E ﬀ ects of Alternating Irrigation with Fresh and Saline Water on the Soil Salt, Soil Nutrients, and Yield of Tomatoes

: Saline water irrigation has become extremely important in arid and semi-arid areas in northwestern China. To study the e ﬀ ect of alternating irrigation models on the soil nutrients, soil salts, and yield of tomatoes with fresh water (total dissolved solids of 0.50 g · L − 1 ) and saline water (total dissolved solids of 3.01 g · L − 1 ), a two-year ﬁeld experiment was carried out for tomatoes in the Hetao Irrigation District (HID), containing six drip irrigation models: T1 (all freshwater irrigation), T2 (saline water used in the seedling and ﬂowering stages; fresh water in the fruit-set and breaker stages), T3 (saline water in the ﬂowering and fruit-set stages; fresh water in the seedling and breaker stages), T4 (saline water in the fruit-set and breaker stages; fresh water in the seedling and ﬂowering stages), T5 (saline water in the ﬂowering and breaker stages; fresh water in the seedling and fruit-set stages), T6 (saline water in the seedling and fruit-set stages; fresh water in the ﬂowering and breaker stages). The study found that saline water irrigation tends to have a positive e ﬀ ect on soil total nitrogen and a negative inﬂuence on soil total phosphorus at each growth stage of the tomato. Soil Na + , Mg 2 + , Ca 2 + , K + , and Cl − increased over the growth period, soil HCO 3 − decreased gradually by growth stage, and the salt ions increased with the amount of saline water applied in alternating irrigation. Though the soil salt accumulated in all experimentally designed alternating irrigation models, soil alkalization did not occur in the tomato root zone under the soil matric potential threshold of − 25 kPa. The utilization of saline water resulted in about a 1.9–18.2% decline in fruit yield, but the total soluble solids, lycopene, and sugar in the tomato fruits increased. Ultimately, drip irrigation with fresh water at the seedling to ﬂowering stages and saline water at the fruit-set to breaker stages was suggested for tomato cultivation in HID.


Introduction
Due to low rainfall and intense evaporation, the Hetao Irrigation District (HID) mainly relies on Yellow River water to meet its agricultural water demand. However, with the development of shortage was necessary [27,28]. Wang et al. (2007) conducted a field experiment about mulched drip irrigation with saline water at a conductivity level of 4.2 dS·m −1 , and found that there was no soil salt accumulation during the growth period of the tomato [29]. Karlberg et al. (2007) found that the fruit production of tomatoes with saline water drip irrigation was above the average yield [30]. Wan et al. (2008) have carried out an irrigation experiment with water containing different concentrations of salt for three consecutive years (2003)(2004)(2005); the results implied that saline water drip irrigation with a salt concentration of 1.1-4.9 dS·m −1 has little influence on the root length density, maximum leaf area index (LAI), total chlorophyll content, and fruit yield [31].
Though there are many studies on the impacts of alternating irrigation on the yield and quality of tomatoes, cotton, and lemons, and on the formation of clogging substances inside the emitters [32], few studies have been found to establish an alternating irrigation schedule based on growth stages under mulched drip irrigation. The current experiment was conducted to: (1) estimate the influence of different alternating irrigation models on the total nutrients and available nutrients; (2) explore the effect of different alternating irrigation models on soil salt ions and the sodium adsorption ratio (SAR); and (3) investigate soil salinization under the use of different water resources.

Experimental Conditions
The field experiment was carried out in 2017-2018 at the Jiuzhuang Experimental Station, in HID, China (longitude: 107 • 18 E, latitude: 40 • 41 N 1042 m a.s.l.). According to the meteorological data, which were collected from an automatic weather station (YM-03A, Handan Yimeng Electronics Co., Ltd., Hebei province, China) installed 50 m away from the field experiment site, the average annual temperature at the experimental area was 6.8 • C, and the area is characterized as having a mid-temperate semiarid continental climate with little precipitation (annual average of 140 mm) and large evaporation (annual average pan evaporation exceeding 2032-3179 mm) [33]. The total precipitation was 142 mm and 131 mm in 2017 and 2018, respectively, and mostly occurred from June to August, while the reference crop evaporation and precipitation during the growing period of the tomato are shown in Figure 1. atherosclerosis and some cardiovascular diseases [24]. Though the tomato is considered to be a medium salt-sensitive crop [25,26], supplementary irrigation with brackish water during freshwater shortage was necessary [27,28]. Wang et al. (2007) conducted a field experiment about mulched drip irrigation with saline water at a conductivity level of 4.2 dS·m −1 , and found that there was no soil salt accumulation during the growth period of the tomato [29]. Karlberg et al. (2007) found that the fruit production of tomatoes with saline water drip irrigation was above the average yield [30]. Wan et al. (2008) have carried out an irrigation experiment with water containing different concentrations of salt for three consecutive years (2003)(2004)(2005); the results implied that saline water drip irrigation with a salt concentration of 1.1-4.9 dS·m −1 has little influence on the root length density, maximum leaf area index (LAI), total chlorophyll content, and fruit yield [31].
Though there are many studies on the impacts of alternating irrigation on the yield and quality of tomatoes, cotton, and lemons, and on the formation of clogging substances inside the emitters [32], few studies have been found to establish an alternating irrigation schedule based on growth stages under mulched drip irrigation. The current experiment was conducted to: (1) estimate the influence of different alternating irrigation models on the total nutrients and available nutrients; (2) explore the effect of different alternating irrigation models on soil salt ions and the sodium adsorption ratio (SAR); and (3) investigate soil salinization under the use of different water resources.

Experimental Conditions
The field experiment was carried out in 2017-2018 at the Jiuzhuang Experimental Station, in HID, China (longitude: 107°18′ E, latitude: 40°41′ N 1042 m a.s.l.). According to the meteorological data, which were collected from an automatic weather station (YM-03A, Handan Yimeng Electronics Co., Ltd., Hebei province, China) installed 50 m away from the field experiment site, the average annual temperature at the experimental area was 6.8 °C, and the area is characterized as having a mid-temperate semiarid continental climate with little precipitation (annual average of 140 mm) and large evaporation (annual average pan evaporation exceeding 2032-3179 mm) [33]. The total precipitation was 142 mm and 131 mm in 2017 and 2018, respectively, and mostly occurred from June to August, while the reference crop evaporation and precipitation during the growing period of the tomato are shown in Figure 1.  physical properties, nutrient contents, and salt ions contents were also measured. According to the soil sample analysis, the soil 100 cm underneath the surface at the experiment field can be divided into three layers on the basis of an international soil texture classification system. The detailed physical properties and background values of soil nutrients and soil salt ions are shown in Tables 1-3, respectively.  The groundwater buried depth in the area varies from 1.85 m to 3.68 m below the soil surface during the growth period of the tomato, which is shown in Figure 2. The average groundwater depth during the tomato growth period was 2.510 m in 2017 and 2.46 m in 2018, respectively. The traditional flood irrigation water is mainly introduced from the Yellow River, which has an annual average salt concentration of 0.505 g·L −1 (electrical conductivity of 0.79 dS·m −1 ), while the average annual salinity of the local shallow groundwater is 3.006 g·L −1 (electrical conductivity of 4.70 dS·m −1 ) ( Figure 2). The main eight ion contents, pH, and TDS of shallow groundwater and Yellow River water are shown in Table 4.

Experimental Design
A local variety of tomato (Jinye No.1) was selected as the test crop in the current study; its growth can be divided into five stages according to the growth and reproductive characteristics of the tomato: seedling stage, flowering stage, fruit-set stage, breaker stage, and maturity stage. Based on the five growth stages, five alternating irrigation models with saline water (TDS of 3.0 g·L −1 ) and fresh water (TDS of 0.5 g·L −1 ) were built to a randomized complete block design. The tomatoes were planted in ridged raised beds (height 0.3 m × width 0.9 m × length 25 m) with plastic film over the roots (width 1.0 m × length 25 m) and drip tubes with 0.3 m emitter intervals were located in the center of each bed. Then 36 beds were divided into six plots for each treatment in one plot, and each plot occupied an area of 225 m 2 (width 7.9 m × length 25 m), as shown in Figures 3 and 4. The tomatoes were transplanted in two rows at a row spacing of 60 cm and plant spacing of 30 cm. The irrigated water amount was measured by a water meter installed at the inlet of the drip submain unit, which comprised six beds; each treatment was equipped with three vacuum gauge tensiometers installed 20 cm underneath the emitter for SMP monitoring, and the tensiometers were observed and the measurements recorded three times a day at 8:00, 14:00, and 18:00 during the growth period of the tomato.

Experimental Design
A local variety of tomato (Jinye No.1) was selected as the test crop in the current study; its growth can be divided into five stages according to the growth and reproductive characteristics of the tomato: seedling stage, flowering stage, fruit-set stage, breaker stage, and maturity stage. Based on the five growth stages, five alternating irrigation models with saline water (TDS of 3.0 g·L −1 ) and fresh water (TDS of 0.5 g·L −1 ) were built to a randomized complete block design. The tomatoes were planted in ridged raised beds (height 0.3 m × width 0.9 m × length 25 m) with plastic film over the roots (width 1.0 m × length 25 m) and drip tubes with 0.3 m emitter intervals were located in the center of each bed. Then 36 beds were divided into six plots for each treatment in one plot, and each plot occupied an area of 225 m 2 (width 7.9 m × length 25 m), as shown in Figures 3 and 4. The tomatoes were transplanted in two rows at a row spacing of 60 cm and plant spacing of 30 cm. The irrigated water amount was measured by a water meter installed at the inlet of the drip submain unit, which comprised six beds; each treatment was equipped with three vacuum gauge tensiometers installed 20 cm underneath the emitter for SMP monitoring, and the tensiometers were observed and the measurements recorded three times a day at 8:00, 14:00, and 18:00 during the growth period of the tomato.

Experimental Design
A local variety of tomato (Jinye No.1) was selected as the test crop in the current study; its growth can be divided into five stages according to the growth and reproductive characteristics of the tomato: seedling stage, flowering stage, fruit-set stage, breaker stage, and maturity stage. Based on the five growth stages, five alternating irrigation models with saline water (TDS of 3.0 g·L −1 ) and fresh water (TDS of 0.5 g·L −1 ) were built to a randomized complete block design. The tomatoes were planted in ridged raised beds (height 0.3 m × width 0.9 m × length 25 m) with plastic film over the roots (width 1.0 m × length 25 m) and drip tubes with 0.3 m emitter intervals were located in the center of each bed. Then 36 beds were divided into six plots for each treatment in one plot, and each plot occupied an area of 225 m 2 (width 7.9 m × length 25 m), as shown in Figures 3 and 4. The tomatoes were transplanted in two rows at a row spacing of 60 cm and plant spacing of 30 cm. The irrigated water amount was measured by a water meter installed at the inlet of the drip submain unit, which comprised six beds; each treatment was equipped with three vacuum gauge tensiometers installed 20 cm underneath the emitter for SMP monitoring, and the tensiometers were observed and the measurements recorded three times a day at 8:00, 14:00, and 18:00 during the growth period of the tomato.

Irrigation and Fertilization
According to the extensive investigation in 2016, traditional tomato cultivation is generally under film covering, while surface flood irrigation of 3600 m 3 ·ha −1 in total was applied during the tomato growth period; additionally, supplementary irrigation of 900 m 3 ·ha −1 was applied before transplanting to enhance the survival rate of the seedlings. Meanwhile, base fertilizers, including diammonium phosphate (DAP: 18% N, 46% P, 0% K) of 450 kg·ha −1 and potassium sulfate (including K2SO4, 45%) of 90 kg·ha −1 were uniformly applied to all plots before transplanting, with urea (46.2% N) of 180 kg·ha −1 given as a dressing fertilization.
The fertilization schedule of the present field experiment was based on the traditional fertilization schedule. Diammonium phosphate at a rate of 225 kg·ha −1 and potassium sulphate at a rate of 90 kg·ha −1 were applied as the base fertilization, and the dressing was supplied with urea (46% N) of 90 kg·ha −1 by mixing it with irrigation water at a concentration of 30% (w/w); the topdressing time and other agronomic practices were the same as in traditional tomato planting.
Many studies have indicated that the soil matric potential (SMP), which is measured 20 cm immediately underneath the drip emitter, can be used as an indicator for crop drip irrigation scheduling [34][35][36]. Kang et al. (2012) indicated that the SMP measured at 0.2 m underneath was recommended to be kept above −20 kPa when applying drip irrigation with saline water in Northwest China [37]. Fresh water with an amount of 900 m 3 ·ha −1 was applied for supplementary irrigation before transplanting to leach the topsoil salt. Then, after the tomato seedlings' grafting (13 May 2017 and 2018), irrigation was triggered by tensiometers buried 0.2 m underneath the emitters; as soon as the value of all tensiometers in each treatment was lower than −25 kPa, irrigation was implemented for each treatment with corresponding irrigation water resources (Table 5), and the same irrigation amount, 200 m 3 ·ha −1 . Additionally, no irrigation water was applied to tomatoes at the ripening stage to avoid rotting.

Irrigation and Fertilization
According to the extensive investigation in 2016, traditional tomato cultivation is generally under film covering, while surface flood irrigation of 3600 m 3 ·ha −1 in total was applied during the tomato growth period; additionally, supplementary irrigation of 900 m 3 ·ha −1 was applied before transplanting to enhance the survival rate of the seedlings. Meanwhile, base fertilizers, including diammonium phosphate (DAP: 18% N, 46% P, 0% K) of 450 kg·ha −1 and potassium sulfate (including K 2 SO 4 , 45%) of 90 kg·ha −1 were uniformly applied to all plots before transplanting, with urea (46.2% N) of 180 kg·ha −1 given as a dressing fertilization.
The fertilization schedule of the present field experiment was based on the traditional fertilization schedule. Diammonium phosphate at a rate of 225 kg·ha −1 and potassium sulphate at a rate of 90 kg·ha −1 were applied as the base fertilization, and the dressing was supplied with urea (46% N) of 90 kg·ha −1 by mixing it with irrigation water at a concentration of 30% (w/w); the topdressing time and other agronomic practices were the same as in traditional tomato planting.
Many studies have indicated that the soil matric potential (SMP), which is measured 20 cm immediately underneath the drip emitter, can be used as an indicator for crop drip irrigation scheduling [34][35][36]. Kang et al. (2012) indicated that the SMP measured at 0.2 m underneath was recommended to be kept above −20 kPa when applying drip irrigation with saline water in Northwest China [37]. Fresh water with an amount of 900 m 3 ·ha −1 was applied for supplementary irrigation before transplanting to leach the topsoil salt. Then, after the tomato seedlings' grafting (13 May 2017 and 2018), irrigation was triggered by tensiometers buried 0.2 m underneath the emitters; as soon as the value of all tensiometers in each treatment was lower than −25 kPa, irrigation was implemented for each treatment with corresponding irrigation water resources (Table 5), and the same irrigation amount, 200 m 3 ·ha −1 . Additionally, no irrigation water was applied to tomatoes at the ripening stage to avoid rotting.  T1  FW  FW  FW  FW  T2  SW  SW  FW  FW  T3  FW  SW  SW  FW  T4  FW  FW  SW  SW  T5  FW  SW  FW  SW  T6 SW FW SW FW Note: "FW" is irrigation with fresh water (Yellow River water, TDS of 0.5 g·L −1 ), while "SW" is irrigation with prepared saline water (shallow ground saline water, TDS of 3.0 g·L −1 ).

Soil Salinity
Soil samples were taken during each growth period of the tomato at depths of 0-40 cm and 40-100 cm in the two locations, 0 cm (inside the film) and 75 cm (outside the film) from the drip tube, for the measurement of soil salinity. The soil samples were naturally air-dried and ground well to pass through a 1-mm sieve; then the soil leachates were prepared at a soil to water ratio of 1:5, EC 1:5 was measured by a conductivity meter (FE30, METTLER TOLEDO, Shanghai, China), and the pH value was determined by a pH meter (FE20, METTLER TOLEDO). Soil Ca 2+ and Mg 2+ were determined by an atomic absorption spectrophotometer (Varian spectra AA55, Varian, Palo Alto, California, USA), while soil K + and Na + were measured by a flame emission spectrophotometer (Model 410 Flame photometer, Sherwood Scientific, Ltd., New York, USA) [38]. Soil Cl − was measured by the method of silver nitrate titration, while soil SO 4 2− was measured by the EDTA titration method; meanwhile, the soil HCO 3 − was determined by the neutralization titration method.
SAR is usually selected as the index for measuring the level of soil salinization that results from irrigation models, and the SAR for soil samples was calculated using the following equation: where c(Na + ), c(Ca 2+ ), and c(Mg 2+ ) are the content of Na + , Ca 2+ , and Mg 2+ , which are expressed in meq·L −1 .
The soil salinity of each soil layer was estimated using the soil salt content of samples both 0 cm and 75 cm away from the dripper. Soil salt concentration was estimated by electrical conductivity EC 1:5 via the following equation [39]: where Y is the mass salt content (%) and EC 1:5 is the soil electrical conductivity (dS·m −1 ).

Soil Nutrient Content
The soil samples that were taken before transplanting and during each growth period were used for measuring the total nitrogen and total phosphorus. While the total nitrogen was measured by the Kjeldahl distillation method [40], the total phosphorus was determined following dry combustion at 550 • C for 2 h and extraction with 0.5 M H 2 SO 4 [41].

Fruit Yield and Quality
Tomato fruits were picked twice a week during the harvest seasons for each plot, and all fruits were classified as unmarketable (fruits that were cracked, green, sunburnt, with symptoms of blossom-end rot, or damaged by pests) or marketable ones; the fruit yield was measured for the marketable fruits of each plot. The yield (g/plant) of each plant was determined during the harvest season.
For each treatment, 20 marketable tomato fruits were collected randomly for the measurement of total soluble solids (T SS ), lycopene, total sugar (T s ), and total acid (T a ). T SS was measured by an ACT-1E digital refractometer (ATAGO Co., Ltd., Tokyo, Japan), and the content of lycopene was measured using the spectrophotometry method. T s was determined by the Fehling reagent titration method, and T a was measured by the sodium hydroxide titration method.

Statistical Analysis
The soil nutrient, soil ions, and soil salinity were analyzed with Excel 2016 (Microsoft Corporation, Redmond, Washington, USA), while the layout of the drip lines was drawn by AutoCAD 2018 (Autodesk Inc., San Rafael, California, USA).

Irrigation
Due to the variable sensitivity of tomato roots to soil moisture and soil salinity at different growth stages, the amount of water consumed and the irrigation frequency were usually distinct between models at different growth periods. Details on the amount of water applied according to the design in 2017 and 2018 are shown in Figure 5. For each treatment, 20 marketable tomato fruits were collected randomly for the measurement of total soluble solids (TSS), lycopene, total sugar (Ts), and total acid (Ta). TSS was measured by an ACT-1E digital refractometer (ATAGO Co., Ltd., Tokyo, Japan), and the content of lycopene was measured using the spectrophotometry method. Ts was determined by the Fehling reagent titration method, and Ta was measured by the sodium hydroxide titration method.

Statistical Analysis
The soil nutrient, soil ions, and soil salinity were analyzed with Excel 2016 (Microsoft Corporation, Redmond, Washington, USA), while the layout of the drip lines was drawn by AutoCAD 2018 (Autodesk Inc., San Rafael, California, USA).

Irrigation
Due to the variable sensitivity of tomato roots to soil moisture and soil salinity at different growth stages, the amount of water consumed and the irrigation frequency were usually distinct between models at different growth periods. Details on the amount of water applied according to the design in 2017 and 2018 are shown in Figure 5. As detailed in Figure 5, the irrigation quota for all treatments at the fruit-set and breaker stages was significantly higher than that at the seeding and flowering stages. The saline water used for the T4 treatment was 160 mm in 2017 and 180 mm in 2018, respectively, which was significantly higher than the other treatments, while the fresh water used for the T4 treatment was the lowest. The irrigation quota for both saline and fresh water at the flowering stage was 20 mm, while at the same growth stage, the irrigation quota for brackish water was no less than that for fresh water. Moreover, the irrigation quota for the T2 treatment was lower than in other alternating fresh-saline water irrigation models; in particular, the irrigation quota for the T2 treatment was lower than that for freshwater irrigation. Additionally, continuous irrigation with saline water at the seeding and flowering stages resulted in a decrease in the irrigation quota at the fruit-set and breaker stages.

Total Nutrients
Soil nutrients were essential for the vegetative and reproductive growth of the plant. The variation of soil nutrients in the main root zone (0-40 cm depth inside the film) at different tomato growth periods was measured and shown in Figure 6. As detailed in Figure 5, the irrigation quota for all treatments at the fruit-set and breaker stages was significantly higher than that at the seeding and flowering stages. The saline water used for the T4 treatment was 160 mm in 2017 and 180 mm in 2018, respectively, which was significantly higher than the other treatments, while the fresh water used for the T4 treatment was the lowest. The irrigation quota for both saline and fresh water at the flowering stage was 20 mm, while at the same growth stage, the irrigation quota for brackish water was no less than that for fresh water. Moreover, the irrigation quota for the T2 treatment was lower than in other alternating fresh-saline water irrigation models; in particular, the irrigation quota for the T2 treatment was lower than that for freshwater irrigation. Additionally, continuous irrigation with saline water at the seeding and flowering stages resulted in a decrease in the irrigation quota at the fruit-set and breaker stages.

Total Nutrients
Soil nutrients were essential for the vegetative and reproductive growth of the plant. The variation of soil nutrients in the main root zone (0-40 cm depth inside the film) at different tomato growth periods was measured and shown in Figure 6. As shown in Figure 6, the total nitrogen content in the root zone decreased gradually along the tomato growth period, and reached a minimum at the maturity stage. Obviously, the value of soil total nitrogen was raised for saline water irrigation from the seedling stage to the fruit-set stage, so the impact of saline water irrigation on soil nitrogen content mainly occurred before the breaker stage.
Similarly, the total phosphorus content in the root zone decreased along the tomato growth period; in particular, the total phosphorus decreased sharply at the seedling period. The soil total phosphate content for the T1 treatment was higher than for the other treatments along the growth period; the content of soil total phosphate for the T2 treatment took second place, while the value of soil total phosphate concentration for the T4 treatment was the lowest.

Salt Ions
In general, there are eight main ions considered in the soil: HCO3 − , CO3 2− , SO4 2− , Cl − , Ca 2+ , Mg 2+ , K + , and Na + . Since CO3 2− is not stable and the content in the soil usually close to 0, it was not considered in this paper; the other seven main ions' contents in treatments at different growth periods are shown in Figure 7.
As we can see from Figure 7a,c, the content of Na + , K + , Mg 2+ , and Ca 2+ for treatments in 0-40 cm topsoil increased with growth, and the contents were higher for saline water irrigation than freshwater irrigation at the same growth stage. Na + had the most obvious soil-soluble cations that were influenced by saline water. The higher the brackish water irrigation quota, the more Na + , K + , Ca 2+ and Mg 2+ was brought into the topsoil.
As detailed in Figure 7 b,d, the soil HCO3 − content reached its maximum at the seedling and flowering stages, while the content for treatments during the growth period had little difference. However, the content of soil Cl − increased gradually over the tomato growth period; in particular, soil Cl − content for saline water irrigation was higher than that for fresh water. Moreover, the content of SO4 2− for treatments in 40 cm topsoil first decreased and then increased slowly over the growth period of the tomato, and reached a minimum at the breaker stage; the soil SO4 − tended to increase with the amount of saline water applied at the same stage. As shown in Figure 6, the total nitrogen content in the root zone decreased gradually along the tomato growth period, and reached a minimum at the maturity stage. Obviously, the value of soil total nitrogen was raised for saline water irrigation from the seedling stage to the fruit-set stage, so the impact of saline water irrigation on soil nitrogen content mainly occurred before the breaker stage.
Similarly, the total phosphorus content in the root zone decreased along the tomato growth period; in particular, the total phosphorus decreased sharply at the seedling period. The soil total phosphate content for the T1 treatment was higher than for the other treatments along the growth period; the content of soil total phosphate for the T2 treatment took second place, while the value of soil total phosphate concentration for the T4 treatment was the lowest.

Salt Ions
In general, there are eight main ions considered in the soil: HCO 3 − , CO 3 2− , SO 4 2− , Cl − , Ca 2+ , Mg 2+ , K + , and Na + . Since CO 3 2− is not stable and the content in the soil usually close to 0, it was not considered in this paper; the other seven main ions' contents in treatments at different growth periods are shown in Figure 7.
As we can see from Figure 7a,c, the content of Na + , K + , Mg 2+ , and Ca 2+ for treatments in 0-40 cm topsoil increased with growth, and the contents were higher for saline water irrigation than freshwater irrigation at the same growth stage. Na + had the most obvious soil-soluble cations that were influenced by saline water. The higher the brackish water irrigation quota, the more Na + , K + , Ca 2+ and Mg 2+ was brought into the topsoil.
As detailed in Figure 7b,d, the soil HCO 3 − content reached its maximum at the seedling and flowering stages, while the content for treatments during the growth period had little difference. However, the content of soil Cl − increased gradually over the tomato growth period; in particular, soil Cl − content for saline water irrigation was higher than that for fresh water. Moreover, the content of SO 4 2− for treatments in 40 cm topsoil first decreased and then increased slowly over the growth period of the tomato, and reached a minimum at the breaker stage; the soil SO 4 − tended to increase with the amount of saline water applied at the same stage.

SAR
To make a further analysis of the effect of alternating irrigation with saline water and fresh water, the soil SAR was measured by soil ions content and shown in Error! Reference source not found.. Note: Values in a column in the same year followed by the same letter are not significantly different at p ≤ 0.05.
As shown in Error! Reference source not found., the soil SAR of the root zone was lower than 10 (meq/L) 1/2 , even though the soil SAR increased due to brackish water irrigation, and the soil structure for all treatments with brackish water irrigation was hardly affected; moreover, the soil SAR increased gradually and reached a maximum at the breaker stage, then decreased. As the present paper only focused on the soil SAR in the root zone, further studies on the variation of SAR outside the film are required.

Total Salt
To determine the influence of saline water irrigation on soil salinity, the total salt content of the two soil layers (0-40 cm and 40-100 cm) inside and outside the film was measured after harvest; the results are shown in Figure 8.

SAR
To make a further analysis of the effect of alternating irrigation with saline water and fresh water, the soil SAR was measured by soil ions content and shown in Table 6. Note: Values in a column in the same year followed by the same letter are not significantly different at p ≤ 0.05.
As shown in Table 6, the soil SAR of the root zone was lower than 10 (meq/L) 1/2 , even though the soil SAR increased due to brackish water irrigation, and the soil structure for all treatments with brackish water irrigation was hardly affected; moreover, the soil SAR increased gradually and reached a maximum at the breaker stage, then decreased. As the present paper only focused on the soil SAR in the root zone, further studies on the variation of SAR outside the film are required.

Total Salt
To determine the influence of saline water irrigation on soil salinity, the total salt content of the two soil layers (0-40 cm and 40-100 cm) inside and outside the film was measured after harvest; the results are shown in Figure 8.  After tomato harvesting, the soil salt content of the upper layers both inside and outside the mulch for the T1 and T4 treatments attained the minimum and maximum values, respectively. Meanwhile, the soil salt content at 40-100 cm inside the film for the T1 and T4 treatments achieved the minimum and maximum, respectively; however, soil salt at 40-100 cm outside the mulch showed little difference between treatments.

Tomato Production
Compared with freshwater irrigation, saline water irrigation has dual effects on the plant: on the one hand, more salt was carried into the topsoil, which was not conducive to root water uptake; on the other hand, some salt ions of brackish water are beneficial for the vegetative and reproductive growth of the tomato. The fruit yield of the tomato is listed in Error! Reference source not found.. After tomato harvesting, the soil salt content of the upper layers both inside and outside the mulch for the T1 and T4 treatments attained the minimum and maximum values, respectively. Meanwhile, the soil salt content at 40-100 cm inside the film for the T1 and T4 treatments achieved the minimum and maximum, respectively; however, soil salt at 40-100 cm outside the mulch showed little difference between treatments.

Tomato Production
Compared with freshwater irrigation, saline water irrigation has dual effects on the plant: on the one hand, more salt was carried into the topsoil, which was not conducive to root water uptake; on the other hand, some salt ions of brackish water are beneficial for the vegetative and reproductive growth of the tomato. The fruit yield of the tomato is listed in Table 7. According to Table 7, the tomato production for treatments was in the order of T1 > T4 > T3 > T5 > T6 > T2, and the fruit yield on each plant for the T4 treatment was the highest among alternating irrigation models in both 2017 and 2018. Compared with the T1 treatment, the tomato production for the T4 treatment decreased by just 2.75% in 2017 and 1.90% in 2018, respectively, while the tomato production for the T2 treatment decreased by 18.2% in 2017 and 15.15% in 2018, respectively.

Fruit Quality
To estimate the fruit quality, the common parameters (T SS , lycopene, T s , and T a ) were measured after tomato harvesting, and the results are shown in Table 8.
stages. The results indicated that saline water irrigation may contribute to a certain increase in soil nitrogen and a certain decrease in soil phosphorus, which was in line with the study of Tian et al. (2018), who has studied the effects of soil nutrients under saline water irrigation in 2016 [42]. The ions of saline water accumulated in the topsoil and then affected the soil nitrogen contents, especially at the seedling and flowering stages. Taking account of the nitrogen in the topsoil, saline water irrigation was recommended before the breaker stage (T2, T3, T6 treatment), because of an increase in the soil nitrogen content. The content of phosphate in the topsoil decreased due to saline water irrigation, which was in line with the study of Tian et al. (2018) [42]. As reported by Jin et al. (2014), there are two possible reasons for the influence of saline water irrigation on soil total phosphate content: on the one hand, the soil ion content tends to increase with the amount of saline water applied, which results in more dissociated phosphorus being adsorbed by soil ions; on the other hand, irrigation with brackish water may decrease the soil phosphorus content by reducing the activity of soil microorganisms such as ammonifying bacteria and the bacteria that decompose organic phosphorus and inorganic phosphorus [43].
Soil Ca 2+ and Mg 2+ increased over the growth period, and the content for saline water irrigation was higher than that for freshwater irrigation at the same growth stage, which may arise from the strong adsorption capacity of Ca 2+ and Mg 2+ with soil colloids, and the rich content of irrigation water. As soil Mg 2+ accumulation used to have a deleterious effect on the aggregate soil stability, such as via high swelling by expanding clays, extra measures such as soil amendment, supplementary irrigation, etc., should be taken to decrease the Mg 2+ content of the topsoil. Soil Na + and K + increased over the growth period of the tomato, and the content of topsoil Na + and K + increased especially for the saline water irrigation, which was in accordance with the study of Karlberg and de Vries (2004), who reported that with saline water irrigation, sodium ions (Na + ) tend to accumulate and concentrate in the topsoil due to the soil evaporation [44]. With the characteristics of low ionic charge, small hydration, and large radius, soil K + and Na + usually have a weak adsorption capacity with soil colloids, [45,46]; accompanied by high background values of K + and Na + in the topsoil and the high concentration of K + and Na + in the saline water, K + and Na + tend to accumulate in the topsoil. The accumulation of Na + usually results in dispersion and expansion of soil particles, such as plugging of pores owing to clay particles [45], which may result in the destruction of soil aggregates, a reduction in soil hydraulic conductivity, and a decrease in available water capacity [47][48][49]; moreover, the accumulation of Na + may lead to negative effects on the soil chemical characteristics, such as a reduction of available nitrogen (N), phosphorus (P), and potassium (K) [50,51].
Soil HCO 3 − decreased over the growth period, which indicated that the topsoil gradually transitions from alkaline to neutral in the course of the cultivation. The content of Cl − increased over the tomato growth period, which may be ascribed to the weak adsorption capacity of Cl − to soil colloids, and the high level of Cl − in saline water. The content of SO 4 2− first decreased and then increased, which may be a result of the base fertilizer of K 2 SO 4 and the higher content of SO 4 2− in saline water than in fresh water. Plant tissues usually accumulate Na + and Cl − , which are induced by saline water, which will cause osmotic stress, ion toxicity, and ion imbalance in tissues [52]; the uptake of the nutrients can be interrupted by competitive membrane selectivity or interaction [51]. Moreover, Na + toxicity can lead to K + deficiency, which will result in water loss and necrosis [53]. Additionally, the stomatal conductance and photosynthesis rate of plants may decrease due to the decline in carbon dioxide (CO 2 ) availability that results from Na + toxicity [54]. Moreover, the photochemical system or metabolism may be interfered with [51]. Thus, saline water irrigation at the seedling stage of tomatoes is not suggested.
The soil SAR of the root zone was less than 10 (meq/L) 1/2 , which indicated that an irrigation schedule alternating brackish water and fresh water will not result in soil alkalization in the root zone; however, the application of saline water will cause an increase in the SAR value in the root zone at each growth stage, which is in line with the high content of salt ions contained in brackish water. To avoid topsoil alkalization, saline water irrigation at the seedling and breaker stages was not suggested; however, there is supplementary irrigation after the harvest every year with a large amount of water (90 mm) for leaching the soil salt, so saline water irrigation with a proper schedule is practicable. However, close attention should be paid to the damage soil sodium salt may do to the soil environment. After the tomato harvest, the total content of soil salt at 40-100 cm inside the film and surface layer outside the mulch appeared to be relatively high, and the total salt content increased with the irrigation quota of brackish water; moreover, there was little difference from the soil salt content at 40-100 cm outside the film, which arises from leaching and evaporation. To reduce the accumulated soil salt, an alternating irrigation schedule of the T2 treatment is recommended. Compared with fresh water, there are more soluble salts in saline water, so saline water irrigation not only dilutes the soil solution by increasing the topsoil moisture, but also carries dissolved salt into the topsoil inside the mulch. Additionally, soil salt in the root zone migrates into the deeper layers inside the film and the soil layers outside the film. Moreover, the higher the irrigation quota, the more sufficient the leaching, and the soil salt tends to migrate to the margin of the wetting body, which results in a higher content of soil salt in the upper layers outside the film. Generally, drip irrigation usually supplies water to the topsoil frequently, so soil salt is sufficiently leached; moreover, some salt moves to the fringes of the root zone along with soil water, so the desired desalination circumstance was formed near the emitter for better plant growth [55]. It is evident from the results that the more saline water is applied, the more soil salt accumulates; compared with freshwater irrigation, alternating saline water irrigation tends to promote soil salt accumulation, which corresponds to a higher dissolved salt content in brackish water. To reduce the accumulated soil salt, the alternating irrigation schedule of the T2 treatment is recommended.
Additionally, compared with freshwater irrigation, the tomato fruit yield for alternating fresh-saline water irrigation models decreased by 2.75-18.20% in 2017 and 1.90-15.15% in 2018, respectively, which indicated that alternating fresh-saline water irrigation will result in reduced tomato production; however, with proper irrigation models, this reduction can be controlled to a certain extent. The T SS , lycopene, and T S of tomato fruit for the T2, T3, T4, T5, and T6 treatments were higher than for the T1 treatment, while the T a for freshwater irrigation was higher than for alternating fresh-saline water irrigation, which indicated that trace elements of brackish water are beneficial in terms of the increase in T SS , lycopene, and T S , a result identical to the study of Abdel Gawad, who reported that the soluble solid content of tomato fruits rose when exposed to a certain amount of brackish water [56].

Conclusions
The alternating fresh-saline water irrigation model for the T4 treatment (fresh water in the seedling and flowering stages, saline water in the fruit-set stage breaker stages) was more freshwater-saving than other treatments, and was more suitable for high tomato production and improvement of fruit quality. According to the effects of alternating irrigation models on topsoil nutrient and the SAR of the topsoil, the alternating irrigation model for the T3, T4, and T5 treatments was better than other treatments. Ultimately, the alternating fresh-saline water model of the T4 treatment was suggested to be the proper irrigation style for the utilization of shallow ground saline water in HID.