Growth, yield and fruit quality of cherry tomato irrigated with saline water at different developmental stages

ABSTRACT

Water shortage is a serious environmental and agricultural problem and saline underground water has been widely used to make up the fresh water shortage in northwestern China. An open-field experiment was conducted to establish a proper irrigation scheme with saline water for cherry tomato in the Minqin oasis, where very severe salinization occurs. The experiment had four treatments including fresh or saline irrigation over the crop season (control, C, T3), fresh/saline-water irrigation change on days after thinning 50 (DAT 50, T1) and saline/fresh irrigation change on DAT 50 (T2). Leaf area index (LAI), photosynthesis rate (P_n), transpiration rate (T_r), leaf dry matter (LDM), stem dry matter (StDM), yield, marketable fruit and total soluble solids (TSS) of tomato were measured. Saline irrigation, irrespective of the timing, significantly decreased maximum LAI, LDM and StDM, P_n, T_r and stomatal conductance but significantly stimulated water use efficiency. The reduction in maximum LAI, LDM and StDM was lower in T2 than in T1 and T3. Harvest index (HI) and TSS were higher in T2 and T3 than in T1 and C. Marketable fruit had no significant change in T2 but significantly declined in T1 and T3. Maximum saturated soil conductivity without yield reduction (the salt tolerance threshold) was 3.69 dS m⁻¹. Total yield of tomato would decrease by 9.85% with one unit increase of soil salinityhigher than the threshold. Final yield significantly reduced by 24.6% and 23.1% in T1 and T3 treatments, respectively. Our results suggest that irrigation with saline water before DAT 50 and fresh water after DAT 50 should be advocated for cherry tomato plantation in water-scarce areas like the Minqin oasis.

KEYWORDS:

• Cherry tomato
• dry matter
• saline water
• soil salinity
• salt tolerance
• yield

Introduction

Ground water shortage is a severe environmental and agricultural problem in arid regions worldwide (Bennett 2000). The increasing demand for irrigation water to ensure food security for booming population in arid land necessitates the use of saline water in agricultural activities (Malash et al. 2008). Saline irrigation-induced higher osmotic pressure generally results in the decrease in growth rate, yield, dry matter (Ismail & Burrage 1994), net CO₂ assimilation rate, transpiration and stomatal conductance (Romero-Aranda et al. 2001), but led to fruit quality improvement (del Amor et al. 2001; Huang et al., 2012; Maggio et al. 2004) and fruit preservability (Tedeschi et al. 2011).

Cherry tomato (Lycopersiconesculentum), a variety of tomato that has been cultivated since at least the early 1800s, is planted in China, United States as well as Southern Europe as a cash crop to raise famers' revenue (Mori et al. 2008). Plantation of cherry tomato started from1996 and boomed in 2005 in the Minqin oasis. In addition to the low, irregular annual rainfall (110.7 mm) and very high potential evaporation (2593 mm, Xue et al. 2015), establishment of a reservoir in the upper part of the oasis in 1950s–1960s and misuse of ground water had led to the disappearance of lake and severe water shortage in the downstream part of the oasis (Feng et al. 2005). Thus, underground water use is indispensable to maintain agricultural activities and to ensure the subsistence of local farmers. However, long-term overuse of underground water has resulted in serious environmental and ecological problems, such as intense mineralization of water quality, lowering underground water table and serious salinization (Figure S1). For example, salt concentration of underground water increases from 0.8 g L⁻¹ in the upper part to 5 g L⁻¹ in the downstream part of the oasis (Feng et al. 2005). In local agricultural practice, fresh water is used to leach the soil in the beginning of growing season and fresh water and saline groundwater are alternatively used in the remaining period of the growing season in the oasis (You et al. 2011). Several studies found the timing of saline irrigation plays an important role in partially alleviating the negative effect of saline irrigation on production and keeping the positive effect on product quality (Heenan et al. 1988; Mass & Poss 1989; del Amor et al. 2001; Läuchli & Grattan 2007; Moradi & Ismail 2007; Wan et al. 2007; Bao & Li 2010). Salt tolerance of tomato plants increased when the application of saline irrigation was delayed with 4, 6 and 8 dS m⁻¹ saline irrigation (del Amor et al. 2001), and reproductive growth stage is less sensitive than vegetative growth stage (Läuchli & Grattan 2007). Flowering and first fruit cluster bearing period was considered as the most sensitive stage for tomato yield (Bao & Li 2010). However, saline irrigation (1.4–4.9 dS m⁻¹) had little effect on tomato yield when saline water used from 30 days after transplanting (Wan et al. 2007). We therefore hypothesized a low yield reduction but improvement in fruit quality when saline irrigation (7.03 dS m⁻¹) applied in the reproductive growth stage in the oasis.

Materials and methods

Location and environmental background

The experiment was conducted at the Gansu Desert Control Station (38^o35′N, 103^o03′E), located in the upper part of the Minqin oasis in 2010. Sand and silt account for 90% of the soil, bulk density of soil in the study site was 1.55 g cm⁻³and pH was 8.3 in the 0–60 cm layer (Huang et al. 2011, 2012). Mean annual temperature is 7.6°C and mean rainfall was 110.7 mm (1990–2010 average). Sowing generally started in April and harvesting ended in October in the oasis. Rainfall in the crop season was about 90% of the annual amount. Annual average reference evapotranspiration (ETc) was 995 mm (1998–2008) based on the Penman-Monteith method with a crop reference coefficient of 0.408.

Irrigation practices

Flood irrigation with fresh water (ECw = 0.7 dS m⁻¹) was conducted in spring 2010 before laying the polyethylene filmmulch and drip irrigation facilities. Tomato cultivar zhongzhiNO.1, a determinate type, was widely used by local famers as a cash crop. The seeds from a local seed store were cultivated in a greenhouse in May 2010. Seedlings were transplanted to the field when they were about 15 cm in height in a double row bed. The distance between twin rows was 0.5 m and the distance across the double row beds was 0.7 m. In each row, holes about 9 cm in diameter and 15 cm in depth, spaced 0.4 m were drilled and leached with 300 ml fresh water one day before transplantation. Two seedlings were placed to each hole on 5 June 2010, and were thinned to 4.16 plants per m² land area on 22 June 2010. The row beds and part of the ridge was covered by low-densitywhite polyethylene film. Experimental area was treated with drip irrigation to facilitate the distribution of the saline solution. Drippers at 3 L h⁻¹ was laid at a high density per m² (0.4 × 0.5 m) in the twin rows. This design ensured a supply of 15 L h⁻¹ m⁻² and completely wet of the soil surface. Water amount for each irrigation was adjusted according to the daily ETc by water meters.

The first irrigation was conducted on 6 June 2010 (0 day after thinning, DAT). The crop phenology of cherry tomato and irrigation arrangement was shown in Table 1.

Table 1. Phenology and irrigation calendar for the cherry tomato.

Phenology stages	Irrigation water amount (mm)
Thinning, DAT 0	DAT	Quantity
	0	15
	7	15.6
	14	14.5
Beginning flowering, DAT 28	18	6.5
	21	6.9
	25	6.0
	29	8.7
First harvest, DAT 39	32	7.3
	35	10.1
	39	4.7
	43	8.1
Peak fruit development, DAT 59	48	6.0
	50	7.6
	54	6.0
	62	8.1
	65	6.0
	69	9.1
Maximum harvest, DAT 74	73	9.4
	80	9.4
	86	8.1
	94	8.1
Final harvest and end of the crop season, DAT 100	101	8.1
	Total irrigation	181

Note: DAT means days after thinning.

Experimental design

A random experimental of four treatments with three replicates in each treatment was used in this study. The four treatments included fresh (C) or saline irrigation (T3) all over the crop season, fresh/saline water changing on DAT 50 (T1) and saline/fresh water changing on DAT 50 (T2). Salt concentration of the saline water was 5 gL⁻¹ to reproduce the salt concentration of groundwater in the oasis, which was made up from chemical compounds and fresh water (Table 2). Each plot was 9 m × 6 m and there were 225 plants in each plot.

Table 2. Chemical composition of fresh and saline irrigation water for the experiment.

Ions (mg L^−1)	Fresh water	Saline water
Cl^−	32	890
SO4^2−	152	1640
HCO3^−	248	360
CO3^2−	0	0
Ca^2+	64	280
Mg^2+	33	190
K^+	1	1000^a
Na^+	35
TDS	0.6	5.0
pH	7.4	7.5
Ec (dS m^−1)	0.70	7.03

^aThe value reported in table for Na⁺ and K⁺ for saline water is the sum of the two cations.

Sample collection and measurement protocol

Soil and plant samples were collected in the central part of each plot (inner 3 m × 2 m area). Soil samples were taken on DAT 1, 8, 26, 41, 55, 70, 87 and 102 at the depth of 0–0.2 m and 0.2–0.4 m. One sample was collected in each plot on every sampling day. Soil salinity of the saturated paste (ECe) was determined using the method described by Rhoades (1996). ECe of 20 randomly selected soil samples for each layer were determined using EC_1:2.5 and a linear regression proposed by Huang et al. (2011). Soil chemical characteristics in the study site could be seen in Huang et al. (2012).

Leaf area, fresh weight of leaves and stems of three plants in each plot were measured on DAT 6, 20, 34, 47, 59, 74 and 88. Leaf area was monitored using an area meter (LI-3100C, Li-COR, Lincoln, NE, USA). The fresh leaves and stems were oven-dried at 75^oC for 72 h to obtain the leaf and stem dry matter (LDM and StDM). The data collected to determine the leaf area duration (LAD) was in agreement with Hunt (1982). Leaf weight (LWR) and stem weight ratio (StWR) was defined as the ratio of LDM and StDM to whole plant dry matter. Fruits were harvested on DAT 39, 57, 74, 93 at marketable maturity. The final yield was the sum of the four harvests.

Nine randomly selected fruits in each plot were cut into slices and flesh total soluble solids (TSS) were determined as the ^oBrix index using a portable refractometer. Average TSS of the nine fruits was used for statistical analysis.

Net photosynthesis rate (P_n), stomatal conductance (g_s) and transpiration rate (T_r) were measured at 11 am on DAT 25 and DAT 53 by using Li-Cor 6400 × t (LI-COR, Lincoln, NE, USA). Daily water flux by T_r was calculated by multiplying T_r, time and leaf area. The fifth leaves from the top-down were used for the measurement. The measurement was conducted on three plants in each plot and was repeated for three times for each seedling. Water use efficiency (WUE) was the ratio of P_n to T_r (Fischer & Turner 1978).

Statistical analysis

One-way ANOVA was used to evaluate the effect of treatments on soil salinity on each sampling day. Two-way ANOVA was employed to test the interaction effect of treatment and sampling time on soil salinity at the depth of 0–0.2 m and 0.2–0.4 m, P_n, g_s, T_r, WUE, LAD and yield. Tukey analysis was used to do the multiple-comparison.

Results

Soil salinity

During the crop season, average ECe in 0–0.4 m was generally higher in T3 and T2 than in T1 and C (Figure 1(a)). ECe showed different seasonal variation among treatments. Fresh irrigation tended to decrease ECe (lower ECe in C and T1 from DAT 0 to DAT 40, and in C and T2 from DAT 50 to DAT 88) and saline irrigation increased ECe (higher ECe in T2 and T3 from DAT 0 to DAT 40, and in T1 and T3 from DAT 50 to DAT 88, Figure 1(a)). ECe increased from DAT 40 to DAT 50 in all treatments (Figure 1(a)). Average ECe in 0–0.4 m significantly increased by 1.93 and 2.91 dS m⁻¹ in the end of growing season (DAT 102) in T1 and T3 treatments, respectively (Figure 1(b)). Vertical distribution of ECe was different among treatments, with higher ECe in 0–0.2 m in T2 but lower ECe in 0–0.2 m in C and T1 (Figure 1(c)).

Figure 1. Seasonal variation in soil salinity at the depth of 0–0.4 m ((a), electrical conductivity in saturated paste, ECe, dS m⁻¹), changes in ECe between the beginning of crop season and the end of harvest at 0–0.4 m depth (b), seasonal average ECe at the 0–0.2 and 0.2–0.4 m depth (c) in fresh water irrigation (C), saline water irrigation after DAT 50 (T1), saline water irrigation before DAT 50 (T2) and saline water irrigation (T3). Values followed by the different small letters indicate the significant difference of ECe at p < .05. DAT means days after thinning.

Net photosynthesis, stomatal conductance, transpiration and WUE

Interaction between treatments and measuring time had no significant effect on photosynthesis (P_n), stomatal conductance (g_s), transpiration (T_r) and WUE. P_n, g_s and T_r were higher on DAT 25 than on DAT 53 but WUE was lower on DAT 25 than DAT 53 in all treatments (Figure 2). Daily T_r was 6.4 mm in C on DAT 25 and decreased to 3.1 mm on DAT 50, and had no significant difference among treatments on both measuring dates. Saline irrigation decreased P_n by about 11% in T2 and T3 on DAT 25 and the negative effect of saline irrigation on P_n in T2 sustained on DAT 53 (Figure 2(a)). Effects of saline irrigation on g_s and T_r were similar to P_n on DAT 25 but reduced g_s on DAT 53 (Figure 2(b) and 2(c)). WUE was significantly increased by 17% in T3 on DAT 25 (Figure 2(d)).

Figure 2. Effects of measuring timing and treatments on net photosynthesis (a), stomatal conductance (b), transpiration rate (c) and water use efficiency (WUE, d) in C, T1, T2 and T3 treatments. Bar values indicated by different small letters are significantly different at the p < .05. For abbreviations for treatments see Figure 1. DAT means days after thinning.

Plant growth

The maximum LAI, LDM and StDM were observed on DAT 59. Maximum LAI, LDM and StDM significantly decreased regardless the timing of saline irrigation applied (Table 3) and the reduction of LAI, LDM and StDM in T1, T2 and T3 had no significant difference (Table 3). LWR and StWR declined in T2 and T3 on DAT 59 (Table 3). LAD significantly reduced in T2 and T3 treatments from DAT 40 to the last measurement (Figure 3).

Figure 3. Interaction (treatments × timing) effects on accumulated leaf area duration (LAD, days) in C, T1, T2 and T3 treatments. Values indicated with different letters are significantly different at p < .05. For abbreviations for treatments see Figure 1. DAT means days after thinning.

Table 3. Effects of different irrigation treatments on growth, yield and fruit quality parameters.

Parameters	Unit	Treatments
		C	T1	T2	T3
Maximum LAI on DAT 59	–	0.69a	0.53b	0.56b	0.52b
Maximum LDM on DAT 59	g per plant	40.42a	28.83b	33.53b	32.26b
LWR on DAT 59	g g^−1	0.327a	0.314a	0.244b	0.256b
Accumulated LAD	d	40.69a	40.05a	35.70b	37.59b
Maximum StDM on DAT 59	g per plant	62.66a	50.21b	52.81b	50.48b
StWR at DAT 59	g g^−1	0.507a	0.494a	0.426b	0.421b
Marketable fruit	t ha^−1	14.51a	10.93c	13.61ab	11.15c
HI (FW/TW)	–	0.17b	0.18b	0.21a	0.21a
TSS	^oBrix	8.47c	8.68bc	8.77ab	9.70a

Note: Harvest index (HI) obtained by the relationship between dry weight of fruit and total dry weight of plant; total soluble solids (TSS), accumulated leaf area duration (LAD); leaf dry matter (LDM); stem dry matter (StDM), LWR (leaf weight ratio) and StWR (stem weight ratio).Values followed by different letters are significantly different at p ≤ .05. DAT means days after thinning.

Yield and product quality

Saline irrigation significantly decreased the yield of marketable fruits by 24.6% and 23.1% in T1 and T3, respectively, compared to the C treatment but had no effect in T2 (Table 3). Harvests after DAT 50 accounted for 88–95% of the total yield. Saline irrigation had no significant effect on yield in the first harvest (before DAT 50 in T2 and T3 treatments) but significantly declined yield after DAT 50 in T1 and T3 treatments (Figure 4). Saline irrigation significantly increased the harvest index (HI) and TSS in T2 and T3 treatments but had no significant effect in T1 (Table 3).

Figure 4. Yield at different times in C, T1, T2 and T3 treatments. For abbreviations for treatments see Figure 1. DAT means days after thinning.

Salt tolerance

Yield response to salinity was evaluated as a linear model with a threshold (Figure 5) according to the linear Mass–Hoffman model (1977).

Figure 5. Relative yield responses to average soil salinity (0–0.4 m). The areas indicated as S, MS, MT and T correspond to sensitive, moderately sensitive, moderately tolerant and tolerant according to the Mass–Hoffman model.

where Y is the relative yield (%); b is the slope (yield reduction per unit increase in salinity); ECe is the average (0–0.4 m) of soil electrical conductivity of soil sample and a is the salinity threshold. The relationship expresses how the plant growth rate reduces linearly as soil salinity increases above a threshold a.

The relative yield was calculated by dividing the yield in T1–T3 treatments by the yield of C in each harvest. Saline irrigation had no significant effect on the first harvest; therefore, we used nine pairs of relative yield and soil salinity to do the salt-tolerance analysis.

The ECe threshold beyond which the crop yield decreased was 3.69 dS m⁻¹. Beyond the threshold the yield decreased at the rate of 9.85% per unit increase in ECe. Yield reduction rate would rate the cultivar as moderately sensitive to salinity (Figure 5).

Discussion

The seasonal variation in soil salinity was the result of soil water balance between irrigation and evapotranspiration, and the associated movement of soil salt with soil water. Higher daily T_r than the sum of irrigation and rainfall from DAT 40 to DAT 50 could induce the upward movement of soil water and soil salt. Lower ECe from DAT 0 to DAT 40 in C and T1 and from DAT 50 to DAT 88 in C and T2 (Figure 1(a)) could be induced by leaching of salt to deep soil by fresh irrigation because of lower T_r (Huang et al. 2012). The leaching effect could be partly proved by the soil water content increase from DAT 0 to DAT 40. The significant increase in soil ECe between DAT 102 and DAT 1 in T1 and T3, but no change in T2 (Figure 1(b)) indicates the build-up of soil salt primarily occurred in the later period of the growing season. Observed lower soil salinity in 0–0.2 m than in 0.2–0.4 m in C and T1 treatments but no difference between the two layers in T3 (Figure 1(c)) was a different result from a furrow saline irrigation study in which soil salt aggregated in surface layer (Kaman et al. 2006). Suppression on soil evaporation and consequent depressed salt accumulation toward the top layer due to plastic mulch (Brent et al. 1991; Li et al. 2004) could provide support for a different result in soil salt between the current study and the furrow saline irrigation study.

Several studies showed that soil salinity slows growth, decreases yield and plant photosynthesis and transpiration, but compensates the yield loss by promoting fruit quality through increase in TSS (Ismail and Burrage 1994; Schwarz and Klaring 2002; Botia et al. 2005; Bao and Li 2010; Zong et al. 2011). Decreased maximum LAI, maximum LDM, StDM and accumulated LAD (Table 3) in T1–T3 supported the effect of saline irrigation on plant growth, and indicated the sensitiveness of tomato plant growth to salt stress irrespective of the timing saline irrigation used. Higher relative decrease in maximum LAI, maximum LDM and StDM in T1 and T3 than in T2 treatment confirmed the finding that salt sensitivity at reproductive growth stage is greater than at vegetative growth stage (Läuchli 1990; Läuchli & Grattan 2007). Reduction in P_n (Figure 4) but no change in accumulated LAD (Figure 5) on DAT 25 suggests leaf morphology change through increasing leaf thickness as evidenced in kochia (Masoume et al. 2009). Decrease in LDM and StDM after DAT 50 in T2 indicates the persistence of saline irrigation effects on plant growth even though fresh water irrigation has been used.

Tomato yield is quite sensitive to salinity increase above 3.0 dS m⁻¹ (Cuartero & Fernandez-Munoz 1998; del Amor et al. 2001; Malash et al. 2012). Significant decrease of final marketable yield in T1 and T3 but no significant change (Table 3) suggests the stage-dependent sensitivity of yield to saline irrigation. The highest yield loss of tomato was observed when plants were irrigated with saline water at the flowering and fruit-bearing stage (Bao & Li 2010). Yield loss after DAT 50 contributed 88–94% to the total reduction, which seems different with the results that flowering and fruit bearing stage was the most sensitive stage (Bao & Li 2010). In their study, flowering and fruit-bearing stage was 40 days after topping and greenhouse could advance plant phenology. Therefore, the flowering and fruit-bearing stage could be equivalent to DAT 50 in this study. The significant decline in T1 and T3 but not in T2 treatments suggests that saline-water irrigation at peak-fruit-development stage has great influence on yield. Our determined threshold value of 3.69 dS m⁻¹and the corresponding yield decrease rate of 9.85% per unit increase of ECe was not in agreement with the one reported by Mass and Hoffman (1977). Cuartero and Fernandez-Munoz (1998) and Mass (1986) found a threshold value of about 3.0 dS m⁻¹ and yield decrease rate of about 9–10% per unit of ECe increase. The higher threshold value of this experiment is due to the inclusion of the relative yield of 100 (Katerji et al. 2003). The less sensitivity to salinity in the experiment (Figure 5) could be due to the mixed use of fresh and saline water, as shown by the ECe variation in the crop season.

Cherry tomato yield was highly correlated with LAD (R² = .614, n = 16, p < .01). The corresponding higher reduction in LDM and yield in T1 and T3 treatments (Table 3) confirmed the usefulness of LAD as an indicator of crop yield (Huang et al. 2012). Though the fruit yield reduced in T2 and T3 treatments, HI increased (Table 3, +24% in T2 and T3 treatments) because LDM and StDM reduction was smaller than yield decrease (Table 3). HI has no significant change in T1 since marketable fruit decrease (−24.6% T1 vs. C) could be offset by decrease in LDM and StDM (leaf, −28.0% and stem, −23.4% T1 vs. C). The increase in TSS with saline irrigation is widely documented, which could partly compensate yield loss (Bustan et al. 2005; Magan et al. 2008; Malash et al. 2008; Tedeschi et al. 2011; Zong et al. 2011). Irrigation with saline water before DAT 50 and fresh water after DAT 50 should be advocated for cherry tomato plantation in water-scarce areas like the Minqin oasis.

Acknowledgements

The authors would like to thank the anonymous reviewers and editors for their help and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was financially supported by National Science Foundation of China (41301615), West Light Program for Talent Cultivation of Chinese Academy of Sciences (Y329951) and Cooperative project between Chinese Academy of Science and local.