Introduction

Agricultural activity, particularly in the intensive production areas, is currently facing the important challenge of providing food that is high in quantity and quality, while at the same time producing maximum profits for the farmer, and limiting or having no impact on the surrounding natural environment (Tilman et al. 2011). Intensive vegetable production in Mediterranean climates around the world is a good example of these intensive production areas, in which productivity relies on favorable climate conditions and the availability of water for irrigation. Once climate and water necessities are met, crop productivity largely depends on the amount of nitrogen that the plants can effectively use. Nitrogen supplied externally through fertilizers, in the form of nitrate or ammonium, is, therefore, another factor underlying the high productivity of the Mediterranean climates, and for this reason, a wide variety of synthetic fertilizers containing nitrate and/or ammonium are broadly and abundantly applied in these fields. In leafy vegetables, the nitrogen supply is even more important, as it is necessary to maintain a high rate of vegetative growth and to ensure that the crops are in good condition at harvest time, which is critical for their commercialization. The use of other sources of N, such as ammonium or organic matter, is also common in horticulture, but under Mediterranean conditions, these forms can rapidly be transformed to nitrate. The complexity of providing nitrogen, where, how, and when the plant needs it, is generally compensated for by applying excessive amounts of nitrogen, which will presumably avoid crop N deficiencies during the growing period (Lemaire et al. 2021).

While nitrogen applied in the form of ammonium generally takes one of three routes (absorbed by the plant, fixed by the soil, or used by microbes) (Nieder et al. 2011), nitrate from ammonium nitrification or nitrate that is externally supplied can be absorbed by the plant, used by microbes, or leached to deep soil layers (Sebilo et al. 2013). Nowadays, nitrate leaching is perhaps the main cause of environmental problems resulting from agricultural activities (Quemada et al. 2013; Thompson et al. 2020). Numerous agricultural areas in the world are facing this problem. One of the most prominent problems in Europe is the nitrate pollution of the Mar Menor coastal lagoon in Spain, due to the load of nitrates from the Albujon wadi, which flows into the lagoon from the intensive agriculture areas in its watershed. Under certain soil conditions or agricultural management practices, nitrate used by microbes can also be transformed into gasses such as NO, N2O, and N2, which may also result in major environmental issues and loss of nitrogen use efficiency (Malhi et al. 1990; Abalos et al. 2014; Fentabil et al. 2016).

Making proper use of the water and nutrients in open-field vegetable cultivation is a key aspect for increasing yield while reducing nitrate leaching and mitigating natural ecosystem damage (Quemada and Gabriel 2016; Zinkernagel et al. 2020; Lassaletta et al. 2021). Fertigation, the technique by which nutrients are dissolved in the irrigation water and delivered to the root zone, coupled with drip irrigation system, is an agronomic practice that perfectly match this objective. Fertigation protocols can be easily implemented by farmers in the field, and may therefore have a rapid and effective effect on the plant’s use of water and nitrogen fertilizers. However, the successful achievement of the appropriate on-farm fertigation schedule is a complex and problematic issue. This complexity mainly originates from the interplay between soil characteristics (e.g., texture and organic matter content), crop specie (e.g., root depth, specific nutrients needs), and in many cases, the farm’s infrastructure and technology, which ultimately determines the possibility of applying water and nutrients on a timely basis (Gallardo et al. 2020). For example, increasing irrigation frequency (IF) implies increasing the number of irrigation events per selected period and reducing the timing of these events. It also implies a greater effort by the farmer in irrigation scheduling, and this becomes complicated when there are many irrigation sectors and many crop species or same crops species at different growth stages. At the Campo de Cartagena (Spain), where medium-textured soils predominate, along with medium to large farms (CARM 2021), the most common management strategy is to irrigate with a frequency of two to four irrigations per week. This allows farmers to integrate many different crops at different phenological phases within a farm into the same irrigation controller, facilitating irrigation management and reducing the wear of pumps and irrigation equipment due to start and stop events. However, the management of irrigation events lasting 1–2 h may lead to a high concentration of nitrate remaining in the root influence zone or even lower in the early stages of the crop cycle (Vazquez et al. 2006). This nitrate is prone to leaching from the agro-system during heavy rains, and many efforts should be dedicated to the design of fertigation strategies suitable from an agronomic and environmental point of view.

A critical point in fertigation is how frequent and prolonged the fertigation events should be scheduled. A fertigation frequency that supplies water and nutrients according to crop demand would increase water and nutrient use efficiency. Pioneering studies have demonstrated that a high fertigation frequency implemented under a low nutrient concentration level significantly increased yield by enhancing nutrient uptake (Silber et al. 2003; Xu et al. 2004). Frequent fertigation improved the uptake of nutrients through two main mechanisms: continuous replenishment of nutrients in the depletion zone at the vicinity of root interface, and enhanced transport of dissolved nutrients by mass flow, due to the higher average water content in the medium. This hypothesis was demonstrated in greenhouse conditions with perlite or sand substrates. In sandy-textured soils with a low water and nutrient retention capacity, it seemed clear that a high IF would improve yield in watermelon (Fernandes et al. 2014), or improve vegetative growth and nitrogen efficiency parameters in maize (Hokam et al. 2011). In a sandy loam soil with tomato for processing, yield and water use efficiency (WUE) increased by increasing the IF from 11 to 30 irrigation events during the entire growth cycle that lasted more than 2 months (Liu et al. 2019). Even in loam-type soils, in which the experimental irrigation frequencies were from three events per week to one event every 2 weeks (Rolston et al. 1982), or from three events per week to one event per week (Farneselli et al. 2015), an increase in the frequency of irrigation has agronomic advantages. Rolston et al. (1982) concluded that small and frequent irrigations (three per week) resulted in high NO3− in the upper part of the soil profile, where it was more accessible for plant uptake, and reduced NO3− leaching. Farneselli et al. (2015) found an increased vegetative growth, but not fruit yield, when increasing irrigation/fertigation frequency, perhaps due to an increased uptake of N and improved N nutritional status. However, even in poor clay soils, plants may not always respond positively to increased IF. In mini Chinese cabbage, cultivated in loam soil type, a frequency of irrigation every 4 days, and again every 2 days, was recommended as an appropriate irrigation regime for yield, which reduced N–NO3− residues in the soil (Xiang et al. 2019). Results may be more variable when the cultivation takes place in soils with a higher proportion of clay, where water and nutrients are more easily retained. In these conditions, irrigation events can be more widely distributed in time and of longer durations, as the water spreads equally vertically and horizontally in the soil, hindering the positive effects of increasing the frequency of irrigation. In clay-loam soil cultivated with cotton, a high IF increased plant height and numbers of bolls, but did not improve yield, N uptake, or internal N use efficiency (Ballester et al. 2021). Also in clay-loam soil, but cultivated with muskmelon, weekly fertigation was a better option than daily fertigation, for combining agronomic productivity with environmental sustainability (Abalos et al. 2014). The closest experimental design to the present study, found in an open field cultivated with processing tomato, is one that used a high IF (four to eight irrigations per day) versus one irrigation per day, to reduce drainage without affecting yield (Vazquez et al. 2006). However, during the growth period, the high IF treatment received less water (80% of the reference evapotranspiration) than the low IF, and the nitrate concentration in the root influence zone was not examined. These studies described the large experimental and data variability due to the growth conditions (soil type, irrigation system) and crop type, and they rarely examined different fertigation frequencies that strictly compare the frequency of fertigation, keeping the weekly water and nutrient dosage the same.

We hypothesized that under a production system with a clay-loam soil with plastic mulch, and therefore with minimum soil evaporation, a high IF coupled with a short watering time in a drip-irrigated open-field endive crop would maximize the use of nutrients and water by the plant, improving overall crop performance. But more importantly, the reduction in the nitrate concentration in the root influence zone would justify the use of this irrigation management strategy, despite its implementation disadvantages. By testing two fertigation frequencies (one and three irrigations events per day) against the farmer’s practices (two to four irrigations events per week), this work aimed to unravel how IF (maintaining the same water and fertilizer doses for the entire cycle) affected shoot fresh biomass production, concentration of nutrients in the plant, and nitrate concentration in the soil solution within the root influence zone.

Material and methods

Experimental site and growth conditions

The experiment was carried out in a 0.6 ha field located in Torre Pacheco (37°45′37.18″ N, 0°54′29.75″ W), Murcia, SE Spain. This location has a long history of farming, generally lettuce or endive in the fall–winter season, and melon in the spring–summer season. The soil has a clay-loam texture, representative of the Campo de Cartagena soils, with a bulk density of 1.45 g cm−3, high water holding and cation exchange capacities. The specific soil physical and chemical characteristics are shown in Table 1. It is observed that up to a depth of 40 cm, the soil characteristics are very similar, without a change in horizons. The depth of the groundwater table on the site is more than 5 m. This work includes two fall–winter growing seasons with endive (Cichorium endivia L.) plants cv Cuartana. The specific characteristics of the growing conditions for each crop cycle are shown in Table 2. The plants were planted on raised beds (0.25 m high × 0.5 m width) separated by 0.5 m, and covered with black polyethylene mulch, where a 16 mm drip irrigation pipe (with built-in emitters spaced 40 cm apart, and a flow rate of 2.2 L h−1) was laid at the center of the bed under the plastic. There were five plants per m2. Before planting, the soil was fertilized with 500 and 250 kg ha−1 of Entec Nitrofoska 14–7–17 (N–P2O5–K2O) in September of 2019 and September of 2020, respectively. Also, before planting, the soil was amended with sheep manure at a rate of 16,000 kg ha−1. These practices are common in the area.

Table 1 Physical and chemical properties at different depths of the experimental soil located in Torre Pacheco, Murcia
Full size table
Table 2 Growing conditions of the endive crop in the two growing seasons
Full size table

The irrigation dose was calculated based on the Irrigation-Adviser decision supporting tool (Mirás-Avalos et al. 2019). In brief, this software tool calculates the daily water needs based on the calculation of a soil water balance in the root zone (40 cm in our case). Soil water balance is initiated with a soil water content close to field capacity, and each irrigation event is scheduled to restore the total amount of water lost by evaporation and transpiration to avoid water stress situations. The tool also considers the weather conditions forecasted, the specific crop, and its canopy cover, and any agronomic practice that could modify soil evaporation or plant transpiration. Soluble fertilizers were used, which were first dissolved in 1000 L tanks, and injected into the irrigation pipe in the desired proportion, using a pump coupled with a venturi and a valve. The concentration of fertilizers in the irrigation water was always the same, and the fertigation duration was only changed in each irrigation event depending on the treatment. Fertilization times always ended 4 min before the end of irrigation, regardless of the treatment. The forms and amounts per week of fertilizers were as proposed by the farmer, with the total amount applied being 3.2 and 39.9 kg of N–NO3− ha−1 in the first and second growing seasons, respectively, and the total amount of N–NH4+ applied being 22.9 and 32.6 kg ha−1 in the first and second growing seasons, respectively. The irrigation water contained 441.6 and 309.8 mg L−1 of Cl− and Na+ respectively, with no presence of nitrate. The fertilizers used were potassium nitrate (13–0–46), calcium nitrate (15.5–0–0–26.5), ammonium nitrate (33.5–0–0), nitrate acid (12% N), ammonium sulfate (21–0–0), monoammonium phosphate (11–54–0), monopotassium phosphate (0–52–34), and phosphoric acid (62% P2O5). Cultural practices, such as weed removal, and pest and disease control, were performed by the farmer.

Treatments and experimental design

Three treatments were established that only differed in the frequency of fertigation and watering times, with both water and nutrients applied per plant at the end of the growth cycle being the same for all treatments (Table 2). The low-frequency treatment (LF) received two to four irrigations events per week at 10:00 a.m., using the recommended standard farmer irrigation practices in this region. This treatment received 16 and 15 irrigation events in the first and second growing seasons, respectively (Supplemental Figs. 1, 2), that ranged from 24 to 135 min for each irrigation event. The medium-frequency treatment (MF) received one irrigation event per day at 10:00 a.m., but not every day, depending on the factors cited above. This treatment received 48 and 41 irrigations events through the first and second growth cycles respectively, with each irrigation event ranging from 24 to 45 min each. The high-frequency treatment (HF) received three irrigations events per day at 8:00 a.m., Noon, and 4:00 p.m., the same days as the MF treatment. This treatment received 144 and 123 irrigation events through the first and second growth cycles, respectively, ranging from 8 to 15 min per each irrigation event. The day before, and coinciding with transplanting, the plants were watered abundantly to reach the 50–60 L m−2 needed to saturate this type of soil. This irrigation was followed by 8–10 days without the application of additional water or fertilizers, a common practice in the area, which forces the plants to develop a deep root system and become well established. The experiment was laid out in four blocks with the three randomized treatments in each block. Each treatment in each block consisted of three rows, with the central row being used for the plant and soil determinations.

Soil profile water content and root determinations

Twelve access tubes, one per replicate and treatment, were installed in the center row. For soil moisture determination, a capacitance sensor Diviner 2000 portable device (Sentek technologies, Australia) was used to measure the volumetric water content (VWC) of the soil at regular 10 cm intervals down through the soil profile. The calibration equation for water content in the soil was provided by the manufacturer. The values obtained are not absolute values, but they are useful for knowing the tendency of water in the soil as affected by the different treatments. The results shown are the VWC averages at the 10–20 cm depth (shallowest layer) and the VWC averages at a 30–40 cm depth (deepest layer). Measurements were made at intervals of 3–4 days between 11:00 and Noon. In the second cropping season, 22 days before final harvest, the soil was excavated under the plant rows to obtain a soil profile showing the distribution of roots down to the first 30 cm in depth.

Determinations in the shoot

Total shoot water content was determined in the first growing season at 36, 62, and 89 days after transplanting (DAT) using 20, 20, and 8 plants per treatment, respectively. Shoot fresh biomass at final harvest was determined at 107 and 105 DAT in the first (120 plants per treatment) and second (240 plants per treatment) growing seasons, respectively. For shoot water content, whole fresh shoots were placed in a force air-oven for at least 2 days, after which the total shoot dry weight was determined. In the final harvest, total shoot biomass determination was performed with portable scales, coinciding with the time of harvesting of the remaining plants in the plot by the farmer. Leaf nutrient analysis was performed using completely developed leaves at the final harvest (first season at 107 DAT, second season at 105 DAT) using eight samples per treatment, with two samples per replicate, each composed of leaves from five plants. All the samples were washed with distilled water and dried in an oven for at least 48 h at 65 °C. Dry tissue was ground and sent to the Ionomics Service (Research Support Service, CEBAS-CSIC, Murcia, Spain) for analysis. Cations were analyzed after digestion with HNO3− HClO4 (2:1) in an inductively couple plasma spectrometer (iCAP series 6500, Thermo Fisher Scientific, Franklin, MA, USA). Nitrate (N–NO3−) and chloride were analyzed by ion chromatography (850 professional IC, Metrohm, Herisau, Switzerland) in water extracts obtained by shaking the dried leaf material for 2 h, and the total N concentration was determined with an Elemental Analyzer (LECO TruSpec Micro Series, St. Joseph, MI, USA). Shoot nutrient uptake was calculated based on the shoot dry biomass at final harvest and the leaf nutrient concentration.

Determinations in the soil solution

Two MicroRhizon soil pore water samplers (E-365-19.21.SA, Eijkelkamp Soil & Water, The Netherlands) per treatment and block were used to extract the soil solution at a depth between 30 and 40 cm in the first growing season, and between depths 5–15 and 30–40 cm in the second growing season. The samplers were placed vertically in the soil, after perforating it to the desired depth and coating the porous polymer with slurry. A 10 mL vacuum tube was used to extract and collect the soil solution sample. Vacuum tubes with the soil solution were collected no more than 4 days before placing the vacuum tubes in the field. Following collection, the tubes were immediately stored in a fridge at 4 °C, until analyzed. In this soil solution, the concentrations of N–NO3− and N–NH4+, EC, and pH, were determined weekly. Ammonium and nitrate were spectrophotometrically determined, with the Berthelot reaction (Kempers and Zweers 1986) and the Doane and Horwath method (Doane and Horwath 2003), respectively.

Statistical analysis

A multi-factor analysis of variance (ANOVA) and a mean separation test (Tukey’s test, p value ≤ 0.05) were performed when appropriate. The data were analyzed using SPSS (version 27.0.1.0) (IBM, NW, USA), and graphs were drawn using Sigma Plot (version 14.5) (Systat Software, San Jose, CA, USA).

Results

Soil volumetric water content and soil temperature

In both experiments, the VWC in the soil, regardless of soil depth, was very close to or slightly above the field capacity and remained stable during the development of the crop (Fig. 1). In both cropping seasons, the VWC at a depth of 10–20 cm remain higher in the HF treatment as compared to other treatments. At a depth of 30–40 cm, the highest VWC values during the growing period were found in the LF treatment, with the differences in VWC between the MF and HF treatments being negligible.

Fig. 1
figure 1

Seasonal changes of the volumetric water content (VWC) of the soil profile (upper panels, 10–20 cm depth, lower panels 30–40 depth) as affected by the irrigation frequency (IF) (LF low frequency, MF medium frequency, HF high frequency) in the two growing seasons. Shown are the mean values ± standard error (n = 3–4). Dotted line indicates the VWC at field capacity, and line with dashes and dots indicates the VWC at wilting point

Full size image

Shoot measurements

Shoot total biomass at final harvest day (107 and 105 DAT in the two consecutive cropping seasons) was higher in the HF treatment than in the LF treatment, increasing by 119 g plant−1 on average for both seasons (Table 3). In the MF treatment, differences were not significant with the LF or HF treatments. Whole shoot water content increased at 89 DAT with regard to previous harvests (Fig. 2). While no differences between treatments were observed at 36 DAT, at 62, and 89 DAT, the shoot water content was higher in the HF treatment than in the LF treatment.

Table 3 Shoot fresh biomass at the final harvest as affected by the IF and season, and their interaction
Full size table
Fig. 2
figure 2

Shoot water content as affected by the IF (LF low frequency, MF medium frequency, HF high frequency) treatments and harvest date at 36, 62, and 89 days after transplanting (DAT). N.s. means that the factor analyzed is not significant, and when significant, a mean separation test is included. For each harvest date, treatments followed by different letters are significantly different at 0.05 level of probability according to Tukey’s test. Shown are the mean values ± standard error (n = 20 at 36 and 62 DAT, and 8 at 89 DAT)

Full size image

Carbon and nutrients in the shoot

Total carbon concentration in the leaves was not affected by the IF, nor Total N and P, K, Ca, Mg, and S (Table 4). Nitrogen concentration in the form of nitrate (N–NO3−), independently of the growing season, decreased in the HF irrigation treatment when compared with the MF or LF treatments. Total N content in the shoot was higher in the LF treatment than in the MF or HF treatments in the first growing season, with no differences between treatments in the second growing season (Table 5). Nitrate content in the shoot increased with LF fertigation with regard to HF fertigation. The concentrations of the other nutrients in the shoot were not affected by IF.

Table 4 Effects of the irrigation frequency (IF) on the total C and macronutrients shoot concentration in the 2019-20 and 2020-21 seasons
Full size table
Table 5 Effects of the irrigation frequency (IF) on the total C and macronutrients shoot concentration in the 2019-20 and 2020-21 seasons
Full size table

Nitrate and ammonium in the soil solution

In the first cropping season (2019–2020), the N–NO3− concentration in the soil solution at a depth of 30–40 cm had a tendency to decrease as the sampling date advanced through the growing season, with no N–NO3− detected in the soil solution at the end of the growing season close to the final harvest (Fig. 3, top panel). In this first year of cultivation, the average N–NO3− concentration in the HF treatment was lower than in the MF and LF treatments (Table 6), with these differences being stable throughout almost the entire growing season, and disappearing by the end of the season. No differences were detected between the LF and MF treatments. In the second year of cultivation (2020–2021), peaks were observed in the N–NO3− concentration in the soil solution in the beginning and close to the end, followed by declines in the middle of the growing season and at the very end (Fig. 3, bottom panels). In this growing season, when the depth factor was added to the analysis, the data showed that the N–NO3− concentration increased in the 30–40 cm depth with regard to the 5–15 cm depth (Table 6). This increase was more evident in the LF and MF treatments than in the HF treatment, where it only increased by 1.67 mg L−1. Independently of the sampling depth, the N–NO3− concentration was lower in the HF and MF than in the LF treatment, being more evident in the 30–40 cm depth than in the upper soil layer. Close to the end of the growing season, and at the 5–15 cm depth, the highest concentration of N–NO3− was detected in the treatment with the HF irrigation.  

Table 6 Effects of the irrigation frequency (IF), depth of sample, sample date, and their 2- and 3-way interactions on different parameters measured in the soil solution in the two growing seasons. The average of each of these parameters as affected by the IF (LF, Low frequency; MF, medium frequency; HF, high frequency), and separated by the depth of the sample, are also presented
Full size table
Fig. 3
figure 3

Soil solution N–NO3− concentration as affected by the IF (LF low frequency, MF medium frequency, HF high frequency) and sampling date in the 2019–2020 (upper panel) and the 2020–2021 growing season (bottom panels). In the 2020–2021 growing season, soil solution was also sampled at 5–15 cm depth. Shown are the mean values ± standard error (n = 6–8)

Full size image

Nitrogen in the form of ammonium (N–NH4+) in the soil solution was very low as compared with that of N–NO3−. (Table 6). The N–NH4+ concentration in the soil solution was affected by the sample date only in the first growing season, decreasing to undetected levels at the end of the growth cycle (data not show), but did not respond to the other factors analyzed (IF and depth).

EC and pH of the soil solution

In the first growing season, the EC of the soil solution did not respond to the IF or sampling date (Table 6), with values close to 3 dS m−1 maintained throughout the growing season (Supplemental Fig. 3). In the second growing season, the data showed a high soil solution EC at a low soil depth, as compared with that at a high soil depth (differences in the order of 1 dS m−1). Also, the effect of the IF was dependent on the sampling depth; in the upper soil layer (5–15 cm), the lowest EC was found in the HF irrigation treatment, while in the lower soil layer (30–40 cm), the MF treatment had the lowest EC.

With regard to the soil solution pH, in the first growing season, it tended to decrease as the sampling date advanced throughout the season (data not shown), being lower in the LF treatment as compared to the MF or HF irrigations (Table 6). In the second growing season, the soil solution pH only responded to the sampling date, but not to the IF or soil depth.

Visual root soil profile

The roots in the soil profile in the LF treatment were uniformly distributed until the 30 cm depth (Supplemental Fig. 4). In the MF and HF treatments, the roots were concentrated in the first 15–20 cm of soil, although less dense roots were also found at a lower depth.

Discussion

In a drip-irrigated open-field endive crop located in one of the most important productive and nitrate sensitive lands in Europe (Torre Pacheco, Murcia, Spain) (de Vries et al. 2011; European Commission 2016), the increase in the fertigation frequency from the common practice in the area of two to four irrigation events per week, to three irrigations events per day or even one irrigation event per day, improved the agronomic value, reduced the risk of nitrate contamination to ground and surface water, and reduced N–NO3− concentration in the shoot tissue.

First, from the agronomic perspective, the high IF increased shoot fresh biomass at the final harvest. Endive plants are harvested to meet a weight of about 1 kg per plant for commercial purposes. This does not mean that the plants could have been harvested earlier, since it is necessary to produce about 1.5 kg per plant to obtain a 1 kg commercial-weight plant, leaving the oldest and more damaged leaves in the field, where they will later be incorporated into the soil organic matter. As our data show, frequent small irrigation events produce the heaviest plants at harvest time, which means that harvesting could have been moved forward, saving water and fertilizers, which is an important agronomic advantage. The findings from many studies on numerous crops (tomato, melon, watermelon, maize, Chinese cabbage) are consistent with ours, pointing to an increase in yield when IF is increased (Rolston et al. 1982; Hokam et al. 2011; Campelo et al. 2014; Fernandes et al. 2014; Xiang et al. 2019; Liu et al. 2019). We hypothesized that a high IF combined with low watering times would maximize the plant’s use of water and nutrients and improve overall plant performance. Based on our results, these two assertions are partially supported. While shoot water content improved at high fertigation frequencies, shoot nutrient content was not improved with this schedule, and total N uptake by the shoot even decreased with one or three irrigations events per day. The importance of IF in plant water status, either because it increased water availability in the root zone or because it decreased both osmotic and matric potential between irrigation events (Yasuor et al. 2020), is evident in our experimental conditions. Other studies with different growing conditions, for example in drip-irrigated Rhododendron plants grown in substrate, also point to a reduction in water stress as IF increases (Scagel et al. 2011). The soil texture and its impact on nutrient holding capacity, where medium- to heavy-textured soils retain more nutrients than light soils, could explain the lack of an effect of the IF on nutrient content in the shoot or the shoot nutrient uptake efficiency. For example, in irrigated cotton with a clay-loam soil, a high IF did not affect nitrogen uptake or lint yield per unit of N uptake (Ballester et al. 2021). On the contrary, in lettuce grown in pots with perlite (low cation exchange capacity as compared with clay-loam soil), a high fertigation frequency increased yield, even with a low concentration of nutrients in the nutrient solution (Silber et al. 2003). The Mediterranean climate is characterized by a high evapotranspiration demand, even in autumn–winter seasons. Throughout the two-cropping seasons analyzed in this experiment, the crop water demand calculated according to the Irrigation-Advisor decision support tool (Mirás-Avalos et al. 2019) (mostly transpiration since plants were grown with plastic mulch) was 234 and 208 mm, respectively. Our results indicated that high fertigation frequencies were able to better fit this crop’s water requirements, maintain a better average plant water status, which could have helped boost growth.

Second, and the most important in the context of our objective, a high IF reduced nitrate concentrations in the root influence zone, thus preventing the drainage loss of nitrate with heavy rains, from contaminating groundwater and surface waters. The highly soluble N–NO3− form was by far the most predominant plant-available inorganic N compound in the soil solution, even when N–NH4+-containing fertilizers were used. This is common in many agricultural soils in temperate regions (Di and Cameron 2002), where ammonium is used by the plants, fixed, or rapidly converted to nitrate (NO3−) by microbial oxidation in the process of nitrification (Norton and Ouyang 2019). In the first growing season, the lowest nitrate concentration in the soil solution in the root zone was observed in the three irrigation events per day, and in the second growing season, this was evident in the one or three irrigation events per day. The nitrate concentration in the soil solution, as evaluated by the MicroRhizon probes and vacuum tubes, was validated as a good technique for determining nitrate, ammonium, and other characteristics of the soil solution, when compared with measurements in solutions obtained with the saturated-media extraction method (Argo et al. 1997; Kabala et al. 2017; Capstaff et al. 2021). Moreover, our clay-loam-textured soil facilitated its usability. The importance of increasing the IF and reducing the watering time to limit nitrate loss through leaching has also been successfully tested in other open-field trials with different irrigation methods; on sprinkler-irrigated tomatoes, three irrigation events per week maintained a high NO3− concentration in the upper part of the soil profile, being more accessible for plant uptake, and reducing NO3− leaching (Rolston et al. 1982); on drip-irrigated tomato with plastic mulching, four to eight daily irrigations reduced nitrate leaching by reducing drainage, without a crop yield reduction (Vazquez et al. 2006). In this last experiment, the high IF treatment has lower water dose (80% ETc) than the low IF (100% ETc).

And third, all the plants, in general and independently of the irrigation schedule, had a low nitrate concentration in the shoot at final harvest (less than 250 mg kg−1 of fresh weight, accounting for 15 g H2O g−1 of dry weight in the shoot at the final harvest). This is much lower than the maximum levels permitted by the EU (< 2000 mg NO3− kg−1 fresh weight in Belgium) (Santamaria 2006; European Commission 2011). The low nitrate concentration found in the shoot is a direct consequence of a low nitrate concentration in the root zone, as observed in the topsoil, mainly due to moderate NO3−-based fertilizer inputs received by all the treatments. Under these low nitrate conditions in the root zone, most of the nitrate taken up by the root was assimilated in the root, with small amounts translocated to the shoot (Andrews and Raven 2022). Despite these low nitrate leaf concentrations in all treatments, frequent and short watering periods (high IF treatment) reduced nitrate concentration in leaves, which is another important aspect from the human health point of view, thus favoring the use of a high fertigation frequency. This is an important matter, as excess nitrate in leafy vegetables is a widespread problem for the human consumption of these crops (Santamaria 2006), in which nitrate plays a unique role in maintaining turgor and driving expansive growth (Andrews et al. 2005). There is scarce literature addressing shoot nitrate concentration as affected by fertigation frequency. Similar to our finding, in drip-irrigated Chinese cabbage, a high IF increased water use efficiency and yield, and reduced nitrate concentration in the leaves (Xiang et al. 2019). As discussed below, a possible explanation for the decrease in nitrate concentration in leaves of plants irrigated with a HF could derive from a decrease in the nitrate found in the root zone, related to the loss of nitrate to the atmosphere in the gaseous form, favored by the clay content of the soil. Also, an explanation for the highest nitrate concentration in plants with low irrigation frequencies and highest watering periods may be related with the foraging for water in deep soil layers, which may promote nitrate uptake (Pedersen et al. 2010; Lilley and Kirkegaard 2011). As shown in our results, an infrequent and large watering regime concentrates much more nitrate in the subsoil layer; these plants presumably developed deeper root systems, and were thus able to use and accumulate more nitrate in the leaves coming from deeper soil layers. Dry matter (C accumulation) and N accumulation are closely associated (Justes et al. 1994), and a delay in growth, for example in the plants irrigated less frequently, could also lead to an increase in total and reduced N concentrations.

We observed that a high IF decreases nitrate concentration in the root influence zone and in the shoot [final harvest and previous harvest (data not show)]. If nitrate is low in the root zone and low in the shoot, an explanation is needed for the destination of the nitrate that is either directly applied by fertigation, or originated from the nitrification of ammonium supplied by fertilizers, or from the mineralization of organic matter. Considering the average values of nitrogen content in the shoot, 3.02 and 2.76 g shoot−1, in the low and HF irrigation treatments, respectively, and the 50,000 plants per ha, we could suspect that this reduction of 11 kg ha−1 of N, that were absorbed by the plants in the HF irrigation treatment, could presumably have been lost in the form of gas. We suggest that nitrate, when supplied through HF irrigation, could be partially lost through denitrification, due to the high temporal water content in the topsoil, which would facilitate the activity of the bacteria responsible for the transformation of nitrate to nitrous oxide (Bateman and Baggs 2005). Previous studies support this claim, in field-grown tomato irrigated with sprinklers (Rolston et al. 1982), in Chinese cabbage with subsurface-drip fertigation (Hamad et al. 2022), or in drip-irrigated apple (Fentabil et al. 2016). Denitrification rates increase linearly at a 25–45% VWC (Beggs et al. 2011), coinciding with the water contents found in the topsoil of our endive field. The authors identified threshold VWC limits of 40–45% for denitrification. Other studies have used water-filled pore space (WFPS) as a surrogate of moisture status, to understand N dynamics in the soil (Bateman and Baggs 2005; Medinets et al. 2015). The most suitable condition for nitrification is at < 60% WFPS (at field capacity) due to the available substrate and O2 for nitrifiers (Linn and Doran 1984). On the other hand, a higher WFPS (> 60%, above field capacity) favors denitrification due to the lack of enough O2 in the available pore space, which inhibits the diffusion of O2 into the soil microsites where anoxic conditions prevail for denitrifiers (Alaoui-Sossé et al. 2005; Hou et al. 2012). Transforming our VWC data into percentage of WFPS, according to our soil bulk density and particle density (2.65 Mg m−3) results (Abalos et al. 2014), the percentage of WFPS was higher than 60% in the topsoil, which is a favorable condition for denitrification and or nitrate ammonification.

Overall, the novel results from this study, in which the doses of water and nutrients per week were maintained equal between treatments throughout the cultivation cycle, confirm the great influence of the fertigation management (coupled with water and nitrogen inputs) on soil nitrogen dynamics, and how the optimization of fertilization and irrigation could help reduce the impact of agricultural activity on the environment, without causing penalties on crop production. Irrigation management through the intensification of fertigation (more frequent irrigation events per crop cycle with shorter watering times) has proven to be agronomically effective in an intensive open-field leafy crop, helping to reduce nitrate concentration in the root influence zone and in the shoot. Large areas of vegetable production around the world could benefit from this simple strategy of irrigation management. For example, the Region of Murcia has a total of 52,716 ha of irrigated horticultural crops, almost half of which are leafy crops (CARM 2021). The drawbacks of this intensification, such as a low nutrient use efficiency, and the possible loss of nitrate through denitrification, could perhaps be accentuated by the medium texture soil, and reduced in lighter soils with less capacity to hold water and nutrients (Silber et al. 2003; Xu et al. 2004). A weakness of the implementation of this management practice is the infrastructural and technological development needed to irrigate crops with higher irrigation frequencies. This strategy would require a modern irrigation infrastructure with irrigation controllers that allow increasing the number of fertigation events in many irrigation sectors, which would imply more starting and stopping periods of the different equipment and water pumps, and the consequent fatigue of these pieces of equipment.

Conclusion

The present study has demonstrated that the irrigation scheduling with varying fertigation frequency without altering the total water and nutrient application influenced both crop performance and nitrogen dynamics in the soil–plant system. The beneficial effect on the biomass production of the high fertigation frequency was due to a more efficient use of the irrigation water, but not of nitrogen or other nutrients. The characteristics of the soil, with its high water and nutrient holding capacity, and the use of plastic mulching, could be the determining factor in masking possible differences in nutrient use. The concentration of N–NO3− in the root influence zone decreased by frequent and short watering periods, thus allowing this strategy to be implemented as an effective and simple agronomic practice to reduce possible loss of NO3− from cultivated land, and to reduce contamination of ground and surface water. Our data and the analysis of current literature suggest that some of the nitrate supplied or formed with a high IF may be lost to the atmosphere in gaseous form. This would also explain the reduction in the nitrate concentration in leaves when a HF irrigation strategy is used. Considering our local yet omnipresent problem of nitrate contamination in ecosystems and water sources, we should decide on the suitability of reducing nitrate losses below the root zone, as opposed to possible losses to the atmosphere. Future work on intensive drip-irrigated vegetables should address gaseous N loss under different fertigation management regimes, as well as the nutrient use efficiency, by reducing fertilizer inputs while increasing fertigation frequency.