Article pubs.acs.org/JAFC
Cite This: J. Agric. Food Chem. 2018, 66, 831−841
Impact of Conventional and Integrated Management Systems on the Water-Soluble Vitamin Content in Potatoes, Field Beans, and Cereals Sabine Freitag,*,† Susan R. Verrall,† Simon D.A. Pont,† Diane McRae,† Julia A. Sungurtas,† Raphael̈ le Palau,† Cathy Hawes,† Colin J. Alexander,‡ J. William Allwood,† Alexandre Foito,† Derek Stewart,†,§ and Louise V.T. Shepherd† †
Environmental and Biochemical Sciences, The James Hutton Institute, Invergowrie, Dundee DD2 5DA, UK Biomathematics and Statistics Scotland, Invergowrie, Dundee, DD2 5DA, UK § School of Engineering and Physical Sciences, Heriot-Watt University, Edinburgh, EH14 4AS, UK ‡
S Supporting Information *
ABSTRACT: The reduction of the environmental footprint of crop production without compromising crop yield and their nutritional value is a key goal for improving the sustainability of agriculture. In 2009, the Balruddery Farm Platform was established at The James Hutton Institute as a long-term experimental platform for cross-disciplinary research of crops using two agricultural ecosystems. Crops representative of UK agriculture were grown under conventional and integrated management systems and analyzed for their water-soluble vitamin content. Integrated management, when compared with the conventional system, had only minor effects on water-soluble vitamin content, where significantly higher differences were seen for the conventional management practice on the levels of thiamine in field beans (p < 0.01), Spring barley (p < 0.05), and Winter wheat (p < 0.05), and for nicotinic acid in Spring barley (p < 0.05). However, for all crops, variety and year differences were of greater importance. These results indicate that the integrated management system described in this study does not significantly affect the water-soluble vitamin content of the crops analyzed here. KEYWORDS: barley (Hordeum vulgare L.), field beans (Vicia faba L.), integrated management, liquid chromatography−triple quadrupole mass spectrometry, potato (Solanum tuberosum L.), water-soluble vitamins (WSVs), wheat (Triticum aestivum L.)
■
systems on nutritional value,4−9 but also yield and cellular processes9−12 mainly in potatoes, while the impact of lower input (sustainable, organic, integrated) farming practices on crop nutritional value has rarely been investigated.13,14 Conventional, integrated, sustainable, and organic cultivation strategies have different goals in relation to crop yield, land and pesticide use, and also environmental impact. While conventional agricultural practices utilize high yield crop varieties, chemical fertilizers and pesticides, irrigation and mechanization, sustainable farming agricultural practices are designed to promote environmental health and the social and economic equity of a region. Sustainable agriculture practices are hard to define as conditions can vary greatly depending on the crop, environment, and issues important to a region.13 The term “nutritional value” as such has been defined as the chemical composition of food, in particular the amounts of key compounds that are essential for functioning of human organisms.15 Vitamins are a broad group of organic bioactive compounds that are minor but nutritionally essential constituents of food. Generally, vitamins can be divided into fat soluble (vitamin A, D, E, and K1) and water-soluble vitamins (WSVs), which include the B-group vitamins thiamine (B1),
INTRODUCTION The agricultural policies and strategies of the UK1 and EU2 are geared toward a shift to a more sustainable use of resources and conservation of farmland biodiversity. Therefore, it is important to understand the impact of the environment, crop management, and soil fertility on both the yield and nutritional quality of economically important crops in Scotland. Currently, the long-term, systems-level impact of changes in management, especially in terms of sustainable treatment strategies, has been little studied. There clearly is a need for balancing environmental management while maintaining crop yield and nutritional quality. In 2009, The James Hutton Institute set up a long-term experimental platform3 for cross-disciplinary research on the use of both integrated and conventional management systems in agricultural ecosystems. The platform is based on a framework for designing and testing cropping systems that hopes to optimize the balance between crop yield and product quality on one hand, with biodiversity and ecosystem services on the other (in essence, attempting to assess the balance between economic and environmental demands). The aim is to reduce the environmental footprint of crop production by minimizing the use of fossil-fuel-derived inputs and maximizing the benefits from renewable resources. The long-term, wholesystems approach adopted at the platform is essential if the potential conflicts between food production and environmental health are to be reconciled. So far, much of the literature has focused on the effect of organic versus conventional cultivation © 2017 American Chemical Society
Received: Revised: Accepted: Published: 831
August 1, 2017 November 17, 2017 December 19, 2017 December 19, 2017 DOI: 10.1021/acs.jafc.7b03509 J. Agric. Food Chem. 2018, 66, 831−841
Article
Journal of Agricultural and Food Chemistry
Figure 1. Platform layout and rotation of each crop across the six different experimental fields for 2011−2015. Each field is divided by a grass buffer strip (dark lines) with integrated management (I) on one-half and conventional management (C) on the other. In each half, five different varieties were grown in 18 m strips (faint lines) to determine varietal responses to the two contrasting management systems. In each strip, crop samples were harvested immediately prior to harvest from five permanently marked GPS locations (dark circles), except for the field beans, where five aliquots of beans were removed postharvest, from the end of each variety strip. Although Winter oilseed rape was included in the experimental plots, it was not analyzed with respect to water-soluble vitamin content. Source: http://csc.hutton.ac.uk/.
flour, tomato pulp, kiwi to semicoarse wheat flour, wheat bread, and toasted wheat bread.30,31 Liquid chromatography−triple quadrupole mass spectrometry has been optimized as a rapid procedure for multivitamin quantification in combination with an optimized extraction procedure.30,31 For this study, a method by Nurit et al.31 has been adapted to investigate whether our integrated management system impacts upon the content of five WSVs in a range of different crop matrices including field beans (Vicia faba L.), Spring and Winter barley (Hordeum vulgare L.), Winter wheat (Triticum aestivum L.), and potato (Solanum tuberosum L.)the latter of which was also analyzed for vitamin C content.
riboflavin (B2), pyridoxine (B6), nicotinic acid (B3), pantothenic acid (B5), and vitamin C. These act as coenzymes and are therefore essential for a range of different metabolic processes.16 Most vitamins, apart from vitamin D and nicotinic acid, are essential in the human diet as the body cannot synthesize them. With the exception of vitamin B12, only plants have the ability to synthesize B vitamins.17 Postharvest processing (such as drying, storage, dehulling, milling, soaking, blanching, fermentation, and cooking) can impact the levels of WSVs in various crops;18,19 however, the number of studies on the effects of preharvest conditions (particularly different agricultural management practices) on content of WSVs have been limited. Most studies in the literature, on the impact of different cropping systems on WSV contents, have focused on vitamin C content in some fruits, vegetables, and eggs.8,20−24 While vitamin C is generally the most studied vitamin, little data have been published on B vitamin levels in food crops. The major challenge previously in quantifying WSVs has been sensitivity, as the low levels in crops (such as cereals) require very sensitive methods,25 and their simultaneous quantification of these diversely different chemistries. Previous methods have mainly focused on the quantification of single vitamins, e.g. Martins-Junior et al.,26 and methods range from microbiological assays,17 Biosensor/ELISA27 to chromatographic procedures with UV or fluorescence detection.28,29 These extraction and detection methods have several drawbacks including the number of B vitamins quantified and lack of both precision and sensitivity. Only recently have advances been made regarding increased sensitivity and simultaneous detection of WSVs in various food matrices ranging from maize
■
MATERIAL AND METHODS
Experimental Design and Preparation of Sample Material. The Balruddery Farm platform, Dundee,3 comprises a 42 hectare (ha) contiguous block of six arable fields, northeast Scotland (56.48 latitude, 3.13 longitude). Balruddery Farm is a 178 ha arable farm, 67 to 163 m above sea level on the south facing slopes of the Sidlaw Hills. The farm is typical of temperate Atlantic maritime arable environments, with an average annual rainfall of 800 mm, an average annual accumulated temperature of 1100−1375 day °C (above 5.6 °C), and a mean annual potential water deficit of 50−75 mm. The area is moderately exposed (2.6−4.4 m second−1 wind speed) and has moderate winters of 50−110 day °C of accumulated frost. The soils are imperfectly draining Balrownie Series32 with an average pH of 5.7. Topsoil depths range from 25 to 40 cm, textures from sandy loam to sandy silt loam, and stone contents of 10−20% volume. Each of the six fields are divided in half, and the integrated and conventional management systems randomly allocated to each field half in 2010, at the start of the first rotation (Figure 1).3 These systems then remain in place for the duration of the experiment to allow detection of a build832
DOI: 10.1021/acs.jafc.7b03509 J. Agric. Food Chem. 2018, 66, 831−841
Article
Journal of Agricultural and Food Chemistry Table 1. Summary of Crops and Their Varieties Used for the Analysis of Water-Soluble Vitamins over Five Yearsa year 1
year 2
year 3
year 4
year 5
crop
2010/2011
2011/2012
2012/2013
2013/2014
2014/2015
potato
Cabaret Lady Balfour Maris Piper Mayan Gold Vales Sovereign Ben Fuego Maris Bead Pyramid Tattoo 4-Comp Mix Concerto Optic Wagon Westminster 4-Comp Mix Flaggon Retriever Saffron Sequel Alchemy Consort Istabraq Viscount Zebedee
Cabaret Lady Balfour Maris Piper Mayan Gold Vales Sovereign Ben Fuego Maris Bead Pyramid Tattoo 4-Comp Mix Concerto Optic Wagon Westminster 4-Comp Mix Cassata Retriever Saffron Sequel Alchemy Beluga Consort Istabraq Viscount
Cabaret Lady Balfour Maris Piper Mayan Gold Vales Sovereign Ben Fuego 1 Fuego 2 Pyramid Tattoo 4-Comp Mix Concerto Optic Wagon Westminster 4-Comp Mix Cassata Retriever Saffron Sequel Alchemy Beluga Consort Istabraq Viscount
Cabaret Lady Balfour Maris Piper Mayan Gold Vales Sovereign Babylon Boxer Fanfare Fuego Pyramid 4-Comp Mix Concerto Optic Wagon Westminster 4-Comp Mix Cassata Retriever Saffron Sequel Alchemy Beluga Consort Istabraq Viscount
Cabaret Lady Balfour Maris Piper 1 Maris Piper 2 Vales Sovereign Babylon Boxer Fanfare Fuego Pyramid 4-Comp Mix Concerto Optic Wagon Westminster 4-Comp Mix Cassata Retriever Saffron Sequel Alchemy Beluga Consort Istabraq Viscount
field beans
Spring barley
Winter barley
Winter wheat
a
Bold text denotes the industry standard. For Spring barley, the same four varieties, and a four-component mix (4-Comp Mix), comprising each of those varieties, were grown over 2011−2015. However, for the other crops, adjustments had to be made as to which varieties were grown in each year, depending on seed availability. For example, the potato variety Mayan Gold was not available for growth in 2015; however, no alternative variety was substituted. Rather, Maris Piper (the industry standard) was grown in two adjacent plot strips (Table 1). For Winter wheat in 2011, Beluga was not available, and Zebedee was grown in its place. Similarly, for Winter barley in 2011, Cassata was not available and Flaggon was grown in its place. This had implications for the 4-Comp Mix in 2011, as it was not comparable with the 4-Comp Mixes generated for Winter barley in 2012−2015. Field beans were more complex, where only two varieties, Fuego and Pyramid, were consistently grown over 2011− 2015. Two of the varieties, Ben and Tattoo, were only grown over 2011−2013. Maris Bead was only grown over 2011−2012. Finally, Babylon, Boxer, and Fanfare were available for 2014−2015. This meant that in total, over the five years, eight varieties of field beans were grown. The above has been considered and factored into the resulting statistical outputs, which will be explained in the Statistical Analysis section. Cereals and potatoes were harvested from 1 m × 1 m quadrats at five, fixed global positioning system (GPS) locations in each variety strip as indicated in Figure 1. With regards to the field beans, the GPS locations were not physically accessible in the standing crop without causing pod shatter and yield loss. Therefore, after each variety strip was harvested, five hand-sampled aliquots were collected (providing five technical, rather than spatial, replicates) for WSV analysis. Following harvest, the sample material was prepared as follows: the three cereal crops were dried down to industrial standards (10−15% moisture still present), threshed with a small combine harvester, and then graded (barley: sieve size of 2.5 mm; wheat: sieve size of 2.25 mm). The seeds were then milled with Retch ZM 200 ultracentrifugal mill (Tecator Udy, sieve size 0.5 mm). The milled powders were then
up in response to cropping system over time. The integrated management system is a composite treatment, including tram-line management in cereals and tied-ridging in potatoes to reduce soil, water, and nutrient loss;33 non-inversion tillage to improve physical structure and decrease nutrient losses;34 green waste compost addition and crop residue incorporation to build up soil carbon and improve physical structure;35 green cover (forage radish) over Winter before potato to reduce nitrogen (N) losses and increase phosphorus (P) uptake;36 clover under sowing of Spring barley crops for additional renewable N input to the rotation;37 lower doses of artificial N fertilizer (taking approximately 75% of the standard rate as a reasonable starting point for this site, based on expert agronomic advice, with further reductions planned as soil fertility improves) to reduce environmental footprint, leaching, and emissions;38 lower herbicide dose and alternative chemicals to encourage a diverse weed understorey, aiming at approximately 10% ground cover of noncompetitive dicotyledonous weeds,39 and threshold crop protection applications based on the Home-Grown Cereals Authority (HGCA), now the Agriculture and Horticulture Development Board (AHDB) dose response curves40 (Table 2). The crop rotation is potato followed by Winter wheat, Winter oilseed rape, Winter barley, field (Spring) beans, and Spring barley. These crops were selected as typical for the Tayside farming region and representative of the most common cropping systems in Scotland (and much of the U.K.). Winter crops are sown in late summer/autumn (August−October) and harvested the following summer (July−September). Spring crops are sown between March and May, and harvested August−September of the same year. Within each half-field, five different varieties of each crop were sown to assess variety-specific responses to the change in management systems. For each crop, one variety was selected which was an industry standard, providing a comparator to annual U.K. performance. The remaining varieties were selected for specific environmental traits such as disease resistance, resource use efficiency, and weed tolerance3 (Table 1). 833
DOI: 10.1021/acs.jafc.7b03509 J. Agric. Food Chem. 2018, 66, 831−841
Article
Journal of Agricultural and Food Chemistry individually packed in polyethylene bags (VWR, U.K.) and stored at −20 °C until required for analysis. After the potato harvest, the potato tubers were stored in the dark at ambient temperature for a minimum of 1 week to facilitate skin set, which is the common U.K. postharvest practice. For each replicate (GPS point), four to five average-sized potato tubers, with a combined fresh weight of approximately 800 g, were selected, washed by hand, and each tuber then transected into eight segments. Two diametrically opposed segments (taken to provide a representative sample of the whole tuber) were taken from each tuber, and the opposite eighths from the tubers per replicate combined, then flash frozen in liquid N and stored at −20 °C. The frozen material was then freeze-dried for five days, then the dried material milled using a Retch ZM 200 ultracentrifugal mill (Tecator Udy; 0.5 mm sieve). The milled potato powders were then individually packed in polyethylene bags (VWR, U.K.) and stored at −20 °C (in the dark) until required for analysis. Five “replicate” aliquots of field beans, each comprising 60 beans, were flash frozen in liquid N, stored at −20 °C, freeze-dried overnight, then milled as per the potatoes. The milled field bean powders were then individually packed, as for potato, and stored at −20 °C (in the dark) until required for analysis. Chemicals and Reagents. All chemicals used for the present study were of analytical grade (purity >98%). Analytical standards of thiamine hydrochloride, nicotinic acid, pyridoxine hydrochloride, pantothenic acid, and riboflavin were obtained from Scientific Laboratory Supplies Ltd. (Newhouse, U.K.). Individual stock solutions for all five vitamins were prepared in 50:50 (v/v) acetonitrile/water (4 mg mL−1). A stock solution of all five vitamins was prepared in 50:50 (v/v) acetonitrile/water with a concentration of 100 μg mL−1. The labeled internal standard pantothenic acid-13C3,15N hemicalcium salt was obtained from LGC Standards (Teddington, U.K.). Thiamine-4methyl-13C-thiazol-5-yl-13C hydrochloride and pyridoxal-methyl-d3 were purchased from Sigma-Aldrich (Dorset, U.K.). Internal standard solutions were prepared by dissolving 1 mg in 1 mL 50:50 (v/v) acetonitrile/water. Prior to batch extraction, a mix of all three internal standard solutions was prepared resulting in a final concentration of 13.6 mg mL−1. Sodium acetate trihydrate, glyoxylic acid monohydrate, L-glutathione reduced, ethylenediaminetetraacetic acid (EDTA), sodium hydroxide, iron(II) sulfate heptahydrate, formic acid, and glacial acetic acid were purchased from Fisher Scientific (U.K., Analytical grade). Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), metaphosphoric, sulfuric and ascorbic (AsA) acids were purchased from Sigma-Aldrich (Dorset, U.K.). HPLC-grade acetonitrile was purchased from VWR (West Sussex, U.K.). Ultrapure water (18.2 MΩ.cm) was obtained from an Elga Purelab-Option Q System (High Wycombe, U.K.). Quantification of Vitamin C. The vitamin C content of freezedried potato powders were quantified as follows: 100 mg of powder was weighed into 2 mL microfuge tubes and resuspended in 1 mL of 5% (w/v) metaphosphoric acid containing 5 mM TCEP, which acts as a reducing agent by converting dehydroascorbate (DHA) to AsA. Consequently, results are presented as total vitamin C (total AsA). The suspension was vortexed for 10 s and transferred onto a blood rotator for 30 min at 5 °C. Following this, the suspension was centrifuged at 5 °C for 10 min. The supernatant was then transferred into a clean 2 mL microfuge tube, and the remaining pellet was reextracted as before. Both extracted supernatants were combined, centrifuged to pellet any remaining debris, and the supernatant transferred into 0.3 mL transparent polypropylene-short threaded high-performance liquid chromatography (HPLC) microvials sealed with a 9 mm polypropylene screw cap (VWR, U.K.), and subjected to HPLC (ASI-100 autosampler, and Ultimate 3000 pump) coupled to a UV−visible detector (UVD340U, Dionex, ThermoFisher Scientific, U.K.). Autosampler and column temperature were maintained at 4 and 50 °C, respectively. Sample (20 μL) was injected onto an ICSep COREGAL-64H column (ChromTech, USA), with the dimensions of 7.8 × 300 mm and particle size of 10 μm and cross-linkage of 6.4. An isocratic run of 30 min was applied with a mobile phase containing 4 mM sulfuric acid in ultrapure water. AsA was detected by absorbance using a diode array detector and quantified at 245 nm. Quantification
was performed at 245 nm against external calibration of AsA in a range of 20−75 μg mL−1. Sample Extraction, Dilution and Preparation of the Standard Curve for Quantification of WSVs. Extraction procedures followed the protocol from Nurit et al.31 and were as follows: for extraction, 622 mg (±2 mg) was weighed into 50 mL tubes (Sarstedt, Germany). Following this, 100 μL of internal standard solution (13.6 μg mL−1) was added in addition to 4.75 mL of sodium acetate (pH 4.5, concentration of 0.5 mM, pH adjusted with glacial acetic acid), 1.25 mL of 0.5 M glyoxylic acid solution, 0.25 mL of 1% (v/v) Lglutathione reduced solution, 0.25 mL of 1% ethylene-diaminetetraacetic acid solution (adjusted with NaOH for solubility), and 0.2 mL of 2% iron(II) sulfate heptahydrate to the powder. The mixture was strongly mixed for 30 s using a vortex and then incubated in the dark at 37 °C for 16 h on a shaker (1500 rpm). Following this, the cooled sample was vortexed and centrifuged at 12 000g for 10 min at 3 °C. The supernatant was filtered through a 0.22 μm filter vial (polytetrafluoroethylene; PTFE) with a preslit cap (Thomson, BioProcess Engineering Services Ltd., Kent, U.K.). Reference materials were prepared to account for extraction and instrument stability (see Supporting Information (SI); Reference Generation and Quality Control, and Tables S2−S6). Chemical Analysis. Chemical analysis of the potato, field bean and cereal powders were performed on an Agilent 1260 HPLC system consisting of a quaternary pump, a Diode Array Detector (DAD), a Temperature Control Device, and a solvent Thermostat module (Agilent Infinity 1290) coupled to an Agilent 6460A Triple Quadrupole Mass Spectrometer (Agilent Technologies, Santa Clara, CA, USA). Sample extract (5 μL) was injected onto a 100 × 3 mm (2.5 μm) Synergy Hydro-RP C18 column with polar end-capping, fitted with an AQ C18 4 × 2 mm security guard TM cartridge (Phenomenex, Cheshire, U.K.). Samples were eluted at a flow rate of 0.5 mL min−1 using a gradient consisting of two mobile phases: A = 0.1% (v/v) formic acid in deionized water and B = 100% acetonitrile. The elution gradient was as follows: A/B 98/2 (v/v) hold for 2 min; ramped up from 2 to 60% B in 3 min and hold for 1 min, and further ramped up from 60% to 90% in 0.1 min and hold for 1.9 min. Within 0.1 min the gradient was returned to the initial composition and held for 5 min until the next injection. One analytical run lasted 13.1 min (SI, Figure S2). Mass detection was carried out in positive-ion mode for all vitamins apart from the isotopically labeled pantothenic acid (pantothenic acid 13 C3, 15N) using a jet stream electrospray ionization (ESI) interface coupled to the triple quadrupole system. For ESI, the gas temperature, gas flow, nebulizer pressure, sheath gas temperature, sheath gas flow, capillary cap voltage, and nozzle voltage were set to 350 °C, 11 L min−1, 50 psi, 300 °C, 11 L min−1, 4 kV (3 kV for negative-ion mode) and 500 V, respectively. Collision energies for transition states of the five standard compounds including thiamine, nicotinic acid, pyridoxine, pantothenic acid, and riboflavin, as well as the isotopically labeled internal standards thiamine-4-methyl-13C-thiazol-5yl13C, pyridoxal methyl-d3, pantothenic acid 13C3, 15N, were optimized for optimal fragmentor voltage and collision energies (SI, Table S1). Hereby, the most sensitive transitions (i.e., transitions with the highest intensity of the product ions) were chosen to build the final multiple reactions monitoring (MRM) method. As shown in Table S1, 16 transitions were part of the MRM mode, each with a dwell time of 20 ms (ms) and a delay time of 3.5 ms, leading to a total cycle time of 376 ms, and thus 2.7 Hz. Peaks of the five B vitamins and three labeled B vitamins were integrated with Agilent MassHunter Quantitative Software (Agilent, U.S.A.). Quantification of the WSVs. As listed in Table S1, seven protonated molecular ions [M + H]+ and one deprotonated molecular ion [M − H]− (pantothenic acid-13C3,15N) were chosen as precursor ions in the MS/MS (tandem Mass Spectrometry) experiment. For confirmation, qualifier ions were also included in the method. The most abundant product ions were used for quantification, measured in MRM mode. A 10-point calibration curve for each B vitamin in the potato, field bean, and cereal samples were calculated ranging from 2 ng mL−1 up to 1 μg mL−1. As in Nurit et al.,31 calibration curve 834
DOI: 10.1021/acs.jafc.7b03509 J. Agric. Food Chem. 2018, 66, 831−841
Article
Journal of Agricultural and Food Chemistry standards were prepared by adding 100 μL of mix unlabeled external standards (10 ng mL−1−5 μL min−1) into microfuge tubes containing 100 μL of thiamine-4-methyl-13C-thiazol-5-yl-13C hydrochloride (1 μL min−1), 100 μL of pantothenic acid-13C3,15N hemicalcium salt (1 μL min−1), 100 μL of pyridoxal-methyl-d3 (1 μL min−1), and 100 μL of acetonitrile/water (50:50; v/v). Due to matrix effects, the response ratio of each vitamin was calculated against a labeled internal standard of the same or as similar chemistry as possible: thiamine with thiamine-4-methyl-13C-thiazol-5-yl-13C hydrochloride; riboflavin and pantothenic acid with pantothenic acid-13C3,15N hemicalcium salt, and pyridoxine; and nicotinic acid with pyridoxal-methyl-d3. Statistical Analysis. A separate statistical analysis was performed on the vitamin measurements for each crop using a linear mixed model approach. The two main effects of Variety and Input (management system) were fitted as fixed effects along with an interaction term. In addition, the effect of Year was analyzed as a fixed, rather than random, effect because variance component estimates with so few levels can be unreliable. The Years here are considered as experimental replicates and account for variability in field and environmental conditions. Three of the terms in the random model account for the physical layout of the design. The nested block terms of Replicate (Rep; samples taken at the five GPS points, except for field beans, as described below) within Variety strip within half-field were fitted as random effects. In addition, interaction terms for Year × Variety, Year × Rep, and Year × Input × Rep were also included as random effects. The Year × Variety term has been found to be a particularly important component of variance, and it is necessary for a good model fit. The terms for Year × Rep and Year × Input × Rep capture further sources of variance within the field. They also account for any variation due to the order in which samples were processed in the laboratory stage, as during an analysis run Rep blocks are processed in a sequence with sample order randomized within these (see SI). Between years there were changes to the selection of varieties grown for field beans, Winter wheat, Winter oilseed rape, and Winter barley (Table 1). Varieties which were present in only one year (Winter wheat−variety Zebedee and Winter barley−variety Flaggon) were excluded from the analysis. Data from field beans had a slightly different structure in that the five samples from each strip were technical replicates, rather than spatial. However, the same model was used since Year × Rep and Year × Input × Rep terms were still required to capture variability in laboratory processing of the technical Rep blocks. The complexity of the experimental design meant that there were many potential random effects which were difficult to estimate in combination. The model we selected here ensured that the fixed effects were estimated against the appropriate level of random variation with degrees of freedom estimated from the design. Measured values were logarithmically transformed to base 10 (log10) before the analyses to account for variance heterogeneity in the residuals. All analyses were performed using Restricted Maximum Likelihood (REML) procedures in GenStat for Windows 17th edition (VSN International Ltd., Hemel Hempstead, U.K.).
riboflavin and thiamine concentrations when compared with the other four varieties. Broadly speaking, Vales Sovereign had the highest concentrations of pantothenic acid, pyridoxine, and vitamin C. Figure 2 shows that Input did not significantly (Table 3) affect the concentrations (Table S8) of any of the six vitamins quantified over the five years studied.
Figure 2. Mean value concentrations (five replicates × five years) of six vitamins, including riboflavin, pyridoxine, pantothenic acid, nicotinic acid, thiamine, and vitamin C for the potato varieties Cabaret, Lady Balfour, Maris Piper, Mayan Gold, and Vales Sovereign, grown under Conventional and Integrated management systems. Where values are shown on the log10 transformed (black values) and natural scale (blue values: expressed as μg g−1 dry weight for the first five vitamin, and mg g-1 dry weight for vitamin C). The average standard error of difference (s.e.d.) is for the Variety × Input term. Con = Conventional; Int = Integrated; log10 = logarithmically transformed to base 10.
■
RESULTS Five WSVs (nicotinic acid, pyridoxine, thiamine, riboflavin, and pantothenic acid; Figure S1) were quantified in five varieties of five different crops: potato, Spring barley, Winter barley, Winter wheat, and field beans. For potato, vitamin C content was also quantified due to its intake being of high importance in western diets.41,42 In potato, highly significant differences (p < 0.001; thiamine p < 0.01; Table 3) in all six vitamin concentrations were seen in all five varieties (Table S8). Most obvious from Table S8 were the levels of nicotinic acid, which varied from ∼1−46 μg g−1 DW, between the varieties (p < 0.001). Cabaret had the lowest nicotinic acid levels, whereas the levels in Vales Sovereign were 7−40-fold higher when compared with the other four varieties. Cabaret also had the lowest levels of pantothenic acid,
Table 3 and Table S9 show significant differences were observed across the field bean varieties in the concentrations of nicotinic acid (p < 0.05), pyridoxine (p < 0.01), and riboflavin (p < 0.001). Figure 3 and Table S9 show that the variety Ben had the lowest levels of nicotinic acid compared to the other seven varieties, including Maris Bead, Boxer, Fanfare, Fuego, Babylon, Pyramid, and Tattoo. The lowest concentration of 835
DOI: 10.1021/acs.jafc.7b03509 J. Agric. Food Chem. 2018, 66, 831−841
Article
Journal of Agricultural and Food Chemistry
Figure 4. Mean value concentrations (five replicates × five years) of five vitamins, including riboflavin, pyridoxine, pantothenic acid, nicotinic acid, and thiamine of the Spring barley varieties, including a four-seed mixture (4-Comp Mix), Concerto, Wagon, Westminster, and Optic, grown under Conventional and Integrated management systems. Where values are shown on the log10 transformed (black values) and natural scale (blue values: expressed as μg g−1 dry weight). The average standard error of difference (s.e.d.) is for the Variety × Input term. Con = Conventional; Int = Integrated; log10 = logarithmically transformed to base 10.
Figure 3. Mean value concentrations (five technical replicates × five years,) of the five vitamins riboflavin, pyridoxine, pantothenic acid, nicotinic acid, and thiamine of the field bean varieties Ben, Fuego, Maris Bead, Pyramid, Tattoo, Babylon, Boxer, and Fanfare, grown under Conventional and Integrated management systems. Where values are shown on the log10 transformed (black values) and natural scale (blue values: expressed as μg g−1 dry weight). The average standard error of difference (s.e.d.) is for the Variety × Input term. Con = Conventional; Int = Integrated; log10 = logarithmically transformed to base 10.
thiamine (p < 0.001). Input only significantly affected nicotinic acid and thiamine (p < 0.05), which were lower in varieties grown under the integrated management system (Figure 6; Table S12).
pyridoxine was found in Maris Bead, whereas this variety had the highest concentrations of riboflavin. With the exception of thiamine (Figure 3), whose concentration was significantly lower (p < 0.01) in varieties grown under the integrated management system (Table 3), Input did not significantly affect any of the other analyzed vitamins (Table S9). For Spring barley, with the exception of thiamine (Table S10), highly significant variety differences (p < 0.001; Table 3) were observed in the concentrations of four of the WSVs analyzed. As observed for the field beans, Input only affected the concentration of thiamine (Figure 4), which again was lower (p < 0.05; Table 3) in varieties grown under the integrated management system (Table S10). Table 3 shows that for Winter barley all five WSV concentrations had highly significant (p < 0.001) concentration differences between varieties. However, Input did not significantly affect the concentration of any of these (Figure 5; Table S11). Significant varietal differences were observed for Winter wheat (Table 3) for nicotinic acid (p < 0.001), pantothenic acid (p < 0.001), pyridoxine (p < 0.05), riboflavin (p < 0.01), and
■
DISCUSSION This study provides quantitative data on WSVs in potato, field beans, and cereals grown under both conventional and integrated management systems, in a rotational system for 5 consecutive years (2011−2015). In terms of experimental design, crop management system, study duration, and type of chemical analysis, this represents the most extensive comparison of the impacts of conventional and integrated management systems in the U.K. (and beyond), on an important aspect of nutritional quality (i.e., WSVs) in a wide range of crops. The adoption and adaptation of the method of Nurit et al.31 for the quantitation of WSVs further extended this ability to a range of different crop matrices of economic importance and represents a rapid and robust approach (as confirmed by the quality controls; see SI) for the simultaneous quantification of five B vitamins. While Variety and Year differences in individual vitamin concentrations within each crop were often found to be 836
DOI: 10.1021/acs.jafc.7b03509 J. Agric. Food Chem. 2018, 66, 831−841
Article
Journal of Agricultural and Food Chemistry
Figure 5. Mean value concentrations (five replicates x five years) of five vitamins, including riboflavin, pyridoxine, pantothenic acid, nicotinic acid, and thiamine of the Winter barley varieties, including a four-seed mixture (4-Comp Mix), Retriever, Saffron, Sequel, and Cassata, grown under Conventional and Integrated management systems. Where values are shown on the log10 transformed (black values) and natural scale (blue values: expressed as μg g−1 dry weight). The average standard error of difference (s.e.d.) is for the Variety × Input term. Con = Conventional; Int = Integrated; log10 = logarithmically transformed to base 10.
Figure 6. Mean value concentrations (five replicates × five years) of five vitamins, including riboflavin, pyridoxine, pantothenic acid, nicotinic acid and thiamine of the Winter wheat varieties Alchemy, Beluga, Consort, Istabraq, and Viscount, grown under Conventional and Integrated management systems. Where values are shown on the log10 transformed (black values) and natural scale (blue values: expressed as μg g−1 dry weight). The average standard error of difference (s.e.d.) is for the Variety × Input term. Con = Conventional; Int = Integrated; log10 = logarithmically transformed to base 10.
significant, Input had few significant impacts on the vitamin concentrations. Potato, as the third most important food crop worldwide43 is of direct importance for human nutrition in Great Britain. Within Great Britain, 23% of the total potato planted area can be found in Scotland, where 47% of this area is used for seed potatoes.44 To highlight the importance of the Scottish potato varieties, around 65 000 t (tonnes) of seed potatoes and 10 000 t of ware potatoes are produced each year in Scotland and exported outside the EU (including the Canary Islands).45 Out of the six quantified vitamins in potato, vitamin C was the predominant vitamin. There were no significant differences in vitamin C concentration in potatoes grown under either agricultural management system, although there is a trend toward marginally higher levels grown under our integrated management system. Previous studies on potato showed higher vitamin C concentrations in organically grown products, and an inverse relationship between fertilizer (N) Input and vitamin C levels in plants has been reported in previous studies.46,47 Because the integrated management system refers to the reduced nonrenewable-derived fertilizer Input (Table 2) as opposed to an organic system (where inorganic fertilization is
completely replaced by manures), it might have influenced the results here; hence, no statistically significant differences were observed. However, Hajslova et al.5 highlighted findings from an organic versus conventional agricultural management comparison, where pooled results from four consecutive harvests of potato plants showed that year to year variation and variety differences were significantly greater factors determining the nutritional quality (micronutrients, metals, secondary metabolites, enzymic browning and organoleptic properties) of potatoes than the agricultural management system. The integrated management system described in our study is not directly comparable with the organic management system described by Hajslova et al.;5 however, despite our study measuring only vitamin C and other WSVs, we describe similar response patterns, with regards to Variety, Input, and Year differences. In this report, significantly lower thiamine concentrations were observed for field beans, Spring barley, and Winter wheat varieties grown under the integrated management system. Thiamine is a nitrogen-rich vitamin (4 N atoms; Figure S1), and its activated form, Thiamine diphosphate, is a key 837
DOI: 10.1021/acs.jafc.7b03509 J. Agric. Food Chem. 2018, 66, 831−841
Article
Journal of Agricultural and Food Chemistry
Table 2. Summary of the Main Differences in Agronomic Practices between the Integrated and Conventional Management Systems at the Platforma crop input nitrogen fertilizer (kg/ ha) tillage compost (t/ha) clover cover crop
field beans
Spring barley
Winter wheat
Winter barley
Int
potato Con
Int
Con
Int
Con
Int
Con
Int
Con
147
196
0
0
83
112
145
198
130
170
non-inversion 35 0 oil-radish
plough 0 0 0
non-inversion 35 0 0
plough 0 0 0
non-inversion 35 under-sown 0
plough 0 0 0
non-inversion 35 0 0
plough 0 0 0
non-inversion 35 0 0
plough 0 0 0
Where values are averages over the first five years of the six-year crop rotation covering seasons 2011−2015. Con = Conventional; ha = hectare, Int = Integrated; kg = kilograms; t = tonnes. Herbicides were applied to the integrated management field half, at half of the rate of the conventional system, unless specific weed problems required targeted treatment with a different product. Full details of agrochemical and other inputs to both cropping systems can be found, by request, at http://csc.hutton.ac.uk/. a
Table 3. p-Values below Are Derived from the log10 Transformed Dataset for Years 2011−2015, for the Terms Variety, Input, Year, and the Interaction between Variety × Inputa crop potato
field beans
Spring barley
Winter barley
Winter wheat
a
term Variety Input Year Variety Variety Input Year Variety Variety Input Year Variety Variety Input Year Variety Variety Input Year Variety
× Input
× Input
× Input
× Input
× Input
nicotinic acid
pantothenic acid
pyridoxine
riboflavin
thiamine
vitamin C
