Quality, yield and nitrogen fixation of faba bean seeds as affected by sulphur fertilization

Abstract

Faba bean (Vicia faba L.) is an important source of plant protein for humans and animals; however, nutritional value of seeds is notoriously deficient in sulphur (S)-containing amino acids. In this article, the effect of S fertilization on faba bean's capability of N₂ fixation, grain yield and chemical characteristics in terms of protein fractions, fatty acids and minerals composition is reported. A randomized, complete block design with three replicates was used, and three S applications (0, 30 and 60 kg ha⁻¹, respectively) for faba bean were performed. The S fertilization was split into two applications: 50% before sowing and 50% in the beginning of March as K₂SO₄. At the same time, both the legume and oat crops were fertilized uniformly with 10 kg N ha⁻¹ as ¹⁵N NH₄ ¹⁵NO₃ (10% ¹⁵N atomic excess) in solution form. In a Mediterranean climate under optimal spring rainfall situations, faba bean produced high yield of grain and protein. Sulphur application resulted in an increase in overall plant yield and N₂ fixation. In addition, S fertilization enhanced the protein quality, increasing its degradable fraction. Fertilizing faba bean with 30 kg ha⁻¹ of S resulted in a more appropriate dose in order to obtain a quantitative and qualitative crop improvement. From our findings, it can be concluded that S fertilization to faba bean should be recommended to soils with suboptimal S levels to obtain maximum seed and protein yields.

Keywords:

• Fatty acid
• legume
• N fixed
• protein fractions
• seed production

Introduction

Faba bean (Vicia faba L.) is an important legume crop in Mediterranean countries and many parts of the world. Its seeds exhibit high levels of protein [28–36% of seed dry matter (DM)]. And it is used in a popular breakfast food and also used as a vegetable green, resh or canned. Also, it is an important crop for soil improvement and used as break crop in cereal rotation to keep the soil fertile and productive through nitrogen fixation (Mona et al., 2011).

Sulphur (S) is one of the elements known to be essential for the legume-rhizobium system with specific physiological and biochemical roles. The S demand of legume crops is higher than that of cereal crops. Studies on different legumes have shown that the concentration of the S-containing amino acids was markedly declined with decreasing S supply (Gaylor & Sykes, 1985). Sulphur fertilization was also found to increase N accumulation and yield of legumes on S-deficient soils (Eppendorfer & Bille, 1992). Scherer and Lange (1996) found a significant increase of DM yield and N₂ fixation in faba bean plants. These authors also reported that S fertilization significantly increased the number and weight of nodules per pot as well as the specific nitrogenase activity. In addition to being an important food crop, faba bean also plays an important role in cereal-cropping systems because of its N₂-fixing capacity. Substantial yield increments have been reported for wheat planted after faba bean in many countries (Amanuel et al., 2000). Faba bean among the grain legumes is reported to derive the highest percentage of N from the atmosphere (Hardarson et al., 1991). Amanuel et al. (2000) reported the per cent of N derived by faba bean from the atmosphere (%Ndfa) to vary from 62 to 74% with corresponding N₂ fixed from 152 to 189 kg N ha⁻¹ under the semi-humid conditions.

Sulphur is becoming deficient in soils due to the introduction of high-yielding varieties, use of high-grade S-free fertilizers and reduced emission of S from industrial units (Scherer, 2009). It is therefore very important to study the response of different crops to S application under different agro-ecological conditions and cropping systems and its effect on soil nutrient balances. Sulphur is a vital part of the ferredoxin, an iron-sulphur protein occurring in the chloroplasts. Ferredoxin has a significant role in NO₂ and sulphate reduction, the assimilation of N by root nodule bacteria and frees living N-fixing soil bacteria (Scherer et al., 2008). Sulphur application has an important potential of increasing the amount of N fixed by legumes, thus improving the fertility status of soil (Habtemichial et al., 2007). However, to date studies on the effects of S fertilization on grain legume yield and quality as well as N fixation are limited (El-Fiel et al., 2002). Therefore, a field trial was conducted in southern Italy under Mediterranean conditions to evaluate the effect of S fertilization on faba bean's capability of N₂ fixation, grain yield and chemical characteristics in terms of protein fractions, fatty acids (FAs) and minerals composition.

Materials and methods

Field trial and treatments

A field trial was carried out in southern Italy at Gaudiano di Lavello – Potenza (41° 06′ N; 15° 51′ E; 145 m above sea level) on a sandy-clay soil, characterized as sub-alkaline, low in total N (0.77‰ Kjeldahl method) and high in available P (83 mg kg⁻¹; Olsen method), exchangeable K (482 mg kg⁻¹; BaCl₂; Triethanolamine (TEA) method) and low soluble S as SO₄-S (4.4 mg kg⁻¹; KH₂PO₄ as extractant; Woomer et al., 2001). The experimental site was characterized by a summer dry climate with a total annual rainfall of 560 mm distributed from autumn to spring and a mean temperature of 16°C. During the experimental period (November–June), the total rainfall was 402 mm, and temperatures did not show any significant variation from the average.

Faba bean (Vicia faba L. cv. Prothabat 69) and oat (Avena sativa L. cv. Argentina) were sown on November maintaining a row distance of 20 cm for both species at seed rate of 300 and 180 kg ha⁻¹, respectively. A andomized, complete block design with four replicates and a plot area of 15 m² was utilized. The macroplot sizes were 3×5 m, whereas microplot sizes were 1.5×1 m; each microplot was located in the centre of individual macroplots. Three S fertilization levels (0, 30 and 60 kg ha⁻¹) were applied. The S fertilization was split into two applications: 50% before sowing and 50% in the beginning of March as K₂SO₄. At the same time, both the legume and oat crops were fertilized uniformly with 10 kg N ha⁻¹ as ¹⁵N NH₄ ¹⁵NO₃ (10% ¹⁵N atomic excess) in solution form (fertilizer dissolved in 1 L H₂O). The macroplots also received 10 kg N ha⁻¹ as unlabelled NH₄ NO₃. The K contained in K₂SO₄ was compensated with KCl for the plots not treated with S. Faba bean crop was grown under rain-fed conditions, and no irrigation was used. At physiological maturity, plants were harvested from 1 m² microplots so that the outer rows were not used for both faba bean and the reference oat crop. Plants were divided into straw and pods or panicles. These plant parts were oven-dried at about 70°C for 48 h. The plant material was ground to pass through a 0.2 mm sieve. Total N and%¹⁵N atomic excess (a.e.) of plant samples were analysed at Iso-Analytical Limited (Cheshire, UK) using elemental isotope ratio mass spectrometry (EA-IRMS) analysis. The %¹⁵N excess was calculated by the difference of the atomic %¹⁵N in the plant material (straw in faba bean and oat, pods in faba bean and panicles in oat) and that of the natural abundance in the atmosphere (0.3663%). The N fixation in faba bean was calculated using the ¹⁵N-isotope dilution method as described by Hardarson et al. (1991), using Equation 1; and the amount of N₂ fixed was determined from the product%Ndfa and N yield for each replication, and the average was then calculated using Equation 2:

1

2

Chemical analysis

Samples of each treatment were ground in a hammer mill with a 1-mm screen and analysed in duplicate for DM, ash, crude protein (CP, Kjeldahl N×6.25) and crude fat (ether extract with previous hydrolysis) according to the procedures described by the AOAC (2000). Neutral detergent fibre (NDF) and acid detergent fibre (ADF) were determined according to Van Soest et al. (1991), and were corrected for residual acid insoluble ash. Acid detergent lignin was determined by the method described in the study of Van Soest and Robertson (1985).

Representative samples of oven-dried seeds weighing 1 g were placed in a muffle furnace at 550°C for four hours for total ash determination. The ash was wet with sulphuric and perchloric acids and diluted with distilled water (AOAC, 2000). Sodium and K concentrations were determined by flame photometry (Agilent 6890 Series GC), P was determined according to the AOAC (2000) method (931.01) by absorption spectrophotometry and concentrations of Ca and Fe were determined by atomic absorption spectroscopy (Jorhem, 2000). All results were expressed on DM basis. In preparation for the analysis of FA composition, samples of faba bean (1 g each) were freeze-dried and then ground. Methyl heptadecanoate (No. 51633, Fluka, USA) was dissolved into n-hexane (1 mg ml⁻¹) as an internal standard. Methyl esters of the FA were prepared (Sukhija & Palmquist, 1988); samples (300 mg each) and 5 ml internal standard were incubated (2 h at 80°C) with methanolic acetyl chloride in a total volume of 9 ml. After cooling to room temperature, 7 ml of 7% (w/v) K₂CO₃ was added with mixing, and then the organic phase was collected after centrifuging at 1500 g for 2 min at 4°C. FA methyl esters were fractionated over a CP-SIL883 column (100 m×0.25 mm i.d., film thickness 0.20 µm fused silica; Varian, Palo Alto, CA, USA) in a Shimadzu (Model 2GC17A) gas chromatograph with a Hewlett–Packard HP 6890 gas system and using flame ionization detection. Helium was used as carrier gas at a constant flow rate of 1.7 ml min⁻¹. The oven temperature was programmed as follows: 175°C, held for 4 min; 175–250°C at 3°C per minute; and then maintained for 20 min. The injector port and detector temperatures were 250°C. Samples (1 µl) were injected with an autosampler. Output signals were identified and quantified from the retention times and peak areas of known calibration standards. Composition was expressed as percentages of the total FA.

Fractionation of crude protein

Fractionation of CP was carried out by the Cornell Net Carbohydrate and Protein System (Sniffen et al., 1992). According to this system, CP was partitioned into three fractions: A fraction was non-protein N×6.25; B fraction was true protein and C fraction was unavailable protein. The B fraction was further divided into three sub-fractions B₁, B₂ and B₃ of rapid, intermediate and slow rates of ruminal degradation, respectively. CP fractions A and B₁ were soluble in borate-phosphate buffer, CP fraction of B₂ was insoluble in the buffer but soluble in neutral detergent solution and the B₃ fraction was insoluble in the buffer and in the neutral detergent, but soluble in acid detergent. Crude protein C fraction was insoluble in acid detergent (acid detergent insoluble protein, ADIP) and contained protein associated with lignin, tannin–protein complexes and Maillard products that are known to be highly resistant to microbial enzymes. Precipitated true protein, buffer insoluble protein, neutral detergent insoluble protein (NDIP) and ADIP were analysed as described by Licitra et al. (1996). The A fraction was calculated as the difference between the total CP and precipitated true protein determined by Kjeldahl analysis of the residue resulting after precipitation with tungstic acid followed by filtration. The CP fraction B₁ was estimated as true protein minus buffer insoluble protein, fraction B₂ as buffer insoluble protein minus NDIP and fraction B₃ by subtracting the ADIP (fraction C) from the NDIP.

Statistical analysis

Data were analyzed by analysis of variance using CoStat version 1.03 Software (CoHort Software Inc., Monterey, CA, USA). The statistical analysis was applied to data following the one-way ANOVA design. Differences among treatment means for significant effects were detected using LSD procedure. Unless otherwise stated, significance was declared at P<0.05.

Results

There was a significant effect of S application on grain and protein yields of faba bean. As reported in Table I, faba bean resulted in higher grain production when S₃₀ was applied compared to the S₀ treatment (4.65 vs. 4.09 t ha⁻¹; P=0.018). Data also showed that grain yield of S₆₀ resulted only in a better trend with respect to S₃₀ application. Furthermore, when S₃₀ fertilization rate was used, faba bean grains produced more protein compared with the unfertilized S₀ treatment (1237 vs. 1087 kg ha⁻¹; P=0.019). However, the higher sulphur application rate to faba bean leads to a slight increase in protein yield compared with S₃₀ fertilization rate. The increase in faba bean protein yield in response to the different sulphur fertilization rates was due to the improvement in grain yield, since seeds’ CP content did not result in a significant variation after S applications. In fact, CP levels in faba bean varied slightly from 26.52 to 26.84% (S₀ and S₆₀, respectively; P>0.05).

Table I. Effect of sulphur fertilizer application rate on grain yield, protein production, Ndfa and N₂ fixation of faba bean.

	Treatments^1
Item	S0	S30	S60	Mean	SEM	P-value	LSD0.05
Grain yield, t ha^−1	4.09^b	4.65^a	4.87^a	4.54	0.19	0.018	0.43
Protein yield, kg ha^−1	1087^b	1237^a	1307^a	1210	54.9	0.019	126.6
Crude protein,%	26.52	26.63	26.84	26.66	0.14	0.111	ns
Ndfa,%	90.9	89.9	89.0	89.9	1.99	0.759	ns
N2 Fixed, kg ha^−1	225.3^b	253.2^ab	274.8^a	251.1	7.32	0.042	35.4
ns: not significant (P>0.05); Ndfa: N derived from the atmosphere; ^1S0, 0 kg ha^−1 of sulphur fertilizer application; S30, 30 kg ha^−1 of sulphur fertilizer application; S60, 60 kg ha^−1 of sulphur fertilizer application; ^a, ^bMeans within a row with no common superscript differ significantly (P<0.05).

The %¹⁵N a.e. in the reference crop oat was always greater than that of faba bean on a whole-plant basis, a clear indication that faba bean had fixed N₂ from the atmosphere. The per cent of N derived from the atmosphere for faba bean (%Ndfa) was not significantly increased with S fertilization, reporting an overall mean of 89.9% (Table I).

The amount of N₂ fixed was found to be significantly (P < 0.05) improved by S fertilization. The total amount of N₂ fixed by faba bean (seed and straw) ranged from 225.3 kg ha⁻¹ for S₀ treatment to 253.2 and 274.8 kg ha⁻¹ (in S₃₀ and S₆₀, respectively), corresponding to an average increase of 12.4 and 22.0%, relative to the control.

Our results indicated that the different S fertilization rates did not influence the chemical composition of faba bean grains, as shown in Table II. However, in spite of the similar CP content of faba bean grains obtained after different S applications, data indicated a significant variation in terms of protein fractionation (Table III). In particular, S addition linearly increased the protein fractions A and B₃, and decreased the fractions B₁, B₂ and C (P<0.05).

Table II. Effect of sulphur fertilizer application rate on nutritional composition of faba bean grains.

	Treatment^1
Item	S0	S30	S60	Mean	SEM	P-value
Chemical composition, %
DM	88.37	88.21	88.31	88.30	0.51	0.745
Crude protein	26.52	26.63	26.84	26.66	0.42	0.095
Crude fat	1.83	1.82	1.84	1.83	0.09	0.115
Ash	4.71	4.65	4.59	4.65	0.13	0.236
NDF	31.58	31.36	31.84	31.59	0.39	0.177
ADF	11.21	10.95	11.38	11.18	0.16	0.108
ADL	0.58	0.59	0.48	0.55	0.03	0.085
Acid insoluble ash	0.27	0.21	0.22	0.23	0.02	0.154
Hemicellulose	18.89	18.96	18.95	18.93	0.33	0.106
Cellulose	11.84	11.62	12.17	11.88	0.21	0.096
DM, dry matter; NDF, neutral detergent fibre; ADF, acid detergent fibre; ADL, acid detergent lignin. ^1S0, 0 kg ha^−1 of sulphur fertilizer application; S30, 30 kg ha^−1 of sulphur fertilizer application; S60, 60 kg ha^−1 of sulphur fertilizer application.

Table III. Effect of sulphur fertilizer application rate on total protein content and protein fractions of faba bean grains.

	Treatment^1
Item	S0	S30	S60	Mean	SEM	P-value
Crude protein (g kg^−1 DM)	265.2	266.3	268.4	266.6	42.2	0.095
Protein fractions ^1 (g kg ^−1 CP)
A (NPN)^2	198.3^b	202.1^b	206.3^a	202.2	30.5	0.039
B1 (TSP)^2	190.3^a	188.2^a	181.2^b	186.6	20.8	0.025
B2 (NDSP)^2	242.7^a	239.1^b	235.3^c	239.0	32.4	0.041
B3 (NDIP)^2	315.7^c	319.3^b	328.7^a	321.2	35.7	0.033
C (ADIP)^2	53.1^a	51.7^a	48.7^b	51.2	5.5	0.029
^1S0, 0 kg ha^−1 of sulphur fertilizer application; S30, 30 kg ha^−1 of sulphur fertilizer application; S60, 60 kg ha^−1 of sulphur fertilizer application; ^2A, non protein-N (NPN); B1, true soluble protein (TSP); B2, neutral detergent soluble protein (NDSP); B3, neutral detergent insoluble protein (NDIP); C, acid detergent insoluble protein (ADIP). ^a–cMeans within a row with no common superscript differ significantly (P<0.05).

Table IV shows the effects of S fertilization on the major mineral elements in faba bean grains. Data showed that with increasing S fertilization rates the Ca, Na and K levels in faba bean seeds resulted in an increase (P<0.05) when compared with control unfertilized treatment. Conversely, the P and Fe grain contents were not influenced by the different fertilization treatments.

Table IV. Effect of sulphur fertilizer application rate on major mineral elements in faba bean grains.

	Treatment^1
Item	S0	S30	S60	Mean	SEM	P-value
Mineral elements, g kg ^−1 DM
Ca	1.41^a	1.82^b	1.82^b	1.68	0.08	0.034
P	3.90	4.21	3.72	3.94	0.19	0.061
Na	0.33^a	0.39^a	0.44^b	0.39	0.05	0.043
K	9.98^a	10.09^b	12.01^c	10.69	0.41	0.032
Fe	0.07	0.06	0.06	0.06	0.02	0.095
^1S0, 0 kg ha^−1 of sulphur fertilizer application; S30, 30 kg ha^−1 of sulphur fertilizer application; S60, 60 kg ha^−1 of sulphur fertilizer application; ^a–cMeans within a row with no common superscript differ significantly (P<0.05).

The results of FA analysis (% on total FA) of faba bean seeds obtained under different S fertilization are given in Table V. Overall, total saturated FA (SFA) decreased significantly from 24.41 to 18.86% (S₀ and S₃₀, respectively), whereas S₆₀ application did not lead further variation (18.47%) compared with S₃₀ rate. In particular, among the individual SFA, significant decrease was observed for myristic, heptadecanoic, stearic and tetracosanoic acids when S fertilization increased up to S₃₀. No effects of S fertilization were reported for total monounsaturated FA in faba bean grains that in overall represented about 25% of total FA. Conversely, total polyunsaturated FA (PUFA) content increased significantly from 51.35 to 56.12% (S₀ and S₃₀, respectively); the higher S application rate did not resulted in a further increase of PUFA level compared with the intermediate S dose. Particularly, in our cultivar of faba bean seeds, the main PUFA was the n-6 linoleic acid representing about the 90% of total PUFA.

Table V. Effect of sulphur fertilizer application rate on fatty acid (FA) composition of faba bean grains.

	Treatment^1
Item	S0	S30	S60	Mean	SEM	P-value
Fatty acids (% on total FA)
12:0 Lauric	0.02	0.02	0.01	0.02	0.01	0.255
14:0 Myristic	6.61^a	0.50^b	0.49^b	2.53	0.11	0.014
15:0 Pentadecanoic	0.10	0.27	0.26	0.21	0.05	0.101
16:0 Palmitic	11.93^b	13.26^a	13.30^a	12.83	0.21	0.042
16:1 n-7 Palmitoleic	0.01	0.04	0.09	0.05	0.02	0.095
17:0 Heptadecanoic	0.52^a	0.19^b	0.17^b	0.29	0.04	0.052
18:0 Stearic	3.01	2.83	2.80	2.88	0.09	0.069
18:1 cis9 n-9 Oleic	22.31^c	23.24^b	23.87^a	23.14	0.38	0.045
18:2 n-6 Linoleic	44.85^c	49.82^b	50.07^a	48.25	0.45	0.014
18:3 n-3 α-Linolenic	0.46^a	0.11^b	0.01^c	0.19	0.04	0.011
18:3 n-6 γ-Linolenic	4.94	5.03	4.96	4.98	0.12	0.052
20:1 n-11 Eicosenoic	1.94	1.75	1.65	1.78	0.08	0.069
20:4 n-6 Arachidonic	1.10	1.16	0.87	1.04	0.05	0.051
22:0 Docosanoic	1.01^a	0.80^b	0.84^b	0.88	0.13	0.039
24:0 Tetracosanoic	1.21	0.99	0.60	0.93	0.15	0.058
Σ SFA	24.41^a	18.86^b	18.47^b	20.58	0.28	0.021
Σ MUFA	24.26	25.03	25.61	24.97	0.37	0.117
Σ PUFA	51.35^b	56.12^a	55.91^a	54.46	0.49	0.012
SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; ^1S0, 0 kg ha^−1 of sulphur fertilizer application; S30, 30 kg ha^−1 of sulphur fertilizer application; S60, 60 kg ha^−1 of sulphur fertilizer application; ^a–cMeans within a row with no common superscript differ significantly (P<0.05).

Discussion

The present work demonstrated the positive effect of S fertilization on grain yield, protein production and total amount of N₂ fixed by faba bean. Fertilizing of faba bean with S was found to increase grain yield and N₂ fixation when the soil contained suboptimal amounts of sulphur (Eppendorfer & Bille, 1992; Scherer, 2001).

Sulphur is reported to enhance the photosynthetic assimilation of C and N in crops (Ahmad & Abdin, 2000), which in turn increases the DM and grain yield, as 90% of plant's dry weight is considered to be derived from products formed during photosynthesis (People et al., 1980). Leguminous plant species require a relatively higher amount of sulphur compared to small cereals, due to their increased concentration of proteins and the concentration of methionine and cystein in seeds (Gaylor & Sykes, 1985). In our study, the amount of N₂ fixed and %Ndfa in faba bean were higher if compared with other earlier trials (Evans et al., 1989; Habtemichial et al., 2007). However, many studies conducted under similar environmental conditions have reported comparable N₂ fixation and %Ndfa (Rennie & Dubetz, 1986; Beck et al., 1991; Turpin et al., 2002). Almost certainly, this last result could be attributed to the low N content of the soil of the experimental site. In fact, the contribution of symbiotic nitrogen fixation (SNF) to overall nitrogen acquisition by legumes (%Ndfa) is known to be affected by mineral N availability in the soil (Voisin et al., 2002). Indeed, the negative effect of nitrate on SNF, even at the smallest concentrations, has been reported by several authors (MacDuff et al., 1996 for white clover; Waterer & Vessey, 1993 for pea).

In our study, fat content in faba bean was not influenced by S fertilization resulting in overall 1.8% on DM. This value is in agreement with El Tinay et al. (1989) who found fat content in faba bean ranging from 1.1 to 2.2%. Conversely, Elsiddig and Abdulhafize (1997) reported fat values ranging from 0.7 to 0.92% that resulted in a decrease when compared with our data. However, the same authors found no difference in fat content when faba bean was fertilized with an S rate of 25 kg ha⁻¹ compared with unfertilized control plants (Elsheikh & Elzidany, 1997).

Up to now, data on protein fractions determined according to the CNCPS system (Sniffen et al., 1992) of faba bean grains are lacking, so the determination of protein quality is useful, given the first-hand idea about the nutritive value and the degradability properties of grains under examination. The undegraded and indigestible CP fraction C made up some 51 g kg⁻¹ of total CP. These results illustrate that faba bean CP is likely to be largely degraded in the rumen (Infascelli et al., 1995; Tufarelli et al., 2012). The available CP fraction should include the majority of CP fraction B₃ and a small proportion of CP fraction B₂. The CP fraction B3 presented a higher proportion of total insoluble CP (fractions B₂, B₃ and C) (Caballero et al., 2001). In our study, B₃ fraction represented the higher proportion of insoluble proteins, especially when faba bean grains were fertilized with 60 kg ha⁻¹ of sulphur. As CP fraction B₃ represents CP insoluble in NDF but soluble in ADF, the proportion of CP fraction B₃ is linked to NDF (Elizalde et al., 1999). As a whole, CP fraction A increased and CP fraction B₁ decreased from unfertilized grains to the higher sulphur fertilization rate. Both soluble fractions, the instantaneously fermented (CP fraction A) and the rapidly fermented (CP fraction B₁), represented together an overall value of 389 g kg⁻¹ of total CP.

In the present study, the experimental treatments affected FA profile of faba bean. We could not locate any literature concerning the effects of sulphur fertilization on FA composition of faba bean seeds; therefore, this subject should be considered in new investigations.

The present findings are confirmed by Abramovic and Abram (2005) who find that the differences in the composition of the FA in seed oil can be caused not only by the different cultivars, but also by environmental conditions. Zubr (2003) similarly concluded that the variations are due to the combined effects of climatic and soil conditions on the crop. This study revealed an effect of nitrogen/sulphur fertilizer application on oil content and FA composition. The highest N dose significantly reduced the oil content of seeds in comparison with the lowest one applied with the same S rate. The S fertilization caused the highest contents of both palmitic and oleic acids. The concentration of linolenic acid, which is the important FA with respect to nutrition in humans, remained unaffected by the fertilizer treatments applied. This indicates that oil of our faba bean variety have a good level of unsaturation and a high linoleic content as also reported by Hossain and Mortuza (2006) on local and exotic varieties of faba beans.

Faba bean contains appreciable amount of essential minerals (Sammour, 1985). In the present study, the five major mineral elements (Ca, P, Na, K and Fe), contents of faba bean, were determined. Our grains reported values comparable to those found in other previous research works (Hossain & Mortuza, 2006; Mona et al., 2011). The main significant effects with S fertilization were obtained for Ca, Na and K that increased with the S application. In particular, the highest variations were found for K content that ranged from 9.98 to 12.01 g kg⁻¹ of DM.

In conclusion, in a Mediterranean climate under optimal spring rainfall situations, faba bean produced high yield of grain and protein. Sulphur application resulted in an increase in overall plant yield and N₂ fixation. From the current results it can be concluded that S fertilization to faba bean should be recommended to soils with suboptimal S levels to obtain maximum seed and protein yields. In addition, S fertilization enhanced the protein quality increasing its degradable fraction. Finally, fertilizing faba bean with 30 kg ha⁻¹ of S resulted in a more appropriate dose in order to obtain a quantitative and qualitative crop improvement.