Crop and soil nitrogen responses to phosphorus and potassium ...

Nutr Cycl Agroecosyst (2012) 93:151–162 DOI 10.1007/s10705-012-9506-0

ORIGINAL ARTICLE

Crop and soil nitrogen responses to phosphorus and potassium fertilization and drip irrigation under processing tomato K. Liu • T. Q. Zhang • C. S. Tan • T. Astatkie G. W. Price

•

Received: 24 November 2011 / Accepted: 30 April 2012 / Published online: 13 May 2012 Ó Springer Science+Business Media B.V. 2012

Abstract Shortage of water or nutrient supplies can restrict the high nitrogen (N) demand of processing tomato, leaving high residual soil N resulting in negative environmental impacts. A 4-year field experiment, 2006–2009, was conducted to study the effects of water management consisting of drip irrigation (DI) and non-irrigation (NI), fertilizer phosphorus (P) rates (0, 30, 60, and 90 kg P ha-1), and fertilizer potassium (K) rates (0, 200, 400, and 600 kg K ha-1) on soil and plant N when a recommended N rate of 270 kg N ha-1 was applied. Compared with the NI treatment, DI increased fruit N removal by 101 %, plant total N uptake by 26 %, and N harvest index by 55 %. Consequently, DI decreased apparent field N balance (fertiliser N input minus plant total N uptake) by 28 % and cumulative post-harvest soil N in the 0–100 cm depth by 33 %. Post-harvest soil N concentration was

K. Liu Department of Soil Science, University of Manitoba, Winnipeg, MB R3T 2N2, Canada T. Q. Zhang (&) C. S. Tan Greenhouse and Processing Crops Research Center, Agriculture and Agri-Food Canada, 2585 County Road 20 E., Harrow, ON N0R 1G0, Canada e-mail: [email protected] T. Astatkie G. W. Price Department of Engineering, Nova Scotia Agricultural College, P.O. Box 550, Truro, NS B2N 5E3, Canada

not affected by water management in the 0–20 cm depth, but was significantly higher in the NI treatment in the 20–100 cm depth. Fertilizer P input had no effects on all variables except for decreasing N concentration in the stems and leaves. Fertilizer K rates significantly affected plant N utilization, with highest fruit N removal and plant total N uptake at the 200 kg K ha-1 treatment; therefore, supplementing K had the potential to decrease gross N losses during tomato growing seasons. Based on the measured apparent field N balance and spatial distribution of soil N, gross N losses during the growing season were more severe than expected in a region that is highly susceptible to post-harvest soil N losses. Keywords Nitrogen balance Nitrogen harvest index Nitrogen uptake Soil profile nitrogen Drip irrigation Abbreviations DI Drip irrigation K Potassium N Nitrogen Ncum Cumulative soil inorganic nitrogen Nmin Soil inorganic N concentration NCSL Nitrogen concentration of stems and leaves NHI Nitrogen harvest index NO3-–N Nitrate nitrogen NI Non-irrigation P Phosphorus

123

152

Introduction Processing tomatoes require adequate supplies of water and a proper balance of nutrients to achieve optimum yields (Patane` and Cosentino 2010; Zhang et al. 2010). As described by Hartz and Bottoms (2009), nitrogen (N) uptake by drip-irrigated (DI) processing tomatoes varies from 222 to 466 kg N ha-1. The nutrient uptake ratio of potassium (K) to N for tomato ranges from 1:1 to 2.5:1 (Tapia and Gutierrez 1997; Huang and Snapp 2009), suggesting a higher requirement for K than N. In Florida, state-wide average rates of fertilizer N, phosphorus (P), and K used by tomato growers were 300 kg N ha-1, 87 kg P ha-1, and 461 kg K ha-1, respectively (Florida Agricultural Statistics Service 1999). Gunes et al. (1998) reported synergistic effects of N, P, and K on tomato growth, suggesting that a balance of P and K nutrients enhances N utilization in plants supplied with adequate N. In addition to fertilizers P and K, processing tomato responses to fertilizer N is strongly affected by water supply (Tilling et al. 2007). In a study of fresh marketable tomatoes, Santos (2009) reported that sufficient irrigation reduced N application rates from 336 to 224 kg N ha-1 without significant decline in yields. A reduced N requirement under sufficient irrigation is explained by increased N availability from the soil and fertilizer under conditions of adequate soil moisture (Kim et al. 2008). Similarly, Gheysari et al. (2009a) and Santos (2009) reported that some negative effects of water stress on crop performance are remedied by an adequate supply of fertilizer N. However, in regions susceptible to N losses, substantial increases in N fertilization are associated with negative environmental effects, such as N induced water contamination. South-western Ontario has been identified as a region with a high risk of N leaching losses (De Jong et al. 2007). The average annual precipitation at the study sites from 1991 to 2005 was 780 mm, but only 289 mm occurred during the tomato growing season between June and September (unpublished data collected at the weather station of Agriculture and AgriFood Canada, Harrow, ON), suggesting a water shortage during the tomato growing season and a water surplus during the non-growing period. In order to overcome rainfall deficits, DI is increasingly practiced with processing tomato in south-western

123

Nutr Cycl Agroecosyst (2012) 93:151–162

Ontario. Using DI allows water applications to be precisely controlled to meet crop demands. During the non-growing season, water surplus may trigger considerable soil N losses leading to degradation of water quality in surrounding water systems. De Jong et al. (2007) estimated that approximately 74 % of soil inorganic N measured at the end of a crop growing season is leached into drainage water over the winter period in the study region. Therefore, it is important to assess residual soil N after crop harvests to help develop suitable management practices, such as irrigation, which potentially mitigate ground water contamination caused by high residual soil nitrogen. High residual soil N after processing tomato is harvested has been reported to be a function of excessive N fertilizer applications (Florida Agricultural Statistics Service 1999) and low apparent N recovery by the plants (Scholberg et al. 2000). Numerous studies have demonstrated N leaching losses both during crop growing seasons and postharvest periods (Gehl et al. 2006; Zotarelli et al. 2007, 2009). In addition to measuring leaching losses of N from agricultural drainage tiles, N concentration in the soil profile are also used to evaluate potential leaching losses (Gehl et al. 2006; Zhang et al. 2011). Hence, evaluating spatial distribution of N in the soil profile at the end of a crop growing season, together with plant N responses, will enhance N management in processing tomatoes production systems. Nitrogen application rates are known to affect plant and soil N in tomato production systems (Hebbar et al. 2004; Va´zquez et al. 2006; Zotarelli et al. 2009; Zhang et al. 2011). A recent study recommended N fertilizer applications of 271 kg N ha-1 for drip fertigated processing tomatoes (Zhang et al. 2010), approximately two times greater than previous N recommendations. In contrast, increased plant N use efficiency in tomato has been reported at lower fertilizer N rates with the application of fertilizer K (Fitzpatrick and Guillard 2004). This suggests that successful processing tomato production relies on an integrated nutrient management approach where balanced nutrients rather than any single nutrient supply are required. At high fertilizer N rates, however, soil and plant N responses to P and K fertilization, with and without DI, are unknown. The objectives of this study were to: (1) determine plant N responses of processing tomatoes to fertilizers


153

rates of fertilizer P (0, 30, 60, and 90 kg P ha-1), four rates of fertilizer K (0, 200, 400, and 600 kg K ha-1), and two water management practices, drip irrigation (DI) and non-irrigation (NI). Water management was assigned to the whole plots, with the 16 P and K treatment combinations assigned to the sub-plots. Triple superphosphate was used as the fertilizer P source and potassium chloride was used as the fertilizer K source. Prior to tomato transplanting, fertilizer N, as NH4NO3, was applied at a rate of 270 kg N ha-1 to achieve the maximum marketable fruit yield (Zhang et al. 2010). All fertilizers were disked to a soil depth of 15 cm following hand broadcast. Greenhouse-grown, 5-week old tomato plants were transplanted to the fields in late May or early June each year using a plug transplanter (RJ Equipment, Blenheim, Ontario). The plot sizes were 4.5 m by 4.5 m, composed of three twin rows on a flat 1.5 m by 4.5 m bed. Plants were spaced 40.6 cm in a row, and the row spacing was 45 cm within a twin-row, resulting in a transplanting density of 33,333 plants ha-1. Additional field management details have been previously described (Liu et al. 2011b). The amount and frequency for drip irrigation was determined using a simplified evapo-transpiration model which is a product of air temperature and radiation data, from a nearby weather station. The model also included a locally determined crop coefficient, dependent on tomato growth stage (ranged between 0.2 and 1.1), and emitter flow rate, along with soil moisture retention characteristics (Tan and Fulton 1980; Tan 1990; LeBoeuf et al. 2008). One drip line

P and K rates, with and without DI, using a recommended fertilizer N rate, and (2) assess post-harvest soil profile N responses to water management and different rates of fertilizers P and K.

Materials and methods Site description A 4-year study, 2006–2009, was conducted on the Research Farm of the Greenhouse and Processing Crops Research Center, Agriculture and Agri-Food Canada, Harrow, Ontario (42°020 N, 82°930 W). The fields are flat and represent the general topographical conditions in the study region. Air temperature in the area averaged over the past 87 years (1917–2004) was 19.1 °C for the growing season (1 May–30 September). Total precipitation was averaged at 789 mm for the entire year and at 291.8 mm for the growing season. The preceding crops for the 2006, 2007, 2008, and 2009 experiments were corn (Zea mays L.), tomato, corn, and alfalfa (Medicago sative L.), respectively. The soils at all study sites are classified as Granby loamy sands (sandy, mixed, mesic Orthic Luvisol). Selected baseline soil physical and chemical properties are shown in Table 1. Experimental design and field management The experiment was arranged as a split-plot factorial design with four blocks. The treatments consisted of four

Table 1 Selected soil physical and chemical properties in the 0–20 cm soil depth prior to site preparation at the multiple study sites 2006 Sand (g kg-1)a -1

Silt (g kg ) Clay (g kg-1)

2007

2008

2009

829 ± 1.0b

821 ± 1.1

771 ± 1.0

823 ± 0.4

108 ± 0.4 63 ± 0.6

110 ± 0.8 69 ± 0.4

158 ± 0.6 71 ± 0.6

123 ± 0.2 54 ± 0.3

Soil pH

5.8 ± 0.2

6.5 ± 0.1

6.3 ± 0.1

6.8 ± 0.1

Soil total organic carbon (g C kg-1)

7.2 ± 0.3

6.9 ± 0.4

16.7 ± 0.7

8.8 ± 0.2 0.77 ± 0.02

Soil total N (g N kg-1)

0.61 ± 0.02

0.65 ± 0.05

1.24 ± 0.06

2 M KCl extractable N (NO3-–N ? NH4?–N) (mg N kg-1)

11 ± 2

12 ± 2

13 ± 3

16 ± 2

0.5 M NaHCO3 extractable P (mg P kg-1)

39 ± 3

43 ± 4

65 ± 3

37 ± 4

1 M NH4OAc extractable K (mg K kg-1)

148 ± 10

133 ± 14

179 ± 10

193 ± 12

a

Particle size distribution of sand, silt, and clay was determined by soil hydrometer, soil pH was determined using soil:water = 1:1 extract, and soil total carbon and N were determined by combustion using an automatic LecoÒ analyzer b

Values are means ± standard error (SE) with a sample size of 4

123

154


was placed on the soil surface in the middle of each twin-row bed. The emitter spacing was 30 cm to supply a flow rate of 0.47 L h-1 in order to achieve uniform soil wetting patterns. During each growing season, the tomato plants were drip irrigated daily for 1–3 h, depending on the growth stage. Drip irrigation was suspended whenever there was more than 19 mm of daily precipitation and was stopped 2 weeks prior to harvesting. Water supplied through the drip lines was continuously monitored by a water meter connected to an irrigation controller. Plant and soil sampling and laboratory analyses At the 80 % fruit ripening stage, tomato plants, including the stems and leaves and tomato fruits, were hand harvested from a 2 m long central twin row in each plot. Tomato fruits were separated from the stems and leaves and graded into marketable and nonmarketable fruits. Fresh tomato fruits and vegetative parts, including stems and leaves, were sub-sampled, weighed, and dried at 55 °C for 48 h. After being ground through a Wiley mill with a 2-mm stainless steel sieve, fruits and vegetative parts were separately digested with H2SO4–H2O2 (Thomas et al. 1967). Nitrogen concentration in the digest was determined using a Flow Injection Auto-Analyzer (QuikChem FIA ? 8000 series, Lachat Instruments, Loveland, CO). Plant N uptake was calculated according to dry biomass and the corresponding N concentration. Fruit N removal refers to the N uptake of marketable fruits. Plant total N uptake is the sum of N uptake of fruits, stems, and leaves. Apparent field N balance is calculated as the difference between fertilizer N input and plant total N uptake. A Nitrogen Harvest Index (NHI) was calculated as follows NHI ð%Þ ¼

Nmfruit 100 Nfruit þ Nstemsþleaves

ð1Þ

where Nmfruit is N uptake of marketable fruits, Nfruit is N uptake of all fruits, Nstem?leaves is N uptake of stems and leaves. After the tomato harvest, two soil cores (5 cm internal diameter) were randomly taken to a depth of 100 cm in each plot. Each core was sectioned into 20 cm depth increments. One soil core was taken in the middle of the twin rows where the drip lines were placed, while the other was taken between twin-row

123

beds. To account for the variability of inorganic N concentrations (NO3-–N and NH4?–N) in surface soils six additional soil cores (1.8 cm internal diameter) per plot were randomly taken to a soil depth of 20 cm. The fresh soil samples were composited by depth in each plot and extracted with 2 M KCl to determine soil inorganic N concentration (Nmin). One additional soil core to a depth of 100 cm was also taken in 12 randomly selected plots to determine soil bulk density at each 20 cm soil depth increment. Cumulative soil inorganic N (Ncum) in the 0–100 cm depth of each plot was reported on a kg N ha-1 basis, adjusted on the basis of measured soil bulk density. All soil response variables were determined in the first 3 years from 2006 to 2008.

Statistical analysis Soil inorganic N was determined only in the first 3 years of the study and was analyzed using 12 combinations of the blocks in the field and the years as blocks for soil response variables. The plant response variables were determined throughout the 4-year study period and data from the split-plot factorial experiment were analyzed using 16 combinations of blocks for plant variables. For plant response variables, the three factors with interest (water management and P and K rates) were considered as fixed, and block was considered as random. Since Nmin was measured repeatedly at the depth of 0–20, 20–40, 40–60, 60–80, and 80–100 cm, the data were analyzed as repeated measures using the most appropriate covariance structure. The three factors (i.e. water management, P rate, and K rate) of interest and soil depth were considered as fixed, and block was considered as random. Analysis of variance (ANOVA) was completed using the Mixed Procedure of SAS (SAS Institute Inc. 2008), and further multiple means comparison was completed for significant (P value \ 0.05) effects by comparing the least square means of the corresponding treatment combinations. Letter groupings were generated using a 1 % level of significance for interaction effects and a 5 % level of significance for main effects. For each response, the validity of model assumptions was verified by examining the residuals as described in Montgomery (2009) and appropriate transformations were applied on responses with violated assumptions. The results reported in the tables and figures are back transformed to the original scale.


155

irrigation decreased NCSL by 1.3 g N kg-1, an equivalent of 8 %, compared with the NI treatment (Table 3). Both fertilizers P and K inputs had negative effects on NCSL regardless of water management (Table 4). The NCSL was significantly (5 %) lower in the 90 kg P ha-1 treatment than in the control P treatments receiving 0 kg P ha-1. Compared with the fertilizer K control treatment, application of K at rates of 200, 400, and 600 kg K ha-1 significantly decreased NCSL by 7, 11, and 12 %, respectively.

Results Fruit N concentration and fruit N removals Water management had no effect on fruit N concentration, but significantly affected fruit N removal (Table 2). Fertilizer K rates significantly affected both fruit N concentration and removal, while effects of fertilizer P input and any interactions were not significant for either fruit N concentration or removal. Fruit N removal was 101 % higher in the DI treatment than in the NI treatment (Table 3). Fruit N concentrations did not differ between the 0 and 200 kg K ha-1 treatments or between the 400 and 600 kg K ha-1 treatments. However, K application at high rates of 400 or 600 kg K ha-1 resulted in a significant (approximately 4 %) decrease in fruit N concentration relative to the low rates treatments receiving 0 or 200 kg K ha-1. Applications of fertilizer K (from 200 to 600 kg K ha-1) increased fruit N removal compared with the K control treatment, with the highest fruit N removal in the 200 kg K ha-1 treatment (Table 4). Application of K at the rates of 200, 400, and 600 kg K ha-1 increased fruit N removal by 12, 6, and 8 % relative to the K control treatment, respectively.

Plant total N uptake Water management and fertilizer K rate significantly affected plant total N uptake, while fertilizer P rate or all interaction effects were not significant (Table 2). Similar to the water management effects on the fruit N removal, plant total N uptake was 37.8 kg N ha-1 (an equivalent to 26 % increase) higher in the DI treatment than in the NI treatment (Table 3). Application of fertilizer K at 200 kg K ha-1 led to significantly higher (6–9 %) plant total N uptake than the other three fertility K treatments, but there was no difference in plant total N uptake among treatments receiving 0, 400, and 600 kg K ha-1 (Table 4). N harvest index

N concentration of stems and leaves All three main effects of water management, fertilizer P rates and fertilizer K rate significantly affected N concentration of stems and leaves (NCSL), but all interaction effects were not significant (Table 2). Drip

Water management and fertilizer K rate had significant effects on NHI (Table 2). Additions of fertilizer P had no effects on NHI, neither did any treatment interactions. Drip irrigation increased NHI by 55 % relative to the NI treatment. Additions of fertilizer K

Table 2 The degree of freedom (df) of treatments and ANOVA P values for the main and interaction effects of water management (W), phosphorus (P) rates, and potassium (K) rates on fruit nitrogen (N) concentration, fruit N removal,

N concentration of stems and leaves (NCSL), plant total N uptake, N harvest index (NHI), cumulative soil inorganic N in the 0–100 cm soil profile (Ncum), and apparent field N balance in processing tomato, 2006–2009

Source of variation

df

Fruit N concentration

Fruit N removal

NCSL

Plant total N uptake

NHI

Ncum

Apparent field N balance

W

1

0.184

0.001

0.044

0.001

0.001

0.001

0.001

P

3

0.153

0.156

0.025

0.069

0.176

0.520

0.115

W9P K

3 3

0.758 0.001

0.632 0.001

0.753 0.001

0.699 0.001

0.922 0.001

0.581 0.552

0.681 0.001

W9K

3

0.187

0.131

0.895

0.233

0.280

0.406

0.133

P9K

9

0.137

0.299

0.726

0.116

0.251

0.377

0.096

W9P9K

9

0.640

0.546

0.647

0.793

0.398

0.552

0.818

Significant effects that need multiple means comparison are italicised

123

156


Table 3 Least squares means of fruit N removal, N concentration of stems and leaves (NCSL), plant N uptake, N harvest index (NHI), cumulative soil inorganic N in the 0–100 cm soil

depth (Ncum), and apparent field N balance at the two levels of water management in processing tomato, 2006–2009

Water management

Fruit N removal (kg N ha-1)

Drip irrigation

122.6 a

15.8 b

184.9 a

66.0a

46.4 b

88.0 b

Non-irrigation

61.1 b

17.1 a

147.1 b

42.5b

69.0 a

122.9 a

NCSL (g N kg-1)

Plant total N uptake (kg N ha-1)

NHI (%)

Ncum (kg N ha-1)

Apparent field N balance (kg N ha-1)

Means followed by the same letters within each column are not significantly different at the 5 % level

Table 4 Least squares means of fruit nitrogen (N) concentration, fruit N removal, N concentration of stems and leaves (NCSL), plant total N uptake, N harvest index (NHI), and

apparent field N balance at the four rates of potassium (K); and of NCSL at the four rates of phosphorus (P) in processing tomato, 2006–2009

K rate (kg K ha-1)

Fruit N concentration (g N kg-1)

Fruit N removal (kg N ha-1)

NCSL (g N kg-1)

Plant total N uptake (kg N ha-1)

NHI (%)

Apparent field N balance (kg N ha-1)

P rate (kg P ha-1)

0

19.7 a

86.1 c

17.8 a

165.6 b

51.1b

105.6 a

0

16.8 a

200

20.0 a

96.8 a

16.6 b

174.8 a

54.5a

97.0 b

30

16.3 ab

400 600

18.9 b 19.1 b

91.6 b 92.9 b

15.8 c 15.6 c

160.7 b 162.9 b

55.4a 55.9a

110.2 a 109.1 a

60 90

16.6 ab 15.9 b

NCSL (g N kg-1)

Means followed by the same letters within each column are not significantly different at the 5 % level

from 200 to 600 kg K ha-1 had no effects on NHI, but substantially increased NHI compared with the K control treatments. Application of fertilizer K at the rates of 200, 400, and 600 kg K ha-1 increased NHI by 7, 8, and 9 %, respectively.

Post-harvest soil inorganic nitrogen concentration (mg N kg-1) 0

2

4

6

8

10

0 a

a

Post-harvest soil inorganic N concentration (Nmin) in the 0–100 cm soil profile was significantly affected by the interaction of water management and soil depth (P \ 0.001). However, neither effects of fertilizers P and K inputs nor any treatment interactions except for water management 9 soil depth were significant. In the 0–20 cm depth, Nmin was comparable between the DI and NI treatments (Fig. 1). However, Nmin was significantly lower in the DI treatment than in the NI treatment at all depths between 20 and 100 cm in the soil profile. The Nmin in the 20–40, 40–60 cm, 60–80 cm, and 80–100 cm soil depth was 20, 35, 37, and 38 % lower in the DI treatment than in the NI treatment, respectively. The Nmin was significantly higher in the 0–20 cm soil depth than in any depths of 20–100 cm regardless of water management. A substantial decrease in Nmin was found from the 0–20 cm depth to 20–40 cm depth,

123

Soil depth (cm)

20

Post-harvest soil inorganic N

c

b

40 d

bc

d

bc

60

80 d

bc

Drip irrigation Non-irrigation

100

Fig. 1 Least squares means of post-harvest soil inorganic N concentration in the 0–100 cm soil depth for drip irrigated and non-irrigated processing tomato receiving an N rate of 270 kg N ha-1, 2006–2008. Means sharing the same letter are not significantly different at the 1 % significant level

with a decrease of 51 % for the DI treatment and 46 % for the NI treatment. The Nmin in the DI treatment was significantly higher in the 20–40 cm depth than any depths from the 40–100 cm depth. The Nmin remained statistically unchanged in the 40–100 cm depth (2.1–2.4 mg N kg-1) for the DI treatment and in the


20–100 cm depth (3.4–4.1 mg N kg-1) for the NI treatment. When taking soil bulk density into account, cumulative soil inorganic N (Ncum) in the 0–100 cm depth was calculated and expressed on a kg N ha-1 basis. The Ncum in the 0–100 cm depth was significantly affected only by water management among all main and interaction effects (Table 2). In contrast to the responses of plant total N uptake to water management, Ncum was 33 % lower in the DI treatment than in the NI treatment. Approximately 37 % of Ncum in the 0–100 cm soil profile was in the 0–20 cm depth for the DI treatment and 31 % for the NI treatment. Apparent field N balance Apparent field N balance, i.e. difference between fertilizer N input and plant total N uptake, was significantly affected only by water management and fertilizer K rate, but not by P rate or by any treatment interactions (Table 2). Drip irrigation decreased apparent field N balance by 28 % compared with the NI treatment. Application of fertilizer K at the rate of 200 kg K ha-1 significantly lowered apparent field N balance compared with the other three K rates. Apparent field N balance in the 200 kg K ha-1 treatment decreased by 8, 12, and 11 % when compared with the 0, 400, and 600 kg K ha-1 treatments, respectively.

Discussion Water management effects Water management significantly affected NCSL but had no effects on fruit N concentration. Although fruit N uptake in the DI treatment was twice as much as in the NI treatment, fruit N concentration remained unchanged between DI and NI treatment. The lack of difference in fruit N concentration could be attributed to dilution effects by the substantial increase in fruit yield (Liu et al. 2011b). Such dilution effects did not exist for NCSL, as water management had no effects on the dry biomass of stems and leaves (Liu et al. 2011b). Compared with the NI treatment, the significantly low NCSL in the DI treatment could be a combined effect of high N translocation to fruits, as indicated by high NHI, and insufficient N supply at the

157

late growing season, as demonstrated by low soil mineral N concentration determined at the harvest stage. Critical N concentration is defined as the minimum N concentration required to maximize plant growth (Greenwood et al. 1991). The N concentration in tomato plants, including fruits, stems, and leaves, was B20 g N kg-1, which was lower than critical N concentration presented by Tei et al. (2002) and Hartz and Bottoms (2009) for processing tomato. This suggests inadequate N supply at this specific sampling period of harvest stage. The deficient N supply is also reflected by a corresponding yield reduction, especially when compared to previous studies (Zhang et al. 2010). At the same N rate of 270 kg N ha-1, the marketable fruit yield of processing tomato decreased from 127 Mg ha-1 using a drip fertigation technique (Zhang et al. 2010) to 100 Mg ha-1 under the DI approach (Liu et al. 2011b). Despite the same irrigation schedule, N application timing differed fundamentally between the drip irrigation and drip fertigation regimes. Under drip fertigation, N was split and applied according to crop demand at various growing stages to maximize N use efficiency, whereas all N was applied prior to transplanting for drip irrigation. Therefore, the single application of N at the rate of 270 kg N ha-1 in the present study likely caused potential N leaching losses by rainfall events during the tomato growing season, resulting in N deficiency in the late growing season as indicated by the low NSCL. In order to maintain adequate N for crop growth, we could either adjust N application schedule, such as adoption of drip fertigation, or apply more N prior to transplanting to offset N losses during the growing season which may in fact exacerbate N losses. Nitrogen accumulation in the processing tomato plants was strongly affected by water supply. In a previous field study of drip fertigated processing tomato, Zhang et al. (2011) found that plant total N uptake averaged 256 kg N ha-1 at an N rate of 240 kg N ha-1. In the current study, plant total N uptake only averaged 185 kg N ha-1 in the DI treatment and 147 kg N ha-1 in the NI treatment using a higher N fertilizer rate of 270 kg N ha-1. During the growing seasons in the current study, soil moisture monitored by a time-domain reflectometer at the 20 cm depth averaged 25 % (v/v) in the DI treatment and 14 % (v/v) in the NI treatment. High soil moisture in the DI treatment increases N availability (Ferguson

123

158

et al. 2002; Hebbar et al. 2004) and explains the high plant total N uptake when compared with the NI treatment. High plant N uptake in the DI treatment left lower amounts of N exposed to post-harvest N losses. The positive apparent field N balance (e.g., fertilizer N input minus plant total N uptake) demonstrated that part of fertilizer N was not accounted in the plant N uptake. We found that apparent field N balance was 42 kg N ha-1 higher than Ncum in the 0–100 cm depth for the DI treatment, and 54 kg N ha-1 higher for the NI treatment, suggesting that at least 48 kg N ha-1, averaged across the DI and NI treatments, was lost during the growing season. Water management not only affected plant total N uptake, but also influenced N translocation among plant parts. Wang et al. (2005) found that water stress on wheat inhibited the translocation of N from vegetative parts to grains and lowered NHI. Deficient water supply could cause a depression of stem diameter expansion of tomato plants and reduced the translocation of assimilates to fruits (Kanai et al. 2011), consequently decreasing NHI. Studies have demonstrated that a smaller proportion of plant total N partitioned to the fruits when N was applied in excess of crop N requirements (Stark et al. 1983; Scholberg et al. 2000). When considering the higher soil inorganic N at transplanting (13 mg N kg-1) than at harvest (7 mg N kg-1) in the current study, an additional 14.4 kg inorganic N ha-1 could be provided to crops at the measured soil bulk density of 1.2 g cm-3. The combined effects of water stress and excessive N supplies, with the exception of the late growing season, substantially lowered NHI in the NI treatment relative to the DI treatment. Due to the low NHI, more plant residual N remained in the field in the NI treatment (86 kg N ha-1) than in the DI treatment (62 kg N ha-1). High plant residual N in the NI treatment could compound the off-season N losses as the decomposition of plant residues can provide substantial amounts of mineral N in the early fall, posing challenges for post-harvest N management. Inorganic N in the soil profile was affected by plant N uptake and also reflected downward movement of N during the growing seasons. Even though fertilizer N supplies were much higher than crop requirements in the NI treatment compared with the DI treatment, post-harvest soil N concentration in the 0–20 cm depth was not different between the two levels of water management. According to Zotarelli et al.

123


(2009), 51–78 % of the root of tomato was in the 0–15 cm soil depth with additional 15–28 % in the 15–30 cm soil depth. This suggested that majority of N uptake by tomato was from surface soil, resulting in comparable soil N concentration in the 0–20 cm soil between the two levels of water management. Soil inorganic N at depths between 20 and 100 cm was higher in the NI treatment than in the DI treatment. Higher soil N concentration in the NI treatment was partially related to the lower plant total N uptake. In a corn study, Gheysari et al. (2009b) reported that the decrease in corn N uptake increased post-harvest soil residual N for the deficit irrigation system compared with the full rate irrigation system. Similarly, Wang et al. (2005) found that soil nitrate N concentration in the post-harvest soil was substantially (52 %) higher in the water deficit treatment than in the supplemental irrigation treatment as a result of lower crop N uptake. In the DI treatment, water is precisely controlled according to crop needs, ensuring minimal downward movement of irrigated water to deeper soil depths. However, movement of nitrate N is strongly linked with water movement and nitrate accumulation at the boundary of the wetted area under drip fertigation circumstances have been reported (Li et al. 2003). Although the majority of irrigated water was scheduled to remain in the active root zone in the 0–20 cm soil depth, the wetting front could be down to the 20–30 cm soil depth, thus higher Nmin could appear at this lower depth. This might explain the significantly higher soil inorganic N in the 20–40 cm depth than the 60–100 cm depths in the DI treatment. Drip irrigation was conceived as an ideal technology to enhance water and nutrient use efficiency while reducing N leaching losses (Hebbar et al. 2004); however, heavy N application at the beginning of crop growing season and uncontrollable rainfall events during the growing season might cause large amounts of N losses, thereby decreasing N use efficiency. The current study was conducted on a loam sandy soil in a region classified as a high risk area of N leaching losses (De Jong et al. 2007). Our results showed that post-harvest soil N concentration was 7 mg N kg-1 in the 0–20 cm depth and ranged from 2 to 4 mg kg-1 in the 20–100 cm depth, and was much lower than 13 mg N kg-1 determined prior to the pre-transplanting. The low post-harvest soil N concentration suggested that N leaching losses during the following non-growing season might be minimal compared with


the growing season losses. By contrast, in a commercial field with drip irrigated processing tomato, Stork et al. (2003) reported higher N losses in the postharvest season than during the growing season in a clay soil. Differences in results between our study and Stork et al. (2003) can be due mainly to the soil texture affecting water and associated N movement in soil. Substantial amounts of N losses during the growing season in our study led to reduced N losses during the non-growing season. Therefore, more attention needs to be paid to reduce N losses during the growing season rather than in the post-harvest season on highly permeable soils. Water management effects on Nmin were mostly apparent in the subsurface soil compared with the surface soil, with significantly lower Nmin in the DI treatment than in the NI treatment. No difference in Nmin at the 20–100 cm depth for the NI treatment could be a result of N downward movement during the growing season with extreme rainfall induced water percolation (Gehl et al. 2006). A study conducted in the same region showed that N concentration in the tile drainage was occasionally higher than 10 mg N L-1 during the growing season in a clay soil, demonstrating intensive rainfall during the growing season in the study region causes heavy N leaching losses (Drury et al. 2009). Furthermore, no deep accumulation of N in the examined soil depths, along with the positive apparent field N balance, confirmed that N downward moved beyond the depth (0–100 cm) we examined. Such deep movement of N during the crop growing season has been well documented. For example, Stork et al. (2003) found that soil N decreased with soil depths down to 100 cm but accumulated at depths between 150 and 200 cm at the harvest time in a clay soil. Gehl et al. (2006) also reported that N could move below 2 m in a coarse textured soil during the growing season in an irrigated corn production system. During the growing season of processing tomato receiving 200 kg N ha-1 of fertilizer N, Va´zquez et al. (2006) found that 18–188 kg N ha-1 was leached below 1 m soil depth as a result of occasional rainfall events. The N leaching losses determined at a depth of 75 cm averaged 40 kg N ha-1 during the tomato growing season at the N rate varied from 176 to 330 kg N ha-1 (Zotarelli et al. 2009). The rainfall averaged 310 mm during the growing seasons in the current 4-year study, making deep downward movement of N during the growing season a very likely N loss pathway.

159

Fertilizer P effects Wright (2004) reported a strong positive correlation between N and P concentrations in leaves across various species, suggesting plant N uptake was affected by P supplies. The negative P effects on NCSL in the present study could be attributed to dilution effects, since fertilizer P input increased the biomass of stems and leaves (Liu et al. 2011b). However, plant total N uptake was not affected by fertilizer P input as evidenced by the opposite effects of P on N concentration and biomass. In a recent field study of fertilizer N and P effects on high yielding drip fertigated processing tomato, Zhang et al. (2011) reported that fertilizer P rates, ranging from 0 to 87.3 kg P ha-1, had no effects on plant N uptake, apparent N recovery, or post-harvest soil N. The soil P fertility in the current study ranged from medium to high levels for processing tomato according to provincial guidelines (OMAFRA 2008). The medium to high background soil P fertility might provide sufficient P required for healthy tomato growth, thus having no effects on tomato N uptake and soil N. Although soil N was not affected by fertilizer P input, post-harvest water extractable soil P and Olsen P increased linearly in response to P application (Liu et al. 2011a), suggesting high fertilizer P input exacerbated the adverse P effects on surrounding water systems. Therefore, fertilizer P application in the processing tomato production systems in the study region skewed the nutrient balance potentially causing adverse environmental effects. Fertilizer K effects Nitrogen concentration in the processing tomato production system was responsive to fertilizer K application. The decreased NSCL at the high fertilizer K input treatments might be related to K-induced enhancement of assimilate translocation to fruits. With medium or high N supplies, water use increased in response to increasing fertilizer K application (Ebdon et al. 1999). Similarly, Huang and Snapp (2009) found that increasing K to N ratio from 0.8:1 to over 1.7:1 significantly increased water uptake of tomato fruits. The K:N ratios for the treatment receiving 200, 400, and 600 kg K ha-1 are 0.7:1, 1.5:1, and 2.2:1, respectively. The increased water uptake at high K:N ratio might explain the decrease in fruit N concentration in

123

160

the 400 and 600 kg K ha-1 treatments compared with K control or the 200 kg K ha-1 treatment. Fertilizer K input affected plant N responses with the highest N uptake when K rate was 200 kg K ha-1. The soil K fertility in the current study ranged from medium to high levels for processing tomato according to the provincial guidelines (OMAFRA 2008). The N response to external fertilizer K input in the present study demonstrated high K requirements for processing tomato supplied with a high N rate (270 kg N ha-1) and suggested K deficiencies in the K control treatment. Similarly, Liu et al. (2008) found that, under a relatively high background soil K fertility circumstance, applying fertilizer K significantly increased yield of tomato supplied with high N. Without fertilizer K input, high yield potential driven by high N input might deplete soil K supply. The deficient K in the control treatment substantially decreased photosynthesis (Kanai et al. 2011), and limited growth of crops even supplied with adequate N (Fofana et al. 2008), resulting in significantly lower N uptake in the K control treatment. The K deficiency was also reported to depress stem diameter expansion and then limited the translocation of assimilates to fruits (Kanai et al. 2011), lowering NHI in the K control treatment. Plant N response to fertilizer K depended on soil N fertility level. Under the conditions of low soil N fertility, application of K was reported to increase N use efficiency (Fitzpatrick and Guillard 2004). When N supplies were high, appropriate amounts of fertilizer K application were required to increase plant N uptake while reducing N contamination to environment (Niu et al. 2011). In the current study, K application at the rate of 200 kg K ha-1 increased N concentration and uptake compared with the K control treatment and had potentials for decreasing N losses as indicated by the lowest apparent field N balance. Although tomato requires large amounts of K for profitable production, over applied K input had no effects on yield (Liu et al. 2011b) and plant N uptake. Similarly, studies on corn (Bruns and Ebelhar 2006) demonstrated that supplementing extra K had minimal effects on crop when K supplies was adequate for healthy crop growth. However, over applications of K decreases farmers’ net economic returns. Considering the economic and N utilization effects, we suggest that application of K at the rate of 200 kg K ha-1 were required for processing tomatoes supplied with adequate N in the study region.

123


Conclusions Water management (DI vs. NI) had larger effects on soil and plant N response variables, especially soil N, than fertilizers P and K in processing tomato. As indicated by the difference between apparent field N balance (e.g., fertilizer N input minus plant total N uptake) and Ncum, more N was lost beyond the 0–100 cm soil depth in the NI treatment (53.9 kg N ha-1) than in the DI treatment (41.6 kg N ha-1) during tomato growing seasons. Due to such considerable N losses, N applied at the rate required to maximize fruit yield appeared insufficient at the late tomato growing seasons as indicated by the low NSCL. Although the study area was located in a region with high risks of post-harvest soil N losses, post-harvest soil N losses in this study might be minimal when considering the low post-harvest soil N concentration ranging from 2 to 4 mg N kg-1 in the 20–100 cm soil depth. Therefore, more attention needs to be paid to reduce N losses during the growing season rather than after harvesting, particularly for the processing tomatoes receiving a high N rate on a sandy loam soil. The application schedule of N in drip irrigation needs to be evaluated further to reduce potential environmental contamination if used for drip irrigated tomatoes. Plant and soil N variables, except for N concentration of stems and leaves, did not respond to fertilizer P input due to existing high soil P fertility. Application of K at the rate of 200 kg N ha-1 increased plant N utilization which has implications for reducing N losses. Consequently, the 4-year field study indicated that water management and fertilizer K rates should be incorporated into N management in a sustainable processing tomato production, with the goal of achieving high yields while decreasing N losses. Acknowledgments We thank M. Reeb, D. Pohlman, K. Rinas, and B. Hohner for technical assistance and the Ontario AgriBusiness Association, International Plant Nutrient Institute, Canadian Fertilizer Institution, Ontario Tomato Research Institute, Ontario Processing Vegetable Growers, A & L Canada Laboratories Inc., and Agriculture and Agri-Food Canada Matching Initiative Investment (MII) program for financial assistance.

References Bruns HA, Ebelhar MW (2006) Nutrient uptake of maize affected by nitrogen and potassium fertility in a humid subtropical environment. Commun Soil Sci Plant Anal 37:275–293

Nutr Cycl Agroecosyst (2012) 93:151–162 De Jong R, Yang JY, Drury CF, Huffman EC, Kirkwood V, Yang XM (2007) The indicator of risk of water contamination by nitrate-nitrogen. Can J Soil Sci 87:179–188 Drury CF, Tan CS, Reynolds WD, Welacky TW, Oloya TO, Gaynor JD (2009) Managing tile drainage, subirrigation, and nitrogen fertilization to enhance crop yields and reduce nitrate loss. J Environ Qual 38:1193–1204 Ebdon JS, Petrovic AM, White RA (1999) Interaction of nitrogen, phosphorus, and potassium on evapotranspiration rate and growth of Kentucky bluegrass. Crop Sci 39:209–218 Ferguson RB, Hergert GW, Schepers JS, Gotway CA, Cahoon JE, Peterson TA (2002) Site-specific nitrogen management of irrigated maize: yield and soil residual nitrate effects. Soil Sci Soc Am J 66:544–553 Fitzpatrick RJM, Guillard K (2004) Kentucky bluegrass response to potassium and nitrogen fertilization. Crop Sci 44:1721–1728 Florida Agricultural Statistics Service (1999) Vegetabel chemical use. Florida Agricultural Statistics Service, Orlando Fofana B, Wopereis MCS, Bationo A, Breman H, Mando A (2008) Millet nutrient use efficiency as affected by natural soil fertility, mineral fertilizer use and rainfall in the West African Sahel. Nutr Cycl Agroecosys 81:25–36 Gehl RJ, Schmidt JP, Godsey CB, Maddux LD, Gordon WB (2006) Post-harvest soil nitrate in irrigated corn: variability among eight field sites and multiple nitrogen rates. Soil Sci Soc Am J 70:1922–1931 Gheysari M, Mirlatifi SM, Bannayan M, Homaee M, Hoogenboom G (2009a) Interaction of water and nitrogen on maize grown for silage. Agr Water Manage 96:809–821 Gheysari M, Mirlatifi SM, Homaee M, Asadi ME, Hoogenboom G (2009b) Nitrate leaching in a silage maize field under different irrigation and nitrogen fertilizer rates. Agr Water Manage 96:946–954 Greenwood DJ, Gastal F, Lemaire G, Draycott A, Millard P, Neeteson JJ (1991) Growth rate and %N of field grown crops: theory and experiments. Ann Bot 67:181–190 Gunes A, Alpaslan M, Inal A (1998) Critical nutrient concentrations and antagonistic and synergistic relationships among the nutrients of NFT-grown young tomato plants. J Plant Nutr 21:2035–2047 Hartz TK, Bottoms TG (2009) Nitrogen requirements of dripirrigated processing tomatoes. HortScience 44:1988–1993 Hebbar SS, Ramachandrappa BK, Nanjappa HV, Prabhakar M (2004) Studies on NPK drip fertigation in field grown tomato (Lycopersicon esculentum Mill.). Eur J Agron 21:117–127 Huang J, Snapp SS (2009) Potassium and boron nutrition enhance fruit quality in Midwest fresh market tomatoes. Commun Soil Sci Plant Anal 40:1937–1952 Kanai S, Moghaieb RE, El-Shemy HA, Panigrahi R, Mohapatra PK, Ito J, Nguyen NT, Saneoka H, Fujita K (2011) Potassium deficiency affects water status and photosynthetic rate of the vegetative sink in green house tomato prior to its effects on source activity. Plant Sci 180:368–374 Kim K, Clay DE, Carlson CG, Clay SA, Trooien T (2008) Do synergistic relationships between nitrogen and water influence the ability of corn to use nitrogen derived from fertilizer and soil? Agron J 100:551–556 LeBoeuf J, Shortt R, Tan CS, Verhalen A (2008) Irrigation scheduling for tomatoes—an introduction. Factsheet order no. 08-011

161 Li J, Zhang J, Ren L (2003) Water and nitrogen distribution as affected by fertigation of ammonium nitrate from a point source. Irrig Sci 22:19–30 Liu Z, Jiang L, Li X, Ha¨rdter R, Zhang W, Zhang Y, Zheng D (2008) Effect of N and K fertilizers on yield and quality of greenhouse vegetable crops. Pedosphere 18:496–502 Liu K, Zhang TQ, Tan CS (2011a) Processing tomato phosphorus utilization and post-harvest soil profile phosphorus as affected by phosphorus and potassium additions and drip irrigation. Can J Soil Sci 91:417–425 Liu K, Zhang TQ, Tan CS, Astatkie T (2011b) Responses of fruit yield and quality of processing tomato to drip irrigation and fertilizers phosphorus and potassium. Agron J 103:1339–1345 Montgomery DC (2009) Design and analysis of experiments, 7th edn. Wiley, New York Niu J, Zhang W, Chen X, Li C, Zhang F, Jiang L, Liu Z, Xiao K, Assaraf M, Imas P (2011) Potassium fertilization on maize under different production practices in the North China Plain. Agron J 103:822–829 Ontario Ministry of Agriculture, Food, Rural Affairs (OMAFRA) (2008) Vegetable production recommendations 2008–2009. Publication 363. Queen’s Printer for Ontario. Toronto Patane` C, Cosentino SL (2010) Effects of soil water deficit on yield and quality of processing tomato under a Mediterranean climate. Agr Water Manage 97:131–138 Santos BM (2009) Combinations of nitrogen rates and irrigation programs for tomato production in spodosols. HortTechnology 19:781–785 SAS Institute Inc (2008) SAS OnlineDocÒ 9.2. SAS Institute Inc., Cary, NC Scholberg J, McNeal BL, Boote KJ, Jones JW, Locascio SJ, Olson SM (2000) Nitrogen stress effects on growth and nitrogen accumulation by field-grown tomato. Agron J 92: 159–167 Stark JC, Jarrell WM, Letey J, Valoras N (1983) Nitrogen use efficiency of trickle-irrigated tomatoes receiving continuous injection of N. Agron J 75:672–676 Stork PR, Jerie PH, Callinan APL (2003) Subsurface drip irrigation in raised bed tomato production. I. Nitrogen and phosphate losses under current commercial practice. Aust J Soil Res 41:1283–1304 Tan CS (1990) Irrigation scheduling for tomatoes-water budget approach. OMAF, Toronto, ON, Factsheet order no. 90-049 Tan CS, Fulton JM (1980) Ratio between evapotranspiration of irrigated crops from floating lysimeters and class A pan evaporation. Can J Plant Sci 60:197–201 Tapia ML, Gutierrez V (1997) Distribution pattern of dry weight, nitrogen, phosphorus, and potassium through tomato ontogenesis. J Plant Nutr 20:783–791 Tei F, Benincasa P, Guiducci M (2002) Critical nitrogen concentration in processing tomato. Eur J Agron 18:45–55 Thomas RL, Sheard RW, Moyer JR (1967) Comparison of conventional and automated procedures for N, P and K analysis of plant material using a single digestion. Agron J 59:240–243 Tilling AK, O’Leary GJ, Ferwerda JG, Jones SD, Fitzgerald GJ, Rodriguez D, Belford R (2007) Remote sensing of nitrogen and water stress in wheat. Field Crops Res 104:77–85 Va´zquez N, Pardo A, Suso ML, Quemada M (2006) Drainage and nitrate leaching under processing tomato growth with

123

162 drip irrigation and plastic mulching. Agr Ecosyst Environ 112:313–323 Wang Z, Li S, Vera CL, Malhi SS (2005) Effects of water deficit and supplemental irrigation on winter wheat growth, grain yield and quality, nutrient uptake, and residual mineral nitrogen in soil. Commun Soil Sci Plant Anal 36:1405–1419 Wright JJ (2004) The worldwide leaf economics spectrum. Nature 428:821–827 Zhang TQ, Tan CS, Liu K, Drury CF, Papadopoulos AP, Warner J (2010) Yield and economic assessments of fertilizer nitrogen and phosphorus for processing tomato (lycopersicon esculentum Mill.) with drip fertigation. Agron J 102: 774–780

123

Nutr Cycl Agroecosyst (2012) 93:151–162 Zhang TQ, Liu K, Tan CS, Warner J, Wang YT (2011) Processing tomato nitrogen utilization and soil residual nitrogen as influenced by nitrogen and phosphorus additions with drip-fertigation. Soil Sci Soc Am J 75:738–745 Zotarelli L, Scholberg JM, Dukes MD, Mun˜oz-Carpena R (2007) Monitoring of nitrate leaching in sandy soils: comparison of three methods. J Environ Qual 36:953–962 Zotarelli L, Dukes MD, Scholberg JMS, Mun˜oz-Carpena R, Icerman J (2009) Tomato nitrogen accumulation and fertilizer use efficiency on a sandy soil, as affected by nitrogen rate and irrigation scheduling. Agr Water Manage 96:1247–1258