Variation in Brooli Cultivar Phytohemial Content under Organi and Conventional Management Systems: Impliations in Breeding for Nutrition Renaud ENC, Lammerts van Bueren ET, Myers JR, Paulo MJ, van Eeuwijk FA, et al. (2014) Variation in Brooli Cultivar Phytohemial Content under Organi and Conventional Management Systems: Impliations in Breeding for Nutrition. PLoS ONE 9(7): e95683. doi:10.1371/journal.pone.0095683 10.1371/journal.pone.0095683 Puli Lirary of Siene Version of Reord http://dss.lirary.oregonstate.edu/sa-termsofuse
Variation in Brooli Cultivar Phytohemial Content under Organi and Conventional Management Systems: Impliations in Breeding for Nutrition Eria N. C. Renaud 1 *, Edith T. Lammerts van Bueren 1, James R. Myers 2, Maria João Paulo 3, Fred A. van Eeuwijk 3, Ning Zhu 4, John A. Juvik 4 1 Wageningen UR Plant Breeding, Plant Sienes Group, Wageningen University, Wageningen, The Netherlands, 2 Department of Hortiulture, Oregon State University, Corvallis, Oregon, United States of Ameria, 3 Biometris, Plant Sienes Group, Wageningen University, Wageningen, The Netherlands, 4 Department of Crop Sienes, University of Illinois, Urana, Illinois, United States of Ameria Astrat Organi agriulture requires ultivars that an adapt to organi rop management systems without the use of syntheti pestiides as well as genotypes with improved nutritional value. The aim of this study enompassing 16 experiments was to ompare 23 rooli ultivars for the ontent of phytohemials assoiated with health promotion grown under organi and onventional management in spring and fall plantings in two rooli growing regions in the US (Oregon and Maine). The phytohemials quantified inluded: gluosinolates (gluoraphanin, gluorassiin, neogluorassin), toopherols (d-, -, a-toopherol) and arotenoids (lutein, zeaxanthin, -arotene). For gluoraphanin (17.5%) and lutein (13%), genotype was the major soure of total variation; for gluorassiin, region (36%) and the interation of loation and season (27.5%); and for neogluorassiin, oth genotype (36.8%) and its interations (34.4%) with season were important. For d- and - toopherols, season played the largest role in the total variation followed y loation and genotype; for total arotenoids, genotype (8.41 13.03%) was the largest soure of variation and its interations with loation and season. Overall, phytohemials were not signifiantly influened y management system. We oserved that the ultivars with the highest onentrations of gluoraphanin had the lowest for gluorassiin and neogluorassiin. The genotypes with high onentrations of gluorassiin and neogluorassiin were the same ultivars and were early maturing F 1 hyrids. Cultivars highest in toopherols and arotenoids were open pollinated or early maturing F 1 hyrids. We identified distint loations and seasons where phytohemial performane was higher for eah ompound. Correlations among hortiulture traits and phytohemials demonstrated that gluoraphanin was negatively orrelated with the arotenoids and the arotenoids were orrelated with one another. Little or no assoiation etween phytohemial onentration and date of ultivar release was oserved, suggesting that modern reeding has not negatively influened the level of tested ompounds. We found no signifiant differenes among ultivars from different seed ompanies. Citation: Renaud ENC, Lammerts van Bueren ET, Myers JR, Paulo MJ, van Eeuwijk FA, et al. (2014) Variation in Brooli Cultivar Phytohemial Content under Organi and Conventional Management Systems: Impliations in Breeding for Nutrition. PLoS ONE 9(7): e95683. doi:10.1371/journal.pone.0095683 Editor: Hany A. El-Shemy, Cairo University, Egypt Reeived Otoer 14, 2013; Aepted Marh 31, 2014; Pulished July 16, 2014 Copyright: ß 2014 Renaud et al. This is an open-aess artile distriuted under the terms of the Creative Commons Attriution Liense, whih permits unrestrited use, distriution, and reprodution in any medium, provided the original author and soure are redited. Funding: Funding was provided y Seeds of Change (www.seedsofhange.om) and monetary funding from Wageningen University, University of Illinois and Oregon State University. The funder had no role in the study design, data olletion and analysis, deision to pulish, or preparation of the manusript. Competing Interests: The priniple researher, Eria Renaud was employed y Seeds of Change and later Vitalis Organi Seeds, the organi division of Enza Zaden, a vegetale seed ompany. This does not alter the authors adherene to all PLOS ONE poliies on sharing data and materials. * Email: E.Renaud@enzazaden.om Introdution Organi food onsumption is in part driven y onsumer pereption that organi foods are more nutritious and simultaneously less potentially harmful to human health [1 2]. Studies, suh as Smith-Sprangler et al. [3], have onluded that there is little evidene for differenes in health enefits etween organi and onventional produts, ut other studies have indiated that organi vegetales and fruits ontain higher onentrations of ertain plant phytohemials assoiated with health promotion than those produed onventionally [4 8]. A numer of these ompounds are produed y plants in response to environmental stress or pathogen infetion, providing a potential explanation of why onentrations of these ompounds might e higher in plants grown in organi systems without appliation of pestiides [9]. In addition, higher phytohemial levels may e due to the effets that different fertilization praties have on plant metaolism. Syntheti fertilizers used in onventional agriulture are more readily availale to plants than organi fertilizers [10]. Nutrients derived from organi fertilizers need to e mineralized, and the availaility of these nutrients depends on soil moisture, temperature and level of ativity of soil organisms [11]. Conventional systems seek to maximize yields, resulting in a relative derease of plant phytohemials and seondary metaolites [12 15]. Correspondingly, ompounds suh as phenolis, flavonoids, and indolyl gluosinolates may e indued y ioti or aioti stress [16 17]. Brooli is an aundant soure of nutrients, inluding provitamin A (-arotene), vitamin C (asorate), and vitamin E (toopherol) [18]. It is also a soure of phytohemials assoiated with health enefits and these inlude gluosinolates, arotenoids, toopherols, and flavonoids [19 21]. Verhoeven et al. [22], Kek and Finley [23] and Here and Bühler [24], reported that diets PLOS ONE www.plosone.org 1 July 2014 Volume 9 Issue 7 e95683
rih in rooli redue aner inidene in humans. Strong assoiations etween onsumption level and disease risk redution exists for gluosinolates (anti-aner), toopherols (ardiovasular), and the arotenoids (eye-health) [25]. Sulfur ontaining gluosinolates are found in the tissues of many speies of the Brassiaeae family. When gluosinolates are onsumed, they are hydrolyzed into isothioyanates (ITC) and other produts that up-regulate genes assoiated with arinogen detoxifiation and elimination. Aliphati gluoraphanin (up to 50% of total gluosinolates) and the indolyli gluosinolates, gluorassiin and neogluorassiin are aundant in rooli florets [20,19,26]. Gluoraphanin is hydrolyzed either y the endogenous plant enzyme myrosinase [27 28] or y gut miroes to produe sulforaphane, an ITC. The indole gluosinolates are tryptophan-derived in a similar ut alternate iosyntheti pathway [29]. The health promoting effets of the indolyl gluosinolates are attriuted to indole-3-arinol, a hydrolysis produt of gluorassiin, N-methoxyindole-3-arinol and neoasorigen, hydrolysis produts from neogluorassiin, and the ataoli produts derived from alkyl gluosinolates. Clinial studies have shown that the gluosinolate hydrolysis produts redue the inidene of ertain forms of aner (e.g., prostate, intestinal, liver, lung, reast, ladder) [30 35]. The lipophili phytonutrients found in rooli inlude the arotenoids lutein, zeaxanthin, -arotene, and toopherols (forms of vitamin E) [36 37]. In addition to their role as vitamins, these ompounds are powerful antioxidants [38 39]. Consumption of vegetales high in toopherols and arotenoids has dereased the inidene of ertain forms of aner [40]. Lutein and zeaxanthin protet against development of atarats and age-related maular degeneration [41]. Toopherols have also een assoiated with redued risk of ardiovasular disease y preventing oxidative modifiation of low-density lipoproteins in lood vessels [42]. The geneti potential for high nutrient ontent has long een a onern of the organi industry in order to meet the expetations of organi onsumers. This has often een manifested y questioning whether modern elite ultivars may have lower levels of nutritional ontent than older open pollinated ultivars. Indiret evidene supporting this argument omes from Davis et al. [43], who ompared USDA nutrient ontent data for 43 garden rops released etween 1950 and 1999. Statistially signifiant dereases were noted for six nutrients (protein, alium, potassium, iron, rioflavin, and asori aid), with delines ranging from 6% for protein to 38% for rioflavin. Crop varieties in 1950 had een red to e adapted to speifi regions and a relatively low input agriulture system, ut ontemporary ultivars are seleted for yield, disease resistane, road adaptation to high input agriulture systems, and for inreased shipaility and shelf life. Traka et al. [44] reommend reeding with greater geneti diversity when the goal is enhaned phytohemial ontent y exploiting wild rop relatives. The genotype is important in determining the level of nutrients in a rop ultivar [45 47]. What is unlear, however, is whether the nutritional ontent of a ultivar is assoiated with ertain genotypi ategorization, e.g. old versus modern, open pollinated versus F 1 hyrid ultivars. In addition, there is no lear differentiation as to what extent nutritional ontent in a rop is determined y genotypi or y field management fators or y the interation of oth. Some studies omparing performane of genotypes in organi and onventional prodution systems have shown that for ertain agronomi traits, ultivars perform differently etween the two prodution systems (e.g. for winter wheat: Murphy et al. [48], Baresel et al. [49]; for lentils: Vlahostergios et al. [50]; for maize: Goldstein et al.[51]), while others have shown no differenes in ranking performane (for maize: Lorenzana and Bernardo [52]; for onions: Osman et al. [53]; for ereals: Przystalski et al.[54]). The results of these studies have profound impliations for organi ultivar seletion and reeding strategies and raise questions as to the need for ultivars to e red with road adaptaility or speifi adaptation for the requirements of regional organi prodution and for designing reeding programs that optimize phytohemials in an adapted management system. Previous studies omparing organially versus onventionally grown rooli for nutritional quality have een market asket (off-the-shelf) studies [55 56]. Harker [57] explained that the limitation of market asket studies is that they either have purhased the produts from the store shelf and annot relate differenes to speifi growing onditions or that the numer of ultivars is too small to generalize the results. While other studies have ompared ultivars from one prodution season time period to another, knowledge of the atual ultivar and prodution system (soil quality, temperature, rainfall) was not availale [58,43]. The onentrations and form of health-promoting nutrients in Brassia vegetales have een reported to vary signifiantly due to (1) genotype (ultivar and genotypi lass) [59,20,26,60,21,37,61,44], (2) environmental onditions suh as season [62 67], light [19], max/min temperature, irrigation [68 69], (3) genotype y environment interations [19,70 71]; (4) management system inluding soil fertility [72 73], organi versus onventional [13,74 75], days to harvest [63 64], and (5) post-harvest management [76 77]. Identifying speifi growing onditions and genotypes that produe ultivars with varying phytohemial ontent and putative disease-prevention ativity ould offer valueadded ommerial opportunities to the seed and food industry. In addition to researh onduted on how rooli genotypes, management system and environment interat for hortiultural traits [78], we address in this paper the question of how do genotypes, management system and environment interat to determine the nutritional ontriutions of rooli to the human diet. We studied the relative importane and interation among genotypes (ultivars, genotypi lasses) and environment {management system [M: organi (O) or onventional (C)], season (S, a omination of year and season within year, i.e., fall 2006, spring 2007, fall 2007, spring 2008), loation (E)} in a set of 23 rooli ultivars for floret gluosinolate, toopherol and arotenoid onentrations grown under organi and onventional prodution systems in two ontrasting rooli prodution regions of the US: Oregon and Maine. Speifially we addressed the following questions: (1) what is the impat of organi management system ompared to the environmental fators inluding limati region, season and their interations [Genotype (G) x Environment (E) x Management System (M)]? (2) is there a signifiant differene in phytohemial ontent etween different genotypes and genotypi lasses (old and modern ultivars; open pollinated and F 1 hyrid ultivars; early and late maturing ultivars; and etween different ommerial seed soures)? (3) what is the est seletion environment for a rooli reeding program for enhaned phytohemial ontent? Materials & Methods Plant Material and Field Trial Loations Twenty-three rooli ultivars inluding open pollinated (OP) ultivars, inred lines, and F 1 hyrids were inluded in field trials (Tale 1). Cultivars were grown in a randomized omplete lok design with three repliates in Maine (ME)-Monmouth (Latitude 44.2386 o N, Longitude 70.0356 o W); and Oregon (OR)-Corvallis (Latitude 44.5647 o N, Longitude123.2608 o W)] with eah loation PLOS ONE www.plosone.org 2 July 2014 Volume 9 Issue 7 e95683
inluding organially (O) and onventionally (C) managed treatments. Plots ontained 36 plants, planted in three rows of 12 plants at 46 m equidistant spaing within and etween rows. The 2006 trials had only 18 of the 23 entries, and the Oregon 2006 trial had only two repliates at the organi loation. Field trials were onduted for three onseutive years with one prodution yle in Fall 2006, two prodution yles in Spring and Fall 2007 and one prodution yle in Spring 2008. The primary management differenes etween the organi and onventional field trial sites are outlined in Tale S1 in File S1, whih desries the prodution system, soils, fertility appliations, the applied supplemental irrigation, and weather onditions for the area of study. Further details of the field design are reported in Renaud et al. [78]. Field Data Colletion As plots approahed maturity they were evaluated three times a week for field quality and rooli heads that had reahed ommerial market maturity (approximately 10 to 12 m in diameter for most of the ultivars while retaining firmness). Field quality traits evaluated on a 1 to 9 ordinal sale inluded head olor, ead size, and ead uniformity. Average head weight was determined y taking the mean of the five individual heads per plot. Head diameter averaged for five heads at harvest maturity from eah plot. Maturity was ased on days to harvest from transplanting date. Detailed proedures and hortiulture trait performane data are reported in Renaud et al. [78]. Brooli Floret Samples and gluosinolate, toopherol, and arotenoid analysis In order to analyse nutritional ompounds of the rooli heads, the following proedure was followed: As plots approahed maturity, five rooli head tissue samples were harvested fresh from eah suplot at eah trial loation and were omposited into a single sample per repliation. The samples were frozen at 220uC and shipped in a frozen state to the University of Illinois, Urana- Champaign where they were freeze-dried and assessed for nutritional phytohemials. Eah sample was analyzed for the gluosinolates (gluoraphanin, gluorassiin and neogluorassiin), arotenoids (-arotene, lutein, and zeaxanthin), and toopherols (d-, -, a- toopherol) y high-performane liquid hromatography (HPLC) analysis using analytial protools desried in Brown et al. [19] for gluosinolates, and Irahim and Juvik [37] for toopherols and arotenoids. Gluosinolates in lyophilized floret tissue samples were extrated and analysed y HPLC using a reverse phase C18 olumn. Three hundred mg samples of rooli floret tissue were weighed out for extration and the HPLC quantifiation of the toopherols and arotenoids. Statistial Analysis Various linear mixed models were used for the analysis of trait variation. We followed the same methodology as desried in Renaud et al. [78], whih was omparale to the approah followed y Lorenzana and Bernardo [52]. For fitting the linear Tale 1. Overview of ommerially availale rooli ultivars, showing origin, main harateristis, inluded in paired organi - onventional field trials 2006 2008. Cultivar Areviation Origin Cultivar Type a Date of Market Entry Maturity Classifiation Aradia ARC Sakata F 1 1985 L B1 10 B11 Rogers F 1 1988 M Batavia BAT Bejo F 1 2001 M Beaumont BEA Bejo F 1 2003 L Belstar BEL Bejo F 1 1997 L Diplomat DIP Sakata F 1 2004 L Early Green EGR Seeds of Change OP 1985 E Everest EVE Rogers F 1 1988 E Fiesta FIE Bejo F 1 1992 L Green Goliath GRG Burpee F 1 1981 M Green Magi GRM Sakata F 1 2003 M Gypsy GYP Sakata F 1 2004 M Imperial IMP Sakata F 1 2005 L Marathon MAR Sakata F 1 1985 L Maximo MAX Sakata F 1 2004 L Nutriud NUT Seeds of Change OP 1990 E OSU OP OSU Jim Myers, OSU OP 2005 E Pakman PAC Petoseed F 1 1983 E Patriot PAT Sakata F 1 1991 M Patron PAN Sakata F 1 2000 M Premium Crop PRC Takii F 1 1975 E USVL 048 U48 Mark Farnham, USVL Inred not released L USVL 093 U93 Mark Farnham, USVL Inred not released M a Cultivar Type: F 1 : hyrid; OP: Open Pollinated; Inred. Maturity Classifiation: E: Early; M: Mid; L: Late. doi:10.1371/journal.pone.0095683.t001 PLOS ONE www.plosone.org 3 July 2014 Volume 9 Issue 7 e95683
mixed models, GenStat 15 (VSNi, 2012) was used. The models followed the set-up: y~ezr(e)zgzg Eze: Here y is the phytohemial response. Term E represents the environment in a very general sense, it inludes all main effets and interations of Season (S), Loation (L) and Management (M). For analyses per loation, the terms involving L were dropped. Similarly, for analyses regarding a speifi management regime, the terms involving M were dropped. Term R(E) is the effet of repliate within environment, and there were two or three repliates in individual trials. G and G6E are genotype and genotype y environment interation effets, respetively. Finally e is a residual. Variane omponents were reported as oeffiients of variation, i.e., CV = 100!V/x with V the variane orresponding to speifi effets and x the trait mean. Repeataility was alulated from the variane omponents in its most general form as H 2 =V G /(V G V GL /nl ++V GS /ns + V GM /nm + V GLS /(nl.ns) + V GLM /(nl.nm) + V GSM /(ns.nm) + V GLSM /(nl.ns.nm) + V e /(nl.ns.nm.nr)), where the variane omponents orrespond to the terms in the mixed model aove. The terms nl, ns, nm and nr stand for the numer of loations (2: Maine and Oregon), numer of seasons (4: Fall 2006, Spring 2007, Fall 2007, Spring 2008), management (2; organi and onventional), and repliates (2 or 3). Genotypi means were alulated y taking genotypi main effets fixed instead of random in the mixed models aove. Pairwise omparisons etween genotypi means were performed using GenStat proedure VMCOMPARISON. Correlations on the asis of genotypi means were referred to as geneti orrelations. Genotypi stailities under organi and onventional onditions were alulated as the variane for individual genotypes aross all trials in the system. To assess the feasiility of seletion for organi onditions (the target environment) under onventional onditions, we alulated the ratio of orrelated response (for organi onditions using onventional onditions), CR, to diret response (for organi onditions in organi onditions), DR, as the produt of the geneti orrelation etween organi and onventional systems (r G ) and the ratio of the roots of onventional and organi repeatailities (H C and H O respetively): CR/DR = r G H G /H O. A ratio smaller than 1 indiates that seletion is etter done diretly under organi onditions when the aim is indeed to improve the performane in organi onditions. Results Comparison of phytohemials means over the environments Gluosinolates. Aross all trials, gluoraphanin levels were omparale etween loations and seasons ut were more variale at the individual loation and season trial analysis level (Tale 2). Gluoraphanin, gluorassiin and neogluorassiin levels were omparale etween organi and onventional treatments. Comparisons of organi versus onventional y loation and season for the gluosinolate phytohemials are presented in Figure S1. Comparale levels of gluosinolates were oserved in the organi - onventional omparisons within loations and seasons. Toopherols. Aross trials ompared regionally, Oregon had higher levels of all three toopherols ompared to Maine (Tale 2, Figure S2). The toopherols d- and - were higher in Fall ompared to Spring, ut not so for a-toopherol (Figure S2). Organi and onventional levels for all toopherol onentrations were in the same range and not signifiantly different. When the three toopherols were analysed y organi versus onventional within loation and season, there were no lear signifiant differenes in management system aross the season and loation ominations (Tale 2, Figure S2). Carotenoids. Overall, Oregon had higher levels of lutein and -arotene ompared to Maine (Tale 2, Figure S3) and omparative levels of zeaxanthin (Tale 2, Figure S3). Spring produed higher levels of all arotenoids ompared to Fall levels in ontrast to the gluosinolates and the d- & - toopherol onentrations. There were no signifiant differenes etween organi and onventional for any arotenoid measured. When arotenoids were analysed y management system within loation and season, -arotene showed signifiantly lower levels in Maine in the Fall ompared to other loation and season ominations (Figure S3). Partitioning of variane omponents Gluosinolates. For gluoraphanin aross all trials in oth regions, Genotype (G) main effet aounted for the largest proportion of variane, followed y G6L6S interation (Tale 3). There was no Management (M) main effet, ut M ontriuted to the three (L6S6M and G6S6M) and four-way interations (G6L6S6M). In ontrast to gluoraphanin, Loation (L) had the largest effet for gluorassiin and neogluorassiin aross all trials in oth regions, followed y the L6S interations. For neogluorassiin the S and G main effet was more important than for gluorassiin. When trials were further partitioned y loation, a G and S main effet was apparent for neogluorassiin in oth loations; for gluorassiin the S main effets was only apparent in Oregon and not in Maine (Tale S2 in File S1). There was M main effet for gluorassiin and neogluorassiin, ut not for gluoraphanin, and no G6M interation for all gluosinolates. Toopherols. For d- and -toopherol aross all trials in oth regions, the Season (S) main effet aounted for the largest proportion of variane (Tale 3). In ontrast the proportion of the variation assoiated with S for a-toopherol aross all trials was minor. For all three toopherols there was minor to no M effet, ut a large L main effet, eing the greatest for -toopherol. The G main effet showed a similar pattern to L. Carotenoids. For all three arotenoids aross all trials in oth regions, the G main effet desried a signifiant omponent of total variane and was of largest influene for lutein (Tale 3). The S main effet played an important role for zeaxanthin, and to a lesser extent for lutein ut not for -arotene. For all three arotenoids the L effet was minor, ut the L6S interation for - arotene was relatively large and mostly assoiated with Maine (Tale S2 in File S1). There was no M main effet; only for - arotene was there a small effet of the G6M interation (mainly driven y Maine). Repeataility, geneti orrelation and ratio of orrelated response to diret response Organi versus onventional. In the present study, we were ale to estimate the proportion of the genotypi variane relative to phenotypi variane, ut eause we did not have a genetially strutured reeding population, we apply the term repeataility rather than road sense heritaility. Of the phytohemials studied, repeatailities for onentrations of seven of the nine were omparale or higher in organi ompared to onventional systems (Tale 4). Only for gluorassiin and d-toopherol was repeataility under organi onditions lower than under onven- PLOS ONE www.plosone.org 4 July 2014 Volume 9 Issue 7 e95683
Tale 2. Trait means 1 of phytohemials of 23 rooli ultivars grown aross four pair ominations of loation (Maine/Oregon), season (Fall/Spring) two-years omined and management system (Conventional/Organi), 2006 2008. Maine Oregon Fall Spring Fall Spring 2006 2007 Comined 2007 2008 Comined Mean 2006 2007 Comined 2007 2008 Comined Mean C O C O C O C O Gluoraphanin 5.31 e 3.77 3.56 4.06 4.18 3.46 3.03 a 4.64 d 4.51 d 3.91 Gluorassiin 1.06 0.90 a 1.45 1.33 1.19 5.14 5.51 g 2.24 d 2.70 e 3.90 Neogluorassiin 0.46 a 0.40 a 2.16 1.85 1.22 2.34 3.20 d 4.32 e 5.10 3.74 d-toopherol 2.34 2.77 d 1.91 1.70 a 2.18 3.53 e 3.66 e 1.91 2.24 2.83 -Toopherol 4.67 4.40 2.63 a 2.98 3.67 8.48 d 8.73 d 3.31 3.22 5.94 a-toopherol 25.83 a 27.33 a 38.61 40.51 33.07 43.04 43.20 40.52 42.25 42.25 Lutein 11.49 a 12.47 a 15.53 15.93 13.85 15.91 16.04 16.48 17.81 16.56 Zeaxanthin 0.81 a 0.83 a 0.87 a 0.88 0.85 0.83 a 0.84 a 1.02 1.02 0.93 -Carotene 12.98 a 13.25 a 28.73 29.71 21.16 29.10 30.10 25.16 25.80 27.54 1 Values in the tale are means. Means of the same letter in the same row are not signifiantly different at the P,0.05 level. doi:10.1371/journal.pone.0095683.t002 f f Brooli Breeding for Organi Agriulture and Phytohemial Content PLOS ONE www.plosone.org 5 July 2014 Volume 9 Issue 7 e95683
Tale 3. Partitioning of variane omponents (%) presented as oeffiients of variation for phytohemials of 23 rooli ultivars grown aross eight pair ominations of loation (Maine/Oregon), season (Fall/Spring) and management system (Conventional/Organi), 2006 2008. Genotype (G) G6L G6S G6M G6L6S G6L6M G6S6M G6L6S6M Residual L6S6M6R Rep (R) Management (M) L6S L6M S6M L6S6M Season (S) Loation (L) Gluoraphanin 0.01 5.45 0.00 7.20 0.01 0.00 11.86 1.56 17.45 0.01 0.01 0.00 15.97 0.01 7.49 12.62 11.94 Gluorassiin 36.00 0.00 0.00 27.51 5.58 4.18 1.34 1.86 9.42 7.77 0.01 0.00 13.84 0.00 0.00 10.63 10.91 Neogluorassiin 36.81 13.51 0.00 34.36 8.47 6.84 4.50 4.76 15.16 6.24 0.01 0.00 16.40 0.00 0.01 13.40 15.80 d-toopherol 6.83 35.22 0.43 7.87 0.01 0.01 3.39 0.01 5.57 5.65 6.01 0.00 13.65 0.01 0.00 12.21 12.74 -Toopherol 12.02 19.09 0.01 12.11 0.01 3.64 4.47 2.12 13.79 4.85 15.82 0.00 12.95 0.00 0.01 11.08 12.03 a-toopherol 6.73 0.01 0.28 10.20 0.01 0.01 0.01 1.29 2.79 3.42 0.01 0.97 10.07 0.00 0.00 8.18 8.92 Lutein 3.71 4.91 0.00 7.52 0.01 2.70 1.85 1.40 13.03 4.14 0.01 0.01 10.76 0.73 0.01 9.21 9.95 Zeaxanthin 1.97 11.55 0.01 3.99 0.01 0.00 0.01 0.00 8.44 3.18 0.01 1.05 6.91 0.01 0.83 8.52 11.36 -Carotene 4.61 0.00 0.00 17.84 0.01 0.70 0.01 0.71 8.41 4.45 0.00 2.31 11.32 4.63 0.65 12.83 10.99 doi:10.1371/journal.pone.0095683.t003 tional. In the analyses d- and a-toopherol had relatively low repeatailities. The highest repeatailities were for gluoraphanin (0.82 0.84), neogluorassiin (0.75 0.76), -toopherol (0.72 0.75), lutein (0.83 0.85) and zeaxanthin (0.76 0.77). Geneti orrelations were high etween organi and onventional for the gluosinolates, -toopherol and lutein (0.84 0.95), while d- toopherol, a-toopherol, zeaxanthin and -arotene were lower (0.63 0.77). The ratio of the orrelated response to diret response for seletion in the organi system was less than 1.0 for all traits. By loation and season. For the gluosinolates, gluoraphanin and gluorassiin repeataility at eah loation, season and treatment trial were omparale and generally high (0.83 0.97) etween organi and onventional trials, while no lear trend for neogluorassiin repeatailities was oserved etween organi and onventional aside from eing muh lower than gluoraphanin and gluorassiin (Tale S3 in File S1). For - and a- toopherol, repeatailities were omparale etween organi and onventional, while for d-toopherol repeatailities were omparale etween systems or higher in onventional exept for one paired trial. For the arotenoids, repeatailities were omparale or higher in organi for all paired trials, while for lutein in seven of the eight paired trials organi was omparale or greater than onventional. Repeatailities for zeaxanthin onentrations were omparale for six of the eight paired trials. Comparison of ultivar ranking for phytohemial onentration and staility aross trials To determine trends in ultivars with oth the highest onentration of phytohemial groups most stale aross loations, seasons and prodution systems, phytohemial onentrations were plotted against staility per genotype aross trials. A group of ultivars were identified as oth highest in onentration and most stale and are indiated in the highlighted red irle per phytohemial (Figure 1A I). For gluoraphanin, the same group of ultivars had oth the highest onentrations and were the most stale aross prodution systems (Figure 1A; Tale S4 in File S1). While for gluorassiin, a different set of ultivars had the highest onentrations aross prodution systems (Figure 1B; Tale S5 in File S1). Overall staility of all ultivars aross prodution system was less related to ultivar mean onentrations for gluorassiin than for gluoraphanin. None of the ultivars with the highest onentration for neogluorassiin were in the top quartile for staility aross trials; all ultivars with the highest neogluorassiin ontent were in the ottom half for staility (Figure 1C; Tale S6 in File S1). Some ut not all ultivars that had the highest onentrations of a-toopherol were among the top group for d- and/or -toopherol. There was no relationship etween d-toopherol onentrations and staility, ut oth - and a- toopherols had higher onentrations assoiated with greater staility (Figure 1D F; Tales S7 9 in File S1). Open pollinated and early maturing ultivars had the highest and most stale onentrations for all arotenoids. (Figure 1G I; Tales S10 12 in File S1). Comparison of phytohemial onentration y genotype lassifiation The open pollinated and F 1 hyrid ultivars were ompared aross trials for eah phytohemial analysed (Figure 2A). The levels of gluoraphanin in F 1 hyrids tended to e higher than the open pollinated ultivars. But the inverse trend was oserved for gluorassiin, whih was supported y the ranking and staility analysis where the F 1 hyrids showed higher levels and more staility aross trials than the open pollinated ultivars for PLOS ONE www.plosone.org 6 July 2014 Volume 9 Issue 7 e95683
Tale 4. Repeatailities, geneti orrelation and ratio of orrelated response to diret response for rooli phytohemials omparing organi versus onventional management systems over all trial season/loation ominations, 2006 2008. Repeataility (H) C O r A a CR org /R org Gluoraphanin 0.84 0.82 0.84 0.83 Gluorassiin 0.70 0.64 0.88 0.84 Neogluorassiin 0.75 0.76 0.94 0.94 d-toopherol 0.50 0.42 0.73 0.66 -Toopherol 0.75 0.72 0.95 0.93 a-toopherol 0.23 0.35 0.61 0.76 Lutein 0.83 0.85 0.93 0.94 Zeaxanthin 0.76 0.77 0.77 0.78 -Carotene 0.62 0.72 0.63 0.68 a Average geneti orrelation etween onventional and organi prodution systems aross loations. Ratio of orrelated response to diret response. doi:10.1371/journal.pone.0095683.t004 Figure 1. Brooli ultivar staility from trials onduted in two loations over four seasons with two management systems plotted against phytohemial ontent. A. Gluoraphanin, B. Gluorassiin, C. Neogluorassiin, D. d-toopherol, E. -toopherol, F. a-toopherol, G. Lutein, H. Zeaxanthin, I. -arotene. See Tale 1 for ultivar name areviations. The C or O at the end of the ultivar areviation indiates onventional or organi management system, respetively. doi:10.1371/journal.pone.0095683.g001 PLOS ONE www.plosone.org 7 July 2014 Volume 9 Issue 7 e95683
Figure 2. Mean phytohemial ontent of rooli genotypi lasses. A. Mean phytohemial ontent of rooli F 1 hyrids versus open pollinated ultivars, and B. Mean phytohemial ontent of early, mid- and late-maturing ultivars grown aross all trials at two loations (Maine and Oregon), in two seasons (Fall and Spring) and in two management systems (Conventional and Organi) and onventional management systems. See Tale 1 for key to ultivar F1 hyrid versus open pollinated lassifiation and maturity lassifiation. Signifiane (* = P,0.05, ** = P,0.01, *** = P, 0.001). doi:10.1371/journal.pone.0095683.g002 gluoraphanin. The reverse was oserved for gluorassiin. For the arotenoids, the open pollinated ultivars had a signifiantly higher mean value of lutein and zeaxanthin and tended to e higher for -arotene ompared to the F 1 hyrids. Based on the results of our field trials, the 23 ultivars of rooli were grouped into three distint maturity lasses: Early (55 63 days); Mid (64 71 days); and Late (72 80 days) and analysed for the effet of the maturity lass on phytohemial ontent (Figure 2B). For gluoraphanin, late maturing ultivars had signifiantly higher ontent levels, while for the arotenoids, early maturing ultivars tended to have higher onentrations and were signifiantly higher for lutein. When ultivar performane etween geneti material originating from two primary rooli reeding ompanies was ompared for phytohemial ontent there were no signifiant differenes with the exeption of lutein, where ompany 1 s ultivars had signifiantly higher onentrations than those of ompany 2 (data not shown). A negative orrelation etween the date of release and levels of gluorassiin (R 2 = 0.21; p = 0.03) (Figure 3) was oserved, ut no signifiant orrelations for any other phytohemial were seen when 21 ultivars (the total set minus the two inred lines) were analysed y their date of ommerial release (1975 2005). Correlation analysis among phytohemials and hortiulture traits Phytohemial orrelation aross trials. Correlation among phytohemials indiated that gluoraphanin was signifiantly negatively orrelated to gluorassiin (Tale 5). Correlations etween the gluosinolates and the toopherols were not signifiant. Gluoraphanin and neogluorassiin were negatively orrelated to all arotenoids ut only lutein and gluoraphanin were statistially signifiant. Gluorassiin demonstrated a positive trend with all arotenoids. No statistially signifiant orrelations were oserved within toopherols. D-toopherol was positively orrelated, while -toopherol was negatively orrelated to all arotenoids. There were no signifiant orrelations for a- toopherol with arotenoids. All arotenoids were highly positively orrelated with one another. Phytohemial orrelation to hortiulture traits aross trials. A orrelation analysis was onduted for six hortiulture traits, derived from the field study omponent of this researh, Renaud et al. [78], and the nine phytohemials aross trials. The results indiated that greater head weight and head diameter were signifiantly positively orrelated with gluoraphanin and negatively orrelated with gluorassiin, d-toopherol and the arotenoids. Inreasing days to maturity was positively orrelated with gluoraphanin, and negatively orrelated to arotenoids. PLOS ONE www.plosone.org 8 July 2014 Volume 9 Issue 7 e95683
Figure 3. Regression of rooli floret gluorassiin onentrations on date of ultivar release for 23 ultivars grown aross all trials in two loations (Maine and Oregon), in two seasons (Fall and Spring), in two management systems (Conventional and Organi), 2006 2008. doi:10.1371/journal.pone.0095683.g003 Head olor was signifiantly orrelated with d-toopherol and the arotenoids, ut not with gluosinolates or - and a-toopherol. Bead size and ead uniformity were positively orrelated with gluoraphanin, neogluorassiin and -toopherol and negatively orrelated with gluorassiin and the arotenoids. Prinipal omponent iplot analysis: orrelation etween phytohemials and ultivars y prodution system In the prinipal omponent analysis the first PC axis aounted for similar amounts of the total variation in oth onventional and organi prodution systems (43.5% vs. 39.6%). The seond PC axis showed a similar trend with 17.02% for onventional and 16.93% for organi (Figures 4A and 4B). The first two PC axes together aounted for 60.53% and 56.57% of total variation for onventional and organi, respetively. The PCA iplot analysis supported our findings that arotenoids were highly assoiated aross systems, while toopherols were highly assoiated in onventional, ut not in organi (toopherols demonstrated the largest shift etween prodution systems). Gluoraphanin and neogluorassiin were assoiated with one another, ut not with gluorassiin aross prodution systems. Gluoraphanin was assoiated with a-toopherol in organi, ut not in onventional treatments. Gluorassiin was assoiated with d- and a-toopherol in onventional, ut not in organi treatments. d-toopherol had a higher assoiation with the arotenoids in organi than onventional. The iplots show response of oth ultivars and phytohemial traits to environment. Those ultivars lose to the origin reveal little aout the relationship of ultivars and trait vetors, whereas those loated near the extremes of trait vetors are those with the highest (or lowest) values for those traits. Disussion Impat of organi management system ompared to environmental fators on phytohemial ontent Few studies have speifially ompared the levels of health promoting ompounds in Brassia vegetale speies grown under organi and onventional prodution systems [13,74 75]. To our knowledge, this investigation is the most omprehensive study with the roadest range of phytohemial ompounds (9) and a diverse set of rooli ultivars (23) over regions (2), and management systems (2), with Fall and Spring season trials (2 eah). In this study organi versus onventional management systems ontriuted the smallest soure of variation ompared to genotype, region and season. Within the phytohemials studied individual ompound onentrations responded differently. All ompounds showed geneti variation, ut also a sustantial proportion of variane omponents were aounted for y high level interations (Tale 3; Tale S2 in File S1). While M main effet was generally small, it had a sustantial ontriution in three- and four-way interations. In partiular, many G6L6S6M interations were large relative to other variane omponents. This indiates that for the phytohemials, M did have an influene on G, ut that there were no onsistent patterns aross loations and seasons that would have shown up as signifiant G6M. Rather in eah season and loation, the paired organi and onventional environments differed signifiantly from one another ut eah situation was unique. In ontrast to many omparisons etween organi and onventional prodution systems [79], it should e noted that in our trials, yields averaged over the years did not differ signifiantly etween the organi and onventional management systems [78]. Among the nine ompounds, gluoraphanin was the most strongly influened y genotype followed y lutein: supporting the findings of several other rooli studies where variation in onentrations for gluoraphanin [19,70,65 66] and lutein [21,37] was primarily due to genotype. For -toopherol, genotype was a large soure of variation, ut this ompound was equally influened y loation and season (also found y Irahim and Juvik [37]). For gluorassiin and neogluorassiin the loation was the largest soure of variation, ut also L6S interation was very influential, partiularly for neogluorassiin, whih is supported y Kushad et al. [20] and Shonhof et al. [26]. Jasmoni aid, a signal transdution ompound in plants, is upregulated under onditions of plant stress, wounding, and herivory. Inreased endogenous levels or exogenous appliation of this ompound (or methyl jasmonate) inreases iosynthesis and transport of neogluorassiin to rooli florets. This up- PLOS ONE www.plosone.org 9 July 2014 Volume 9 Issue 7 e95683
Tale 5. Correlations oeffiients (r) for six hortiultural traits and nine phytohemials, alulated using data standardized aross trials. Head Color Head Weight Head Diameter Maturity Bead Size Bead Uniformity Gluoraphanin Gluorassiin Neogluorassiin d-toopherol -toopherol a-toopherol Lutein Zeaxanthin -Carotene Head Weight Head Diameter 0.81 Maturity Head Color Bead Size 0.63 0.69 Bead Uniformity 0.49 0.48 Gluoraphanin 0.47 0.44 0.43 0.63 0.51 Gluorassiin 20.54 20.50 20.56 20.64 20.51 Neogluorassiin 0.58 0.48 d-toopherol 20.55 0.49 -Toopherol 0.43 a-toopherol Lutein 20.65 20.70 0.56 20.69 20.41 0.55 20.54 Zeaxanthin 20.68 20.43 20.62 0.49 20.64 0.60 20.42 0.95 -Carotene 20.53 20.54 0.59 20.48 0.50 20.43 0.90 0.90 Correlation results inlude means from 23 ultivars, aross eight pair ominations of loation (Maine/Oregon), season (Fall/Spring) and management system (Conventional/Organi), 2006 2008 a. a For empty ells, r is not signifiantly different from zero (P,0.05). doi:10.1371/journal.pone.0095683.t005 PLOS ONE www.plosone.org 10 July 2014 Volume 9 Issue 7 e95683
Figure 4. Prinipal omponents iplot of phytohemials (vetors) and 23 ultivars (irles) grown in four seasons in Oregon and Maine. A. Biplot for onventional prodution, B. Biplot for organi prodution. See Tale 1 for ultivar name areviations. Trait areviations: GLR: Gluoraphanin; GLB: Gluorassiin; NGB: Neogluorassiin; DTO: d-toopherol; GTO: -toopherol; ATO: a-toopherol; LUT: Lutein; ZEA: Zeaxanthin; BCA: -Carotene. doi:10.1371/journal.pone.0095683.g004 regulation was not oserved for gluorassiin iosynthesis [17]. This ould explain why neogluorassiin was primarily under the ontrol of Loation and L6S interation in our study. Season was the largest variane omponent for d-toopherol and zeaxanthin, whih ontrasts with the work of Irahim and Juvik [37] who found genotype had the largest influene on these ompounds, followed y genotype y environment interation although this study was onstrained y the fat that the experiment was onduted in only one loation over two growing seasons. For the other ompounds suh as a-toopherol and -arotene, L6S and the G6L6S interations were most important. Overall we found high geneti orrelations etween gluosinolates in organi and onventional trials. When trial loations were analysed separately, M main effet was present for gluorassiin and neogluorassiin. The mean onentrations of gluorassiin and neogluorassiin in rooli from Oregon organi trials had higher onentrations ompared to Oregon onventional trials, while Maine trials were omparale etween management systems (Tale 2, Figure S1). These results an e explained y the larger environment effet on gluorassiin and genotype y environment effet on neogluorassiin found in the variane omponent analysis indiating sensitivity of these ompounds to aioti and/or ioti stresses. Our loation speifi findings are supported y those of Meyer and Adam [13] who performed a omparative study of the gluosinolate ontent of store ought organi and onventional rooli and determined that the indolyl gluosinolates, gluorassiin and neogluorassiin were signifiantly higher in the organially grown versus the onventionally grown. Evaluation of 10 rooli genotypes over two years y Brown et al. [19] further supports our findings and those of Rosa and Rodrigues [62], Vallejo et al. [64], and Farnham et al. [70], that variation in onentration for gluoraphanin was primarily due to geneti variation, while differenes in gluorassiin was due to environmental variation (e.g. season, temperature) and genotype y environment interation. The signifiantly higher levels of gluorassiin in Oregon in the Fall harvested trials ompared to Maine ould e attriuted to the higher maximum temperatures and GDD in Oregon ompared to Maine (Tale S1 in File S1). Compared to gluosinolates, there is sustantially less researh on the genotype y environment interation of toopherol and arotenoid phytohemial groups in rooli, and no speifi studies exploring the influene of organi prodution system. In our study, minor management system effet at the overall trial analysis level was oserved for the toopherols and for arotenoids, there was management system effet only for lutein in Oregon Spring trials. Pihi et al. [75] also did not find differenes in levels of arotenoids in auliflower in organi versus onventional systems. In the toopherols, there were no signifiant differenes in loation, ut for d- and - toopherol onentration levels were higher in the fall ompared to the spring, while for a-toopherol, onentration levels were higher in the spring ompared to the fall. For the arotenoids, there were no signifiant loation differenes, however there was a seasonal trend that all arotenoids were higher in spring ompared to fall. Irahim and Juvik [39] found signifiant environmental variation among 24 rooli ultivars for arotenoids and toopherols whih they attriuted to the stressful prodution environments. Fators explaining the genotype and genotype y environment interation omponents of variation in the arotenoids and toopherols ould e larified y the fat that environmental stimuli are oth up- and down-regulating genes assoiated with arotenoid and toopherol iosynthesis. There is evidene in the literature that there are oordinated responses of the arotenoid and toopherol antioxidants in vivo. There was a redution in rape seed (Brassia napus) toopherol ontent in response to inreased arotenoid levels due to over expression of the enzyme phytoene synthase [80]. This response ould explain the negative orrelation etween -toopherol onentration and the arotenoids oserved in our trials. PLOS ONE www.plosone.org 11 July 2014 Volume 9 Issue 7 e95683
Differenes in phytohemial ontent etween different genotypes and genotypi lasses The partitioning of variane indiated that genotype was an important soure of variation for all gluosinolates. The ultivar ranking and rank orrelation analysis demonstrated that there was a pattern in genotype ontent of gluosinolates where ultivars with the highest onentrations of gluoraphanin had the lowest levels for gluorassiin (Figures S1). In our trials, the range in gluoraphanin onentrations aross ultivars was (1.15 7.02 mmol/g DW, Tale S4 in File S1), while gluorassiin was 1.46 3.89 mmol/g DW, Tale S5 in File S1). Several of the ultivars with the highest onentrations of neogluorassiin were those that had the highest onentrations of gluorassiin. Range in neogluorassiin onentrations aross ultivars was 0.68 4.54 mmol/g DW, Tale S6 in File S1). In earlier studies, gluosinolate onentrations in rooli have shown dramati variation among different genotypes. Rosa and Rodriguez [62] studied total gluosinolate levels in eleven ultivars of rooli and found ranges from 15.2 59.3 mmol/g DW. Among 50 aessions of rooli Kushad et al. [20] found gluoraphanin ontent ranges from 0.8 22 mmol/g DW with a mean onentration of 7.1 mmol/ g DW, while Wang et al. [61] found gluoraphanin ontent of five ommerial hyrids and 143 parent materials ranging from 1.57 5.95 mmol/g for the hyrids and 0.06 24.17 mmol/g in inred lines and Charron et al. [65] found ranges from 6.4 14.9 mmol/g DW. While the means in our study are somewhat lower, they are within the range of other studies. A genotype effet was oserved for toopherols, ut predominantly for -toopherol. The PCA iplots (Figure 4AB) and the orrelation analysis (Tale 5) demonstrated the high positive orrelations etween d-toopherol, a-toopherol and the arotenoids (a-toopherol and -arotene were also highly orrelated in the Kushad et al. [20] study). The ultivar relationship to different phytohemials was represented in the iplots as well as in the ultivar ontent and staility analysis (Figure 1). Many ultivars with the highest onentrations in the toopherols and arotenoids were open pollinated ultivars, inreds and early maturing, older F 1 hyrids. Many of this same group were also relatively high in gluorassiin onentrations. Kurilih et al. [38] found that arotenoid and toopherol onentrations among 50 rooli lines were highly variale and primarily genotype dependent. Speifially, levels of -arotene ranged from 0.4 2.4 mg/100 g FW. Irahim and Juvik [37] also found road ranges for total arotenoid and toopherol onentrations among 24 genotypes ranging from 55 154 mg/g DW and 35 99 mg/g DW, respetively. Farnham and Kopsell [21] studied the arotenoid levels of nine doule haploid lines of rooli. Similar to our findings, lutein was the most aundant arotenoid in rooli ranging from 65.3 139.6 mg/g DM. The soures of variation for lutein were predominantly genotype, followed y environment and G6E interation, whih also supports our findings. No genotypi differenes were found for -arotene in Farnham and Kopsell [21], whih is in ontrast to our findings. Overall, they found that most of the arotenoids measured were positively and highly orrelated to one another as was oserved in our study (Tale 5). Kopsell et al. [81] found lutein levels in kale of 4.8 13.4 mg/100 g FW where the primary variane omponents for oth lutein and -arotene were also genotype and season. Our researh aimed also to address the question whether the phytohemial ontent of rooli ultivars is assoiated with ertain genotypi lasses, e.g. open pollinated vs. F 1 hyrids; older vs. newer ultivar releases; and etween ommerial soures. Brooli is typially a ross-pollinated, self-inompatile rop speies and ultivars are either open pollinated and omposed of heterogeneous genetially segregating individuals, or F 1 hyrids produed y rossing of two homozygous inred lines, resulting in homogeneous populations of heterozygous individuals. In the 1960 s virtually all rooli grown was derived from OPs. By the 1990 s almost all ommerial ultivars were hyrids [82]. In our trials with 18 F 1 hyrids (released etween 1975 2005) and 3 open pollinated ultivars (released from 1985 2005), we found several interesting trends related to genotype and genotypi lass performane as it related to the three groups of phytohemials. When analysing F 1 hyrid and open pollinated ultivars, they also demonstrated different performane patterns depending upon the individual phytohemial or group of ompounds analysed. When ultivars were ranked for ontent and staility per phytohemial, there were distint trends for ertain ompounds suh as late maturing, F 1 hyrids outperforming early maturing F 1 hyrids and open pollinated ultivars for gluoraphanin, while the inverse was found for gluorassiin and all arotenoids studied. This analysis was further supported y the PCA iplots that showed a strong relationship for selet ultivars to ertain phytohemials or groups of phytohemials suh as OSU OP to the arotenoids. When the full set of ultivars was divided into F 1 hyrid and open pollinated groups and the means ompared y phytohemial, the results further supported the individual ultivar analysis where F 1 hyrids had higher mean values for gluoraphanin than the open pollinated ultivars (Figure 2A). Clear ultivar performane differenes were identified where early maturing versus late maturing ultivars performed differently depending upon the phytohemial (Figure 2B). We also found that late maturing ultivars had higher onentrations for gluoraphanin than early maturing lines (and the inverse for gluorassiin and the arotenoids). Pihi et al. [75] studied the quantity of gluosinolates of an early and late maturing ultivar of auliflower grown in one onventional and three organi prodution systems, and found a signifiantly higher level of gluoraphanin in the later maturing ultivar ompared to the early maturing ultivar in the organi prodution system. Another interesting trend was that ultivars with higher onentration levels for those phytohemials whose expression is heavily influened y environmental fators were not neessarily the most stale aross trial environments; as was the ase with neogluorassiin, d- and -toopherol in our study. For traits where genotype played a more signifiant role in ontriuting to variation, ultivars with a higher onentration level tended to also e those that were most stale aross environments as was seen for lutein and gluoraphanin onentrations. No signifiant differenes were found for ultivar performane in phytohemial onentrations etween geneti materials originating from two distint ommerial soures, with the exeption of lutein (data not shown). When the full set of rooli ultivars were analyzed for a orrelation etween date of release and mean level of phytohemial ontent aross trials, no signifiant orrelation was found with the exeption of a negative trend for gluorassiin (Figure 3). Our data does not support the idea that modern reeding for high yield performane and disease resistane neessarily leads to a trade-off in level of phytohemials. Previous reports examining the relationship etween year of release and performane had foussed on wheat vitamin and mineral ontent [83 85], and mineral ontent in rooli [86 87]. However these authors did not study phytohemial ontent and their results were equivoal on the question on an innate iologial trade-off etween inreased yield and nutritional ontent. Not many studies have inluded two or more groups of phytohemials. In our study with three phytohemial groups we found that phytohemials demonstrating a negative orrelation PLOS ONE www.plosone.org 12 July 2014 Volume 9 Issue 7 e95683