
N0.6P0.2K0.3 120.0 88.7 0.34 12.77 38.2 11.2 191.3 1.74 0.123 0.91 0.13 0.013 0.10
N0.6P0.2K0.45 87.3 74.7 0.13 12.94 58.8 60.6 211.7 2.10 0.112 0.69 0.12 0.021 0.09

TE
N0.6P0.3K0.3 98.2 78.8 0.11 11.00 62.0 66.3 254.4 2.06 0.117 0.67 0.16 0.015 0.08
N0.9P0.2K0.3 46.9 68.6 0.18 13.14 50.1 59.7 213.5 2.21 0.116 1.12 0.19 0.018 0.11
N0P0K0 43.6 72.2 0.19 11.28 78.4 51.9 214.4 1.87 0.119 0.89 0.19 0.017 0.10
EP
634 Note. N0, N0.3, N0.6, N0.9 indicate four N application rates: 0, 0.3, 0.6 and 0.9 kg plant-1, P0, P0.1, P0.2, P0.3 indicate four P application rates: 0, 0.1,
C
AC

635 0.2 and 0.3 kg plant-1, K0, K0.15, K0.30, K0.45 indicate four K application rates: 0, 0.15, 0.3 and 0.45 kg plant-1. SFW: single fruit weight, TA: total

636 acid of fruit, TSS: total soluble solids, AN: available nitrogen, AP: available phosphorous, AK: available potassium.

32
 ACCEPTED MANUSCRIPT

637 Table 4 Mean squares from ANOVA of yield, fruit quality, and soil residual available

638 nutrients after satsuma harvesting as influenced by N, P and K fertilization

F test source df Yield SFW TA TSS AN AP AK

FN 3 ** ** * ** ** ** NS
FP 3 ** ** ** ** ** ** **

PT
FK 3 ** ** ** * ** ** NS
FN×FP 1 ** ** ** ** NS NS **
FN×FK 1 ** ** ** NS NS ** **

RI
FP×FK 1 ** NS ** * NS ** NS
Note: SFW: single fruit weight, TA: total acid of fruit, TSS: total soluble solids, AN: available

SC
nitrogen, AP: available phosphorous, AK: available potassium. NS, *, ** Nonsignificant or
significant at P<0.05 or 0.01.

U
639
AN
640
M

641
D

642
TE

643
EP

644
C
AC

33
 ACCEPTED MANUSCRIPT

645 Table 5 Effects of single N, P and K on yield, fruit quality and soil available nutrient

646 content of satsumas

Fruit Quality Residual nutrient content in soil

Rate Yield
SFW TA TS AN AP AK
kg plant-1 kg plant-1
g % % mg kg-1 mg kg-1 mg kg-1

PT
N
0 46.4 c 67.0 c 0.23 b 12.0 b 69.7 a 40.8 b 218.0
0.3 80.7 b 74.3 a 0.31 a 13.5 a 59.0 b 28.4 c 216.2

RI
0.6 94.5 a 70.6 b 0.22 b 12.5 b 52.2 b 36.1 b 211.6
0.9 46.9 c 68.6 c 0.18 c 13.1 a 50.1 b 59.7 a 213.5

SC
P
0 64.4 c 53.3 c 0.17 c 12.1 b 61.8 a 38.4 b 210.4 b
0.1 84.4 b 69.0 b 0.31 a 12.9 a 56.3 b 28.8 c 207.3 b
0.2 82.2 b 74.7 a 0.25 b 12.9 a 53.7 b 35.7 b 211.7 b
0.3 98.2 a 78.8 a

U
0.11 c 11.0 c 62.0 a 66.3 a 254.4 a
AN
K
0 54.5 c 72.8 a 0.20 ab 11.2 b 71.7 a 42.9 b 208.6
0.15 103.6 a 76.2 a 0.28 a 13.2 a 53.5 b 42.1 b 216.7
M

0.3 78.8 b 74.7 a 0.25 a 12.8 a 52.7 b 30.3 c 214.0

0.45 87.3 b 67.6 b 0.13 b 12.9 a 58.8 b 60.6 a 211.7
NS, *, ** Nonsignificant or significant at P<0.05 or 0.01
D

647
TE
C EP
AC

34
 ACCEPTED MANUSCRIPT

648 Table 6 Effects of N×P interaction on yield, fruit quality and soil available nutrient content of satsumas

PT
N P Yield Increase Quality Available nutrient content in soil

RI
SFW Increase TA Increase a TSS increase AN increase AP increase AK increase
-1 -1 -1 -1 -1
kg plant kg plant % g % % % % % mg kg % mg kg % mg kg %

SC
0 0 43.6 72.2 0.19 11.28 78.4 51.9 214.4
0 0.2 49.1 12.5 61.7 -14.5 0.26 36.8 12.7 12.2 60.9 -22.3 29.6 -43.0 221.5 3.3
0.3 0.1 56.7 30.0 81.7 13.2 0.41 115.8 12.0 6.4 64.5 -17.7 19.3 -62.8 205.5 -4.2

U
0.2 87.3 100.0 62.9 -12.9 0.27 42.1 14.2 25.9 51.5 -34.3 18.6 -64.2 221.5 3.3

AN
0.6 0 85.1 95.0 34.3 -52.5 0.15 -21.1 12.9 14.0 45.2 -42.3 24.8 -52.2 206.4 -3.7
0.1 87.3 100.0 64.4 -10.8 0.27 42.1 14.0 23.9 49.0 -37.5 13.0 -75.0 197.5 -7.9
0.2 120.0 175.0 88.7 22.9 0.34 78.9 12.8 13.2 38.2 -51.3 11.2 -78.4 191.3 -10.8

M
0.3 98.2 125.0 78.8 9.1 0.11 -42.1 11.0 -2.5 62.0 -20.9 66.3 27.7 254.4 18.7
0.9 0.2 46.9 7.5 68.6 -5.0 0.18 -5.3 13.1 16.5 50.1 -36.1 59.7 15.0 213.5 -0.4

D
Note: Positive signs denote an increased percentage compared with the control, while negative signs denote a decreased percentage compared with the control.

649 TE
C EP
AC

35
 ACCEPTED MANUSCRIPT

650 Table 7 Effects of N×K interaction on yield, fruit quality and soil available nutrient content of satsumas

PT
N K Yield Increase Quality Available nutrient content in soil

RI
SFW Increase TA Increase TSS Increase AN Increase AP Increase AK Increase
-1 -1 -1 -1 -1
kg plant kg plant % g % % % % % mg kg % mg kg % mg kg %

SC
0 0 43.6 72.2 0.19 11.3 78.4 51.9 214.4
0.3 49.1 12.5 61.7 -14.5 0.26 36.8 12.7 12.2 60.9 -22.3 29.6 -43.0 221.5 3.3
0.3 0.15 98.2 125.0 78.3 8.4 0.26 36.8 14.3 26.4 60.9 -22.3 47.2 -9.1 221.5 3.3

U
0.3 87.3 100.0 62.9 -12.9 0.27 42.1 14.2 25.9 51.5 -34.3 18.6 -64.2 221.5 3.3

AN
0.6 0 65.5 50.0 73.3 1.5 0.21 10.5 11.1 -1.9 65.0 -17.1 33.8 -34.9 202.8 -5.4
0.15 103.6 137.5 89.4 23.8 0.31 63.2 12.5 11.0 44.1 -43.8 25.0 -51.8 209.9 -2.1
0.3 120.0 175.0 88.7 22.9 0.34 78.9 12.8 13.2 38.2 -51.3 11.2 -78.4 191.3 -10.8

M
0.45 87.3 100.0 74.7 3.5 0.13 -31.6 12.9 14.7 58.8 -25.0 60.6 16.8 211.7 -1.3
0.9 0.3 46.9 7.5 68.6 -5.0 0.18 -5.3 13.1 16.5 50.1 -36.1 59.7 15.0 213.5 -0.4

D
Note: Positive signs denote an increased percentage compared with the control, while negative signs denote a decreased percentage compared with the control.

TE
C EP
AC

36
 ACCEPTED MANUSCRIPT

651 Table 8 Effects of P×K interaction on yield, fruit quality and soil available nutrient content of satsumas

PT
P K Yield Increase Quality Available nutrient content in soil

RI
SFW Increase TA Increase TSS increase AN Increase AP Increase AK Increase
-1 -1 -1 -1 -1
kg plant kg plant % g % % % % % mg kg % mg kg % mg kg %

SC
0 0 43.6 72.2 0.19 11.3 78.4 51.9 214.4
0 0.3 85.1 95.0 34.3 -52.5 0.15 -21.1 12.9 14.0 45.2 -42.3 24.8 -52.2 206.4 -3.7
0.1 0.15 109.1 150.0 61.0 -15.5 0.26 36.8 12.7 12.4 55.4 -29.3 54.2 4.4 218.8 2.1

U
0.30 87.3 100.0 64.4 -10.8 0.27 42.1 14.0 23.9 49.0 -37.5 13.0 -75.0 197.5 -7.9

AN
0.2 0 65.5 50.0 73.3 1.5 0.21 10.5 11.1 -1.9 65.0 -17.1 33.8 -34.9 202.8 -5.4
0.15 103.6 137.5 89.4 23.8 0.31 63.2 12.5 11.0 44.1 -43.8 25.0 -51.8 209.9 -2.1
0.30 120.0 175.0 88.7 22.9 0.34 78.9 12.8 13.2 38.2 -51.3 11.2 -78.4 191.3 -10.8

M
0.45 87.3 100.0 74.7 3.5 0.13 -31.6 12.9 14.7 58.8 -25.0 60.6 16.8 211.7 -1.3
0.3 0.30 98.2 125.0 78.8 9.1 0.11 -42.1 11.0 -2.5 62.0 -20.9 66.3 27.7 254.4 18.7

D
Note: Positive signs denote an increased percentage compared with the control, while negative signs denote a decreased percentage compared with the control.

TE
C EP
AC

37
 ACCEPTED MANUSCRIPT

652 Table 9 Evaluation of fertilization combinations based on yield, quality and environment

PT
Yield Fruit quality Environment Comprehensive
No. Treatment

RI
Score Rank Score Rank Score Rank Score Rank
1 N0P0.2K0.3 0.284 10 0.521 8 0.556 7 0.480 11

SC
2 N0.3P0.1K0.3 0.182 11 0.458 12 0.626 6 0.436 13
3 N0.3P0.2K0.15 0.663 4 0.770 1 0.437 11 0.682 2
4 N0.3P0.2K0.3 0.537 6 0.759 2 0.711 5 0.705 1

U
5 N0.6P0K0.3 0.511 9 0.534 7 0.756 3 0.574 6

AN
6 N0.6P0.1K0.15 0.790 2 0.507 9 0.442 10 0.551 7
7 N0.6P0.1K0.3 0.537 7 0.559 6 0.823 2 0.608 4

M
8 N0.6P0.2K0 0.727 3 0.399 14 0.534 8 0.492 9
9 N0.6P0.2K0.15 0.094 12 0.613 5 0.756 4 0.538 8

D
10 N0.6P0.2K0.3 0.916 1 0.507 10 0.942 1 0.676 3
11 N0.6P0.2K0.45 0.537 8 0.669 3 0.386 12 0.586 5

TE
12 N0.6P0.3K0.3 0.663 5 0.474 11 0.225 14 0.462 12
13 N0.9P0.2K0.3 0.068 13 0.642 4 0.463 9 0.491 10
EP
14 N0P0K0 0.030 14 0.446 13 0.271 13 0.328 14
C
AC

38
 ACCEPTED MANUSCRIPT

653 A B C D

654

655
F G H
E

PT
656

RI
657

SC
I J K L
658

U
AN
659

660
M

M N
661
D
TE

662
EP

663 Figure 1 Satsuma trees plant during harvesting under different fertilizer combinations. A: N0P0.2K0.3,
664 B:N0.3P0.1K0.3, C: N0.3P0.2K0.15, D: N0.3P0.2K0.3, E: N0.6P0K0.3, F: N0.6P0.1K0.15, G: N0.6P0.1K0.3, H:
665 N0.6P0.2K0, I: N0.6P0.2K0.15, J: N0.6P0.2K0.3, K: N0.6P0.2K0.45, L: N0.6P0.3K0.3, M: N0.9P0.2K0.3, N:
C

666 N0P0K0.
AC

667

668

669

39
 ACCEPTED MANUSCRIPT

Yield Comprehensive
150000
1,15,200 1,11,825
120000 1,10,880 1,05,510

90000

PT
60000

30000 3,840
5,370

RI
5,280 3,375
0
Yield Output Input Net
income

SC
670 (kg ha-1) (yuan ha-1) (yuan ha-1) (yuan ha-1)

U
671 Figure 2 The economic benefit of the best fertilization treatments based on yield and
AN
672 our comprehensive index
M

673
D

674
TE
C EP
AC

40
 ACCEPTED MANUSCRIPT

N P2O5 K2O
** * 200
Y 0.438 0.32 0.09 100
0
** 100
SFW 0.014 0.522 -0.13 50
0
0.5
TA -0.18 -0.09 -0.08 0.25
0

PT
* 20
TSS -0.01 -0.11 0.32 10
0
* 4

RI
LN 0.243 0.00 0.36 2
0
0.2
LP -0.12 -0.14 0.11 0.1

SC
0
* 2
LK 0.173 -0.32 0.06 1
0
** 0.4

U
FN 0.16 -0.41 -0.19 0.2
0
AN
0.04
FP -0.06 -0.08 0.21 0.02
0
0.2
FK -0.18 0.03 -0.05 0.1
M

0
** ** 100
AN -0.59 -0.15 -0.43 50
0
D

100
AP 0.05 0.22 -0.08 50
0
TE

* 400
AK -0.14 0.355 0.04 200
0
EP

675 0 0.5 1 0 0.2 0.4 0 0.5

676 Figure 3. Pearson correlation analysis between yield, fruit quality, soil and plant nutrient
C

content and fertilizer application. Y: yield, SFW: single fruit weight, TA: total acid of
AC

677

678 fruit, TSS: total soluble solids, LN: leaf N content, LP: leaf P content, LK: leaf K

679 content, FN: fruit N content, FP: fruit P content, FK: fruit K content, AN: soil available

680 N content, AP: soil available P content, AK: soil available K content. * and **:

681 Correlation is significant at the 0.05 or 0.01 level (2-tailed).

682

41