1 Introduction

Barley (Hordeum vulgare L.) is primarily used for both malt and beer production and as animal feed; it is also used as food for human consumption. Barley grain is high in carbohydrates and also contains moderate amounts of protein, calcium, and phosphorus and small quantities of vitamin B. Barley is the world’s fourth most important cereal crop (after wheat, rice, and maize) in terms of both quantity of production and area of cultivation. The increase in the total barley production over the last half-century has mainly been due to increases in yield, since the total area under production has remained relatively stable, or has even decreased in recent years (FAOstat 2018). Hordeum species are found in most areas with Mediterranean climates.

Nitrogen (N) is an essential element for crop development. The rate of N fertilizer application is critical for cereal crops since N affects grain yield and plant quality (Lopez-Bellido 2009). Under Mediterranean climatic conditions, barley crops under rainfed Mediterranean conditions extract approximately 25 kg of N per 1000 kg of grain produced from the soil. The need for N fertilization in barley crops differs from region to region. While the need for greater food production has increased, the global consumption of N in its synthetic (commercial) and organic (manure) forms has increased at an even greater rate. Several organic and inorganic N sources can supply the N required for optimum crop growth. In 2018, Spain was the largest pig producer in the EU (Eurostat 2018). These animals produce large quantities of manure every year, which presents a problem in terms of the management of the very large volume of waste produced. The agricultural use of this manure is recommended not only for fertilizing but also to facilitate the disposal of these increasingly important residues. Given its value as a fertilizer, pig slurry is commonly used in agriculture and this is a relatively cheap way of decomposing it and offers a low-cost alternative to mineral fertilizers. However excessive applications of manure may cause the pollution of groundwaters as a result of leaching, especially involving nitrates.

In various regions of Spain, it has been traditional to apply high doses of N fertilizer with the sole aim of achieving high yields. However, there are several drawbacks associated with the overuse of N fertilizers: (1) the cost, from an agronomical point of view overuse does not produce a significant increase in crop yield; and (2) environmental problems, normally associated with N losses to the natural environment, through N leaching to groundwater, the runoff of surface water and emissions of hazardous gases, such us ammonia (NH3), nitrous oxide (N2O), and nitric oxide (NO) (Spiertz 2010). The excessive use of fertilizers constitutes an environmental risk and, according to the European Commission, farming is responsible for over 50% of total nitrogen discharges into surface waters. Nitrogen application is usually based on crop N requirements which are applied in accordance with the EU Nitrates Directive (Council Directive 91/676/EEC n.d.). In most EU countries, slurry application is limited to 170 kg N ha−1 year−1. The original objective of the Nitrates Directive was to protect water quality across Europe by preventing nitrates from agricultural sources polluting ground and surface waters and by promoting the use of good farming practices. In line with this Council Directive, the regional government of Castilla-La Mancha (central Spain) published a set of guidelines for the N fertilization of crops, suggesting stricter limits in areas considered vulnerable to nitrate pollution (Real Decreto 261/1996 n.d., DOCM 2011: Orden 07/02/2011).

It is also important for crops to have a high level of Agronomic Efficiency. Nitrogen agronomic efficiency (AEN) is a parameter that expresses the ability of the plant to increase its yield in response to N application. Improving N use efficiency (NUE) in crops is also economically beneficial to producers and to the environment. A more efficient and ecologically responsible use of N could be achieved by adopting a series of improved strategies and management practices (Abunyewa et al. 2017; Guo et al. 2019; Ladan and Jacinthe 2017; Perchlik and Tegeder 2017; Pinochet et al. 2018; Zhang et al. 2017).

Zinc is an essential micronutrient for plant nutrition and human health. However, Zn deficiency is a very widespread problem throughout the world and results in reduced crop yields and poor crop quality. About 50% of the soils used for cereal production in the world contain low levels of bioavailable Zn (Cakmak and Kutman 2018). Crop biofortification offers a sustainable and relatively cheap way to provide essential micronutrients to people in both developing and developed countries. Biofortification is a process that involves increasing the natural content of bioavailable nutrients in crop plants (Guo et al. 2016). The biofortification of cereals using Zn has been the focus of considerable research in recent years and has become increasingly relevant and important for both crops and humans (Guo et al. 2016; Liu et al. 2017). Applying Zn to barley increases both its fertility (number of grains per ear) and grain quality. Zinc concentrations and contents have been progressively increased through the application of Zn fertilization. In cases of extreme deficiency, Zn application may increase barley crop yields by as much as 48% compared with control crops only receiving NPK fertilizer (Singh 2008).

Zinc deficiencies in plants have traditionally been corrected by the application of inorganic Zn sources, such as ZnO, or salts, such as ZnSO4 or Zn(NO3)2. Organic Zn sources, such as synthetic chelates and organic complexes, are now commonly used. Applications of Zn to calcareous soils in the form of chelates (such as Zn-EDTA –Zn- ethylenediaminetetraacetate-, Zn-DTPA–Zn-diethylenetriaminepentaacetate or Zn-HEDTA–Zn-hydroxyethyl-ethylenediaminetriacetate) can have important implications for crop nutrition and that these are highly effective sources of Zn for crops. Zinc chelates provide the micronutrients needed to produce high concentrations of water-soluble Zn and available Zn in soils, though the effectiveness of these chelates depends on their stability. However, Zn chelates differ in their physical state, chemical reactivity, cost, bioavailability, and susceptibility to leaching. The chelating agents EDTA (ethylenediaminetetraacetate acid), HEDTA (hydroxyethyl-ethylenediaminetriacetate acid), and DTPA (diethylenetriaminepentaacetate acid) are some of the strongest synthetic chelating agents, and when used in combination with Zn, form much stronger chelates than naturally occurring organic ligands (Mortvedt and Gilkes 1993). Zinc chelates enhance Zn availability due to less Zn2+ ion interaction with soil components (Pagani et al. 2013).

Various authors have studied the effects of applying Zn-N combinations using inorganic sources of Zn, such as ZnO, ZnSO4, or Zn(NO3)2 (Cakmak et al. 2010; Erenoglu et al. 2011; Gonzalez et al. 2019; Kutman et al. 2010). All these studies have focused on the impact of applying Zn-N combinations on N and Zn concentrations in plants in just one way: N fertilization combined with soil Zn applications has a significant interaction on grain Zn and N concentrations. Improving the N nutritional status of the plant may enhance the abundance of transporter proteins and nitrogenous chelators involved in the uptake, translocation, remobilization, and grain allocation of Zn in cereals (Cakmak et al. 2010). For example, Kutman et al. (2011) obtained increases of over 50% in Zn concentrations in whole grain wheat when the N supply was increased from low to very high. Zinc-N management appears to offer a promising agronomic strategy for the biofortification of cereals with Zn. On the other hand, to the best of our knowledge, there have been no specific studies of the application of Zn chelates in conjunction with N sources (organic vs. inorganic) to reduce the N rate with the aim of minimizing the environmental impact of fertilization yet maintaining adequate plant nutrient concentrations and a high crop yield.

The main objective of the present study was to determine the influence of the combined application of Zn, in the form of a synthesized Zn chelate fertilizer (Zn-DTPA-EDTA-HEDTA), and different N sources (pig slurry or urea) as a strategy to reduce the N application rate in a barley crop grown in a calcareous soil in order to use fertilizer more rationally and effectively and for improving both nitrogen use efficiency and Zn biofortification of barley.

The specific objectives were:

  1. (1)

    To study the effects of applying a combination of Zn-DTPA-EDTA-HEDTA and urea or pig slurry N fertilization on grain yield, Zn content, and protein content in a barley crop

  2. (2)

    To compare the agronomic efficiency of Zn and N applied via different combinations of Zn and N

  3. (3)

    To establish the relative status of soil Zn fractions and the potential phytoavailability of soil-applied Zn based on different extraction methods

2 Material and Methods

2.1 Field Locations and Soil Characterization

The field experiment was conducted at the “El Encín” Research Farm (40.52° N 3.30° W, 599 m) Madrid, Spain. This area is near “Alcarria-Guadalajara,” which is in a part of Castilla-La Mancha that has been classified as sensitive to nitrate pollution contamination (DOCM 2011). The average temperature from December to June was 11.1 °C (average temperature range = 4.32–21.01 °C). The rainfall in the crop growth period was 344 mm.

Soil samples from the Ap horizon (0–28 cm) were air-dried, sieved (< 2 mm), and then analyzed according to the Spanish official methodology (MAPA 1994). The soil was classified as Calcic Haploxerept (Soil Survey Staff 2010). Its main characteristics were as follows: the soil texture silty-loam (310 g kg−1 soil Sand, 540 g kg−1 soil Silt, and 150 g kg−1 soil Clay) (Day 1965), pH and electrical conductivity (soil-to-water ratio of 1:2.5) were 8.20 and 188.1 μS cm−1, respectively; organic matter, 12.6 g kg−1 soil; total nitrogen, 1.0 g kg−1 soil; available phosphorus (Bower et al. 1952), 30.90 mg kg−1 (Olsen 1954); total Zn, 51.70 mg kg−1 soil, and DPTA-TEA extractable Zn, 0.98 mg kg−1 soil. Soil texture was analyzed using a Bouyoucos densimeter. The percentage of organic matter and total N content were determined using the Walkley–Black and Kjeldahl methods, respectively.

2.2 Experimental Design

The experiment was run with a randomized complete block design that included three replicate plots (7.5 m2 plot−1) for each treatment. Barley (Hordeum vulgare L. cv. Quench) was sown in December and harvested in early-June in the following year. Basal fertilization included 90 kg P2O5 ha−1 and 70 kg K2O ha−1 (aqueous solutions of KH2PO4 and H3PO4).

A combination of three nitrogen (N) levels (30, 60, and 90 kg ha−1) and four zinc (Zn) levels (0, 0.5, 1, and 1.5 kg ha−1) of two different sources was applied during the early vegetative stage (February). Nitrogen was applied in the form of granular urea (46% N) or the liquid fraction of pig manure (total N, 2.0 g kg−1; N-NH4+, 1.23 g kg−1; total P, 97.5 mg kg−1; total K, 1.32 g kg−1, total Zn 22.14 mg kg−1; total Cu 4.87 mg kg−1; oxidizable organic matter, 4.60 g kg−1; dissolved organic carbon, 0.5 g kg−1; pH, 7.61; and dry matter (DM), 16 g kg−1). Zinc was applied as ZnCHE (Zn-DTPA-EDTA-HEDTA) [Zn-DPTA (Zn-diethylenetriaminepentaacetate), EDTA (Zn-ethylenediaminetetraacetate), and HEDTA (Zn-N-2-hydroxyethyl-ethylenediaminetriacetate)].

A control, without applications of either Zn or N fertilizer, was also performed (total plot number 75) to calculate the NUE. This treatment was not used for statistical analysis as it did not conform to a real fertilization strategy used for this crop or in this region. Nitrogen fertilizer applications of about 100 kg N ha−1 are typically recommended for barley crops in this area (Lopez-Bellido 2009). The N application rates were therefore lower than typically recommended (with approximate reductions of 70%, 40%, and 10%, respectively, compared to traditional application rates for the 30, 60, and 90 kg N ha−1 application rates).

2.3 Sample Preparation and Chemical Analysis

A sample of six 0.5-m-long rows of barley plants were harvested from each plot. The different plant parts (root, stem, leaf, and grain) were then separated. The root samples were carefully washed with deionized water and then all the samples were dried at 60 °C for 72 h. Grain from the remaining plants was harvested using a plot combine (Weinterteiger Classic). Soil samples from the rhizosphere were collected and homogenized, air-dried, sieved (< 2 mm) and stored for further analysis.

The Zn, Fe, Mn, and Cu concentrations in the different parts of the plant (root, stem, leaf, and grain) were determined by wet digestion in Sample Preparation Block Systems (SPB PROBE PerkinElmer) using “Teflon tubes” kept at a maximum temperature of 150 °C for 2 h, with 1.5 g of dried samples, 10 mL of HCl (37%), 10 mL of HNO3 (65%), and 5 mL of HF (48%). The grain N concentration was determined using either the Kjeldahl or the Dumas method. Protein concentration was calculated as the N concentration multiplied by 5.83.

Total Zn, potentially available Zn (DTPA-TEA and Mehlich-3), and Zn distribution in the different soil fractions were all determined from the soil samples. Total Zn was determined by acid digestion after treating 1 g of soil with 14 mL HNO3 (65%) and 6 mL HF (48%) followed by digestion in Teflon bombs in a microwave oven with a rotating tray. This involved a three-step process at a maximum pressure of 700 kPa.

The concentration of potentially available Zn in the soil was estimated using two different chemical extraction procedures: DTPA–TEA (diethylenetriaminepentaacetic acid–triethanolamine), (Lindsay and Norvell 1978) and Mehlich-3 (Tran and Simard 1993). The distribution of Zn in the different fractions was obtained by selective sequential extraction (SSE). The SSE used was a six-step procedure which was performed according to Leleyter et al. (1999). The soil (g)/extractant solution (mL) ratio was 1:10. The Zn fractions were sequentially determined using a 2.5 g soil sample combined with the following extractants: NH4Ac 1.0 M, pH = 7.0 (water-soluble plus exchangeable, WSEX); NaOAc 1.0 M. pH = 4.5 (Carbonate bound, CAR); NH2OH·HCl 0.1 M in HNO3, pH = 2.0 (easily reducible oxide bound, RMO); (NH4)2C2O4 0.2 M + H2C2O4 0,2 M, pH = 3.0 (amorphous minerals colloids bound, AMC); (NH4)2C2O4 0.2 M + H2C2O4 0.2 M, pH = 3.0 in ascorbic acid 0.1 M (crystalline Fe oxide bound, CFeO). The organic material and sulphide bound + residual fraction (OM + RES) was calculated as the difference between the total Zn extracted by wet acidic digestion in a microwave oven and the sum of the other fractions.

“PerkinElmer Pure” standard checks were used for the Quality Assurance System (certified by NIST–SRM). Standard solutions of Zn, Fe, Mn, and Cu were prepared for each extraction in background solutions of the extracting agents. In all cases, the Zn, Fe, Mn, and Cu concentrations were determined by flame atomic absorption spectrophotometry (PerkinElmer AAnalyst 700).

2.4 Statistical Analysis

Descriptive, simple, and stepwise multiple regression analyses and other statistical studies were conducted using Statgraphics Plus software, version 5.1 (Manugistic Inc., Rockville, MD).

Multifactor analyses of variance (ANOVA) of the parameters studied were carried out to determine the main effects of the Zn application rate, N application rate, N source, and experimental repetition, and the interactions between them. When two-way interaction between the Zn rate and N rate factors was significant, we performed a new multifactor analysis of variance to determine the main effects of the combined factor. Means were compared using Fisher’s protected least significant difference (LSD) tests. Differences were considered significant at P < 0.05.

3 Results

3.1 Plant Response to Fertilizer Treatments

The analysis of variance (ANOVA) for the plant variables studied in the barley crop to determine the main effects of the ZnCHE (Zn-DTPA-EDTA-HEDTA) application rate (ZnCHE-0, ZnCHE-0.5, ZnCHE-1, ZnCHE-1.5), N application rate (N30, N60, N90), and N source (pig slurry (PS) or urea) and the interaction between these main effects are shown in Table 1. The Zn and N application rates caused significant differences in grain yield and Zn concentrations in the different parts of the plant. However, only grain yield was significantly influenced by the N source. In leaf and grain Zn concentrations, there were interactions between the Zn application rate and N source. In the grain Zn concentration, there was also an interaction between the N application rate and N source.

Table 1 Analysis of variance (ANOVA) of the effects of Zn application rate, N application rate, N source, and their interactions on grain yield; root, stem, leaf, and grain Zn concentrations; N and Zn uptake and grain protein concentration
Full size table

Figure 1 shows the grain yield for each treatment and also the mean values when all the data were studied in combination. The highest grain yield was observed with PS + N90 treatments when ZnCHE was applied at any of the three Zn application rates (3873, 3857, and 3928 kg ha−1, with ZnCHE-0.5, ZnCHE-1.0, and ZnCHE-1.5, respectively). The lowest grain yield was obtained without a Zn treatment (ZnCHE-0) and with N30, applied either as PS or in the form of urea treatments (3015 and 3026 kg ha−1, respectively). According to the statistics, grain yield increased when all of the ZnCHE application treatments were compared to the ZnCHE-0 treatments (by between 9.5 and 11.0% compared to the ZnCHE-0 treatments). An increase in the N application rate also significantly increased the mean grain yield (by between 6.9 and 13.0%) and the application of PS produced a higher mean grain yield than urea (which increased it by 6.0%). Orthogonal contrasts were used to compare the effects of the combined application of Zn and N on mean grain yield. Increases were observed in the mean grain yield produced by combined application of Zn and N compared with the combined application of the nil-Zn treatment (ZnCHE-0) and N. The estimated difference between the treatment means was 1505 kg ha−1 for ZnCHE-0 (N30 + N60 + N90) vs. ZnCHE-0.5 + ZnCHE-1.0 + ZnCHE-1.5 (N30 + N60 + N90) (P < 0.05).

Fig. 1
figure 1

Effect of ZnCHE (Zn-DTPA-HEDTA-EDTA) at doses 0, 0.5, 1, and 1.5 kg ha−1 and N in form of pig slurry or urea (at doses 30, 60, 90 kg ha−1) treatments on grain yield of barley (Hordeum vulgare L.) grown under field conditions. Vertical bar at each of the data points represents the standard deviation from the means. Statistical differences at P < 0.05 (LSD test) are presented by different letters (bold, capital, and roman letters indicate the differences between N source, ZnCHE doses, and N doses, respectively)

Full size image

The mean Zn concentrations in root, stem, leaf, and grain obtained with the different Zn and N application rates and N sources and their respective interactions are shown in Table 2. The different Zn rates employed had an effect on the mean Zn concentrations in the different parts of the plant (root, stem, leaf, and grain), with average increases of up to 75, 106, 59, and 30%, respectively, when the ZnCHE-1.5 treatments were compared to the ZnCHE-0 treatments.

Table 2 Zinc concentrations in root, stem, leaf, and grain obtained with the different Zn and N application rates and N sources. Italic values show mean Zn concentrations and the ANOVA of the effects of Zn application rate, N application rate, N source, or their interactions when these were significant
Full size table

Increasing the N application rate from N30 to N60 significantly increased the mean Zn concentrations in the root and stem (by 10.2 and 26.0%, respectively). In contrast, the mean Zn concentrations in the root and stem were not significantly affected by an increase in the N application rate from N60 to N90. However, increases in the N application rate from N30 to N60 and from N60 to N90 significantly increased the mean Zn concentrations in the leaf (by 8.2 and 14.7%) and grain (by 7.8% and 8.0%), respectively. In contrast, the different N sources employed (PS or urea) had no effect on the mean Zn concentrations in the different parts of the plant.

The combined factor of N sources × Zn application rates significantly affected (P < 0.0001) the leaf and grain Zn concentrations. The treatments in which ZnCHE-1.5 was applied (both with urea and with PS), and also those involving PS + ZnCHE-1.0, produced the highest leaf and grain Zn concentrations. The combined factor of N sources × N application rates employed only had a significant (P < 0.05) effect on grain Zn concentration. The highest grain Zn concentrations were observed with the urea + N90 treatment. Even so, there were no significant differences between the grain Zn concentration obtained with the urea + N90 treatment and the PS + N90 and PS + N60 treatments.

The grain N and Zn uptake and grain protein concentrations obtained with each of the treatments are shown in Fig. 2. As expected, the mean comparisons for the N application showed that N uptake significantly (P < 0.0001) increased with each increment in the N application rate; it increased by 26.9% with an increase in the N application rate from N30 to N60 and by 16.4% with an increase from N60 to N90. The Zn rate administered also significantly influenced (P < 0.0001) the N uptake. An increase in the mean N uptake was observed when the Zn application rate was increased from ZnCHE-0 to ZnCHE-0.5 (with an increase of 32.6% in the mean N uptake) and from ZnCHE-0.5 to ZnCHE-1.0 (with an increase of 11.8% in the mean N uptake). There were, however, no significant differences between the N uptake obtained with the ZnCHE-1.0 and ZnCHE-1.5 treatments.

Fig. 2
figure 2

Effect of ZnCHE (Zn-DTPA-HEDTA-EDTA) at doses 0, 0.5, 1, and 1.5 kg ha−1 and N in form of pig slurry or urea (at doses 30, 60, and 90 kg ha−1) treatments on N and Zn uptake and grain protein concentrations of barley (Hordeum vulgare L.) grown under field conditions. Vertical bar at each of the data points represents the standard deviation from the means. Statistical differences at P < 0.05 (LSD test) are presented by different letters, according to the statistical interactions obtained

Full size image

As expected, the Zn uptake significantly increased (P < 0.0001) with increases in the Zn application rate. Increases of 23.1 and 7.2% were observed in mean Zn uptake, respectively, when the PS + ZnCHE-0.5 and urea + ZnCHE-0.5 treatments were compared to the ZnCHE-0 treatments. Moreover, increases of 59.4 and 25.9% were observed in the mean Zn uptake by changing from ZnCHE-0 to ZnCHE-1.5, applying PS and urea, respectively. Furthermore, increases in the N application rate significantly (P < 0.0001) increased the Zn uptake (with increases of 12.3% and 19.0% in the mean Zn uptake when the N application rate was increased from N30 to N60, applying PS and urea, respectively, or with increases of 22.2% and 46.6% in the mean Zn uptake when the N application rate was increased from N30 to N90, applying PS and urea, respectively).

An increase in the N application rate significantly increased mean grain protein concentrations. On the other hand, applying ZnCHE produced significantly (P < 0.0001) higher mean grain protein concentrations than the ZnCHE-0 treatments. Increases in mean grain protein concentrations of from 21.4 to 34.8% were observed when applying ZnCHE as opposed to ZnCHE-0 treatments. The protein concentration was not significantly affected by an increase in the Zn application rate from ZnCHE-1 to ZnCHE-1.5. Something similar occurred with total N uptake by grain, although the grain protein concentration was significantly affected by the different N sources employed, with higher mean grain protein concentrations being associated with treatments using urea than those applying PS.

The Fe, Mn, and Cu concentrations in several aerial parts of the plant (stem, leaf, and grain) are shown in Fig. 3. In the case of stem Fe concentration, significant (P < 0.0001) differences were observed between N sources. Greater increases in Fe levels were observed in the stem with PS applications than with urea treatments (22.98 and 16.67 mg Fe kg−1, respectively). Significant differences (the P values ranged from 0.0000 to 0.05) were observed between N application rates or N sources in leaf and grain Fe concentrations. As the statistical interaction between the N source and N application rate factors were significant (P < 0.05) for leaf and grain concentrations, we used a factor that combined both of them: N source × N application rates. Significant (P < 0.0001) differences were observed between the N source × N application rates factor in terms of Fe concentration in these parts of the plant. The highest concentrations were observed for the treatment with the highest N application rate (N90) and where PS was applied. Even so, there were no significant differences between the leaf or grain Zn concentrations obtained with PS + N30 and any of the urea treatments. Furthermore, significant differences (P < 0.05) were observed between N application rates or N sources in stem Mn concentrations. Significant differences were also observed between the N source × N application rates interaction in the stem Mn concentration. The highest mean concentrations were observed for the treatment with the highest N application rate (N90) and where PS was applied.

Fig. 3
figure 3

Fe, Mn, and Cu concentrations in leaf, stem, and grain, obtained with each treatment. The vertical line at each of the data represents the standard deviation from the mean. Statistical differences at P < 0.05 (LSD test) are presented by different letters (capital, italic, and roman letters indicate the differences between treatments in grain, stem, and leaf respectively) according to the statistical interactions obtained

Full size image

3.2 Zinc, Fe, Mn, and Cu Status of the Soil

The effect of the fertilizer treatment on the total Zn, DTPA-TEA- and Mehlich-3-extractable Zn, and Zn fractions in the soil at harvest are shown in Table 3. In total and available (Zn-DTPA-TEA and Zn-Mehlich-3) Zn concentrations significant differences were observed between Zn application rates. Significant differences were also observed between Zn application rates in all of the fractions except the CFeO, OM, and RES fractions. Zinc and N application rates and N source factors did not show any statistically significant interactions in any of the cases studied. As expected, when the Zn concentrations in the soil were statistically different, the highest Zn concentrations were observed with ZnCHE applied at 1.5 mg Zn kg−1 and the lowest Zn concentrations were obtained with ZnCHE-0.

Table 3 Analysis of variance (ANOVA) and effect of fertilizer treatment on total Zn, DTPA-TEA, and Mehlich-3 extractable Zn and Zn fractions in soil (mg kg−1) at the time of barley harvest
Full size table

The total available Fe, Mn, and Cu concentrations in the soil (DTPA-TEA and Mehlich-3) and distributions within the different soil fractions did not show any significant differences between Zn application rates (data not shown). Potentially available (DTPA-TEA and Mehlich-3) Fe concentrations in the soil showed significant (P < 0.05) differences between N application rates or N sources. Furthermore, Fe-WSEX + CAR showed significant (P < 0.001) differences between N application rates or N sources. The statistical interaction between N source and the N application rates factors were significant (P < 0.05) in all these cases. Significant (P between 0.008 and 0.046) differences were observed between N source × the N application rates interaction for Fe concentration in the soil. The highest mean concentrations were observed for the treatment with the highest N application rate (N90) and where PS was applied (Fe-DTPA-TEA, 5.08 mg Fe kg−1; Fe-Mehlich-3, 42.90 mg Fe kg−1; Fe-WSEX + CAR, 2.12 mg Fe kg−1). The lowest mean concentrations were observed for the treatments including urea (Fe-DTPA-TEA urea + N60 = 4.26 mg Fe kg−1; Fe-Mehlich-3 urea + N30 = 30.80 mg Fe kg−1; Fe-WSEX + CAR urea + N60 = 1.27 mg Fe kg−1). Even so, there were no significant differences between the potentially available Fe concentration obtained with PS + N30 and the urea treatments.

On the other hand, Mn-Mehlich-3 showed significant (P < 0.05) differences between N application rates or N sources. The statistical interaction between N source and the N application rate factors was significant (P < 0.05). The Mn-Mehlich-3 concentrations ranged from 61.75 mg Mn kg−1 to 76.08 mg Mn kg−1, urea + N60, and PS + N90, respectively. Conversely, the potentially available (DTPA-TEA and Mehlich-3) Cu concentration in the soil and its distribution within the different soil fractions did not show any significant differences between N sources.

4 Discussion

4.1 Influence of Combined Application of Zn and N on Yield, Biofortification, and Plant Quality

Under wet conditions, such as parts of the north of the Spain, with barley yields of between 3000 and 5300 kg ha−1, it is recommended to apply N at a rate of between 80 and 140 kg N ha−1. In contrast, in the arid drylands of Castilla-La Mancha, with average barley yields of around 2000 kg ha−1, the N application rate tends to be below 50 kg N ha−1. In irrigated areas, with yield levels above 5000 kg ha−1, the typical N application rate is around 125 kg N ha−1 (Lopez-Bellido 2009). In our experiment, there was an improvement in grain yield when ZnCHE fertilizer was added. This enhancement (9–11%) was similar to the mean yield increase observed when the N rate was increased from N30 to N60 or N90 (7–13%) (see Fig. 1). This improvement, which was evident following the application of Zn fertilization, could be attributed to an adequate Zn supply that might have affected metabolic and enzymatic activity. There were no significant differences between the grain yields obtained with the different treatments in which Zn was applied. The mean yields obtained with the combined applications of Zn and N were greater than those obtained when applying only N. Harapiak et al. (2000) reported that, in general, the barley yield response to micronutrients was 3% of the total response on the Canadian Prairies, while N was responsible for 73% of the increase in yield. Nitrogen is the key element for achieving high cereal yields. In our experiment, grain yield increased with the N application rate (e.g. the application of N90 + ZnCHE-0 increases 23% and 12.5% in mean grain yield respect to the treatment N30 + ZnCHE-0, for PS and urea, respectively). Plots fertilized using the lowest N application rate produced the lowest grain yields. Limited N availability affects the supply of carbohydrates to developing seeds and thereby reduces grain weight.

On the other hand, the application of PS rather than urea produced higher mean grain yields (see Fig.1). Increases in the N + Zn application rates had a great impact on grain yield, with mean values increasing by up to 30.3% with PS and by 23.1% with urea. This could have been due to an increase in the available nutrients and dissolved organic carbon (DOC) to the crop through the increase in their supply with the application of pig slurry. In our study, significant correlations and positive coefficients were observed between grain yield and the Zn, Fe, and Cu concentrations in grain (P < 0.05, n = 72). Zeidan et al. (2010) reported increases in grain yield with respect to the control for the application of Zn and Fe of 42% and 40%, respectively, in a wheat crop.

In our experiment, the combined application of N and Zn also had a significant, positive effect on grain Zn concentrations (see Table 2). The use of agronomic strategies to increase concentrations of mineral elements in edible plants is commonly known as Agronomic Biofortification. In countries whose populations exhibit a high incidence of micronutrient deficiencies, cereal-based foods generally represent a considerable proportion of the daily diet (Cakmak et al. 2010). According to Velu et al. (2014), the Zn concentration in wheat in many countries ranges from 20 to 35 mg kg−1. In our study, we observed a total Zn concentration in barley of between 37 and 72.4 mg Zn kg−1 and in grain we found Zn concentrations of between 17.4 and 27.5 mg Zn kg−1 (PS × N30 + ZnCHE-0 and urea × N90 + ZnCHE-1.5, respectively).

Erenoglu et al. (2011) and Kutman et al. (2010) observed that the N nutritional status of wheat plants affects root uptake and shoot transport, retranslocation from vegetative tissues to seeds, and the seed allocation of Zn. The redistribution of Zn from leaves to sinks and from roots to shoots is a mechanism that makes it possible to provide emerging organs and maturing seeds with appropriate concentrations of Zn in all cereals. Various authors have studied the mobility of metals within plants by calculating the TF (translocation factor); this is defined as the relationship between the micronutrient concentrations in two different parts of a given plant (Kabata-Pendias 2001; Podlesakova et al. 2001). A high TF value indicates greater mobility (Intawongse and Dean 2006; Almendros et al. 2015).

In our study, significant differences in TF were noted for the different Zn treatments (P < 0.0001 in TF grain/leaf, TF grain/stem, and TF grain/root; P < 0.001 in TF leaf/stem and P < 0.05 in TF stem/root) and N rates applied (P < 0.05 in TF stem/root, TF grain/stem, and TF leaf/stem). The highest TF stem/root values were observed with the ZnCHE-1.5 and N60 or N90 treatments. Kutman et al. (2010) reported that the positive interaction between N and Zn in cereals is due to improvements in root uptake and the translocation of Zn due to the presence of N. Several nitrogenous compounds, such as amino acids and nicotianamine, could be involved in the translocation of Zn from the root into the shoot. However, in our study, it was observed that the TF reached similar values with the N60 and N90 treatments when the highest Zn rate was added (ZnCHE-1.5).

In contrast, for TF grain/leaf, TF grain/stem, TF grain/root (P < 0.0001), and TF leaf/stem (P < 0.001), the highest values recorded in our study were observed with the ZnCHE-0 treatments. This seems to suggest that when the Zn concentration in plant is low, it assures the accumulation of Zn in grain or leaves.

In our experiment, N application rates did not cause any significant differences in TF grain/leaf. These results differed from those reported by Erenoglu et al. (2011) who, in a radiolabelled experiment conducted with 65Zn and carried out in wheat, found that plants with high levels of N translocated more 65Zn from flag leaves into seeds than those containing low levels of N. This difference could be explained by the fact that the N doses applied in our experiment were low (with approximate reductions of 70%, 40%, and 10% compared to the traditional application rates used in our area).

The Zn uptake in plants significantly increased at both of the Zn and N application rates (see Fig. 2). As Zn uptake is calculated as the product of grain Zn concentration and yield, the highest values for this crop were obtained with applications of PS. Combined applications of Zn and N also significantly increase the grain protein concentration and N uptake. Various authors have reported increases in protein content associated with combined applications of N and Zn, in different crops: rice (Guo et al. 2016), maize (Sajad et al. 2014), wheat (Cakmak et al. 2010), and fodder sorghum (Verma et al. 2005). Speciation and localization studies conducted on cereal grain indicate that Zn interacts with proteins in the grain and therefore grain proteins constitute a physiological sink for Zn (Cakmak and Kutman 2018). Zinc plays an indispensable role in protein synthesis and Zn deficiency impedes this process and affects the health and productivity of plants, resulting in low yields and poor plant quality. According to FEDNA (2016), grain protein concentrations of between 9.6 (“lower protein level”) and 11.3% (“upper protein level”) in barley grain are appropriate for feed use. In Spain, average analyses for the last 6 years show a tendency towards a reduction in the protein content of barley grain and an increase in the concentration of starch; this is probably due to the fact that N application has been limited by the Nitrates Directive. In our study, all the treatments with N90 reached these values when ZnCHE was applied (see Fig. 2). Moreover, treatments with N60 and ZnCHE-1 or ZnCHE-1.5 also achieved this range. Conversely, the only treatment involving the lowest N rate applied to reach this range was the urea × N30 + ZnCHE-1.5 treatment. It is noteworthy that in our study PS application affected the grain protein concentration, which decreased by 10% when PS was applied instead of urea. This may reflect a common characteristic of pig slurry: once it has been added to soil, the ammoniacal N in slurry quickly changes into nitric N and this may leach if plants do not assimilate it (Sanchez and Gonzalez 2005). Moreover, N from manure is easily lost via gaseous emissions (NH3, N2O, and NO).

4.2 Influence of Treatment Added on the Efficiency of Zn or N Use

To evaluate the relative effectiveness of the different treatments, we established the agronomic efficiency of Zn or N use (AEZn and AEN, respectively). These parameters were used to study the capacity of yield increase per kg of Zn or N applied. Significant differences between N or Zn application rates were observed in AEN (P between 0.009 and 0.0000). The factors studied did not show any statistically significant interactions in any of the cases. As shown in Fig. 4, it was possible to observe a (non-significant) tendency for AEZn to decrease with increases in the application rates of Zn and N and also with urea application. However, the mean AEN obtained with the application of Zn fertilizer was significantly (P < 0.001) greater than that obtained by only applying N. According to Cakmak and Kutman (2018), N and Zn act synergistically when they are both present in sufficient amounts. Arora and Singh (2004) reported the beneficial effect of Zn application on N availability, possibly as a result of the activation of a physiological process in which Zn acts as a co-enzyme and/or catalyst. Since the agronomic efficiency of N use (AEN) describes the capacity to increase the yield per kg of N applied, the highest AEN values for this crop were obtained with applications of the lowest N doses (N30 treatments). These results are in line with Tilman et al. (2002) who reported the highest efficiency of N fertilizer use being associated with the first increments in added N, with AEN decreasing with higher levels of N application.

Fig. 4
figure 4

Agronomic efficiency of Zn or N use (AEZn and AEN, respectively), Zn and N utilization obtained with each treatment. The vertical line at each of the data represents the standard deviation from the mean. Statistical differences at P < 0.05 (LSD test) are presented by different letters according to the statistical interactions obtained

Full size image

According to Prasad and Sinha (1981), the percentage of Zn used, or Zn utilization, by the crop is a decisive parameter for the relative effectiveness of any fertilizer treatment. In our study, significant differences were observed between N sources (P < 0.0001) or Zn application rates (P < 0.05). The factors studied did not show any statistically significant interactions in Zn or N utilization. As expected, in this crop the highest Zn utilization values were obtained with applications of the highest Zn doses (ZnCHE-1.5 treatments). There were differences between the results obtained from treatments with ZnCHE-1.0 and ZnCHE-1.5, but these were not significant. In addition, the application of Zn fertilizer produced significant (P < 0.0001) increases in N utilization with respect to the control treatment. The highest N utilization values were obtained with applications of the ZnCHE-1.0 and ZnCHE-1.5 treatments. According to several studies (Hu et al. 2003; Montoya et al. 2018), ethylenediamine-based chelating agents, such as DTPA and EDTA, may act as nitrification inhibitors. Nitrification inhibition caused by heavy metals, such as Zn, has also been addressed in the literature (Kapoor et al. 2015). The application of nitrification inhibitors is a strategy employed to increase the efficiency of N use in crops (Abalos et al. 2014).

The highest level of N utilization was obtained with the lowest N doses (N30 treatments); this result was in line with the AEN. The differences in Zn and N utilization between the different N sources applied (PS and urea) were consistent with the results obtained for Zn uptake in plant and grain protein concentration, which was, in itself, directly related to the N concentration in the plant.

4.3 Soil Status and Nutrient Availability to the Plant

The amount of Zn in soils and plants depends on the source applied, its behavior (fixation, leaching, transformation, availability), and several other plant parameters (uptake, transport, and accumulation). Various authors have assessed the mobility of metals by calculating soil-to-plant transfer factors (TFS-P) (Kabata-Pendias 2001; Prasad and Sinha 1981) this concept is defined as the relationship between the Zn concentration in the plant root and the total concentration of this element in the soil. A high TFS-P value indicates greater soil-to-plant mobility and availability, while a low value indicates a greater probability of there being a deficiency. In our study, significant differences between N application rates (P < 0.0001) or Zn application rates (P < 0.0001) were obtained for TFS-P. The mean TFS-P concentration increased with N application rates in the following order: N30 (0.31) < N60 (0.34) ≈ N90 (0.35). Chaudhry et al. (1977) and Singh and Singh (1981) observed higher Zn contents in plant tissues as a result of applying more N in response to an increase in Zn solubility. In contrast, the mean TFS-P concentration increased in the following order: ZnCHE-0 (0.24) < ZnCHE-0.5 (0.31) < ZnCHE-1 (0.38) < ZnCHE-1.5 (0.41). This Zn source: ZnCHE (Zn-DTPA-EDTA-HEDTA), was associated with large amounts of bioavailable Zn in the soil and also with high Zn concentrations resulting from the sum of the most labile fractions (water-soluble plus exchangeable fractions) in the soil with similar properties (Almendros et al. 2015). This behavior could be related to differences in the stability constant (K) of the chelates studied under these soil and plant conditions. The Zn sources that contain the most stable chelates (log K Zn-EDTA = 17.5 and log K Zn-DTPA = 19.5 with an ionic strength of 0.01 mol L−1) (Lindsay 1979; Smith et al. 2004) maintain greater amounts of Zn in their soil solution. In contrast, in Zn sources that contain fewer stable chelates, for example, Zn-HEDTA (log K Zn-HEDTA = 15.3 with an ionic strength of 0.01 mol L−1), the metal is retained by the soil components.

5 Conclusion

The results of our study indicate that N-Zn interaction influences both plant yield and its nutritional composition. The application of Zn-DTPA-HEDTA-EDTA not only increased the Zn concentration in the plant but also affected the amount of N in the plant. We suggest an application rate of 0.5 mg Zn kg−1 to achieve a great grain yield. However, it is recommendable to increase the dose of Zn applied in order to reach high nutritional values as this improves the Zn concentration in the plant or grain protein concentration. Higher Zn doses are also advisable to achieve high levels of Zn and N uptake in plants and to increase root-stem Zn translocation. In contrast, the 90 kg N ha−1 application rate was the most effective in terms of crop yield, grain protein concentration, and plant Zn and N uptake. These treatments would imply a reduction of 25% in the “rational N application rate” for barley in this calcareous Mediterranean soil. Pig slurry application produced a higher mean grain yield, Zn uptake, and Zn utilization but a lower grain protein concentration than urea applications. Our study suggests that the use of combined fertilization (N + Zn-DTPA-HEDTA-EDTA) could be a beneficial strategy in order to obtain greater nutritional quality and higher crop production in both economic and sustainable terms.