George J.Vndemrk*,Michel A.Grusk,Reecc J.McGee

aUSDA-ARS,Grain Legume Genetics and Physiology Research Unit,Pullman,WA 99164,USA bUSDA-ARS,Red River Valley Agricultural Research Center,Fargo,ND 58102,USA

1.Introduction

Dietary mineral deficiencies impair the health of over three billion people globally,and are responsible for illnesses and deaths during all stages of development from infancy through adulthood[1].It is estimated that two billion people globally suffer from iron deficiency[2].Iron deficiency during pregnancy causes increases in premature deliveries,low birth weights,and maternal deaths,while deficiency during infancy and early childhood can reduce physical growth and development of cognitive functions[1].Humans also suffer considerably from diseases caused by zinc deficiency.Adequate levels of zinc are essential to fetal development,healthy birth,and subsequent physical growth[3].Approximately one billion people globally are estimated to be at risk of zinc deficiency[4],primarily in sub-Saharan Africa and southeast Asia[4].In Bangladesh,over 100 million people are at risk of arsenicosis associated with dietary selenium deficiency[5].Diets deficient in selenium have been associated with a higher incidence of prostate cancer[6].

Mineral deficiencies can often be treated by dietary consumption of mineral supplements or fortified foods[7].Unfortunately,in regions such as sub-Saharan Africa and southwest Asia that are strongly affected by mineral deficiencies[8],various socioeconomic constraints can limit availability of mineral supplements and fortified foods.Another approach proposed to reduce the incidence of mineral deficiencies is biofortification,in which crop plants have higher concentrations of minerals in edible parts[9].Biofortification may be accomplished by applying management practices that result in increased concentrations of minerals in edible parts,development of new cultivars with elevated mineral concentrations through plant breeding,or a combination of management and genetic approaches[9].

Several pulse crops,including chickpeas(Cicer arietinum L.)[10,11],and lentils(Lens culinaris Medik.)[11–15]have been proposed as targets for mineral biofortification.Currently,chickpeas are the second most important pulse crop in terms of global production,after dry beans(Phaseolus vulgaris L.),with over 14.2 Mt.produced in 2014[16].Nearly 4.9 Mt.of lentils were produced globally in 2014,fifth among pulse crops after dry beans,chickpeas,dry peas(Pisum sativum L.),and cowpeas(Vigna unguiculata L.)[16].

The development of new cultivars that stably express high concentrations of selected minerals requires an understanding of the magnitude of genetics,environment,and their interaction effects on mineral concentrations.Studies conducted in Saskatchewan(Canada)on chickpeas have shown that genotype,environment,and their interaction effects were significant for seed concentrations of copper,magnesium,selenium,and zinc[10,11].Similarly,studies on lentils grown in Saskatchewan showed that genotype,environment,and their interaction effects were significant for seed concentrations of copper,iron,magnesium,manganese,and selenium[11,12,14].Significant genotype effects on seed mineral concentrations were also observed for lentils grown in Turkey[15].

In the U.S.chickpeas and lentils are produced primarily in the Palouse”region of eastern Washington and northern Idaho and in the dryland regions of Montana and North Dakota.In 2016,approximately 360,000 ha of lentils were harvested in the U.S.with a value greater than$350 million[17].Chickpeas can be divided into kabuli and desi classes[18].Kabuli chickpea seeds typically have a rounder shape and are larger and lighter in color than desi chickpeas.Production of desi chickpeas in the U.S.is negligible[19].In 2016 over 129,000 ha of kabuli chickpeas were harvested in the U.S.with a value greater than$128 million[17].

The objectives of this study were to partition variance components conditioning seed mineral concentrations of advanced breeding lines and cultivars of chickpeas and lentils grown in Washington and Idaho,determine correlations between seed mineral concentrations and several agronomic traits including seed yield and seed size,and compare mineral concentrations of chickpeas and lentils grown in side-by-side field trials.

2.Materials and methods

2.1.Plant materials and field trials

This study examined 22 kabuli chickpea entries(Table 1),which included six cultivars,Dylan,Dwelley,Nash,Royal,Sawyer,and Sierra,and 16 breeding lines.All entries were planted in 2010 and 2011 at three locations:Kendrick,ID(46.6141°N,116.6465°W);Genesee,ID,(46.5507°N,116.9254°W),and Pullman,WA(46.7298°N,117.1817°W).Lentil entries(Table 2)included three cultivars,Avondale,Merrit,and CDC Richlea,and 13 breeding lines planted in 2010 and 2011 at two locations:Fairfield,WA(47.3852°N,117.1716°W)and Pullman,WA.

A seed treatment was applied prior to planting that contained the fungicides fludioxonil(0.56 g kg?1,Syngenta,Greensboro,NC,USA),mefenoxam(0.38 g kg?1,Syngenta),and thiabendazole(1.87 g kg?1,Syngenta),thiamethoxam(0.66 mL kg?1,Syngenta)for insect control,and molybdenum(0.35 g kg?1).Approximately 0.5 g Mesorhizobium ciceri inoculant(1 × 108CFU g?1;Novozyme,Cambridge,MA,USA)was applied to each chickpea seed packet one day before planting.Chickpeas were planted at a density of 43 seeds m?2in a 1.5 m × 6.1 m block(～430,000 seeds ha?1).Lentils were planted at a density of 86 seeds m?2in a 1.5 m ×6.1 m block(～860,000 seeds ha?1).All yield trials used a randomized complete block design with three replications.Weeds were controlled by a single post-plant/pre-emergence application of metribuzin (0.42 kg ha?1,BayerCrop Science,Raleigh,NC)and linuron(1.34 kg ha?1,NovaSource,Phoenix,AZ,USA).

Once plots at Pullman,WA began to flower,they were visited every two days and the percentage of flowering plants within a plot was estimated to determine the number of days after planting required for 50%of the plants to flower(DF).Plots at all locations were mechanically harvested and seed yield(kg ha?1)was determined for all entries at all locations.One hundred-seed weights(g,HSW)were determined for all entries in 2010 and 2011 at Pullman,WA by taking the weight of 100 random seeds sampled from each of three plots for each entry.These data along with pedigrees and cotyledon color are presented for chickpea in Table 1 and for lentil in Table 2.

2.2.Analytical methods

Mineral concentrations of entries were determined using 50 g seed(at 14%moisture)for each plot.Each seed sample was ground in a stainless-steel mill to a uniform powder,following drying in a 60°C drying oven.Two 0.5 g subsamples(dryweight basis)from each plot were wet-digested using HNO3–H2O2[20]and analyzed by inductively coupled plasma opticalemission spectrometry(ICP-OES)(CIROS ICP Model FCE12;Spectro,Kleve,Germany)to determine concentrations of boron(B),calcium(Ca),copper(Cu),iron(Fe),potassium(K),magnesium(Mg),manganese(Mn),nickel(Ni),phosphorus(P),selenium(Se),sulfur(S),and zinc(Zn).Discrepancies higher than 5%between subsamples resulted in the analysis of a third subsample.The subsample values were averaged to determine the mineral concentration of each plot.

Table 1–Selected characteristics of chickpea entries evaluated for seed mineral concentrations.

2.3.Statistical analysis

Entries(genotypes)were considered fixed factors and locations(environments)and replications(blocks)within locations were considered random factors.Data for each mineralwere subjected to Levene's test[21]to determine whether variances were homogeneous across locations and years.Combined ANOVA was conducted across all years and locations to detect effects of genotypes,environments,years,and their interactions.Entry means were compared using Tukey's HSD(α=0.05).Pairwise correlations were determined between seed mineral concentrations and yield from data combined across all locations and years,and correlations were also determined between mineral concentrations and HSW and DF for data obtained at Pullman,WA combined across both years.Seed mineral concentrations were compared between chickpea and lentil entries using data obtained from Pullman,WA in 2010 and 2011,where the two species were grown in adjacent trials.Because the two species were not grown together in a randomized complete block design,only means and standard errors are presented for these sideby-side comparisons.All statistical analyses were performed with JMP software(SAS,Cary,NC,USA).

Table 2–Selected characteristics of lentil entries evaluated for seed mineral concentrations.

3.Results

3.1.Chickpea seed mineral concentrations

Mean squares of combined analysis of variance for chickpea seed mineral concentrations are presented in Table 3.The partitioning of variance components indicates that for the majority of mineral concentrations the location,year,and location×year effects were much higher than genotype effects.Genotype effects were significant for all minerals except Se.Genotype effects were highest for Ca,followed by Cu and Mn.Location effects were significant for all minerals and were highest for Ni,followed by Cu and Zn.Year effects were significant for all minerals except S.For several minerals,including B,Ca,Fe,K,and Mg,the majority of the total variation was explained by year effects.Genotype×location effects were significant for all minerals except B and Se.Genotype×year effects were significant for all minerals except B.For the majority of minerals the highest interaction effect was the location×year effect,which was significant for all minerals.Genotype×location×year effects were significant for all minerals except B,and explained nearly all of the total variation in S concentration.

Means comparisons of seed mineral concentrations between chickpea entries are presented in Tables 4 and 5.Se was the least abundant seed mineral present,with a mean concentration of 0.35 μg g?1,while the most abundant mineral in chickpea seed was K(＞10,000 μg g?1).Significant differences between means were detected for all minerals except B,K,Ni,and Se.Dylan had the highest Ca concentration among entries,significantly higher than those of all other entries except Nash and CA04900421C.CA0790B0053C had the highest Cu concentration among entries.Cu concentration of Royal was significantly lower than those of all other cultivars.CA0469C025C had the highest Fe concentration among all entries.Among cultivars,Dwelley had a significantly higher seed concentration of Fe than Royal and Sawyer.Dwelley had the highest Mg concentration among entries but there were no significant differences among cultivars.CA0790B0155 had the highest Mn concentration among entries.The Mn concentration of Sierra was significantly higher than that of Royal,while that of Nash was significantly lower than those of all other cultivars.Dwelley had the highest concentration of P among all entries,but there were no significant differences among cultivars.Sawyer had the highest concentration of S among all entries,significantly higher than those of all other cultivars.CA0790B0053C had the highest concentration of Zn among all entries and there were no differences between cultivars.

3.2.Lentil seed mineral concentrations

Mean squares of combined analysis of variance for lentil seed mineral concentrations are presented in Table 6.Genotype effects were significant for all minerals except Se.Genotype effects were highest for Fe,followed by Mg and Cu.Location effects were significant for all minerals except Mg and S and were highest for Ni,followed by B and Fe.Year effects were significant for all minerals except Fe and Mg.For several minerals,including K,Mn,P,S,and Zn,almost all of the total variation was explained by year effects.Genotype×locationeffects were significant for Cu,Mn,Ni,and P,although these interaction effects were much lower than other sources of variance.Genotype×year effects were significant for Ni only.For the majority of minerals the highest interaction effect was the location×year effect.Nearly all of the total variation for Mg seed concentration was explained by location×year effects.Genotype×location×year effects were significant for Cu,P,and S,although these interaction effects were much lower than other sources of variance.

Table 3–Mean squares of combined ANOVA and coefficient of variation(CV)for seed mineral concentrations in chickpea cultivars and breeding lines grown in Idaho and Washingtona.

Means comparisons of seed mineral concentrations between lentil entries are presented in Tables 7 and 8.Se was the least abundant seed mineral present,with a mean concentration of 0.43 μg g?1,while the most abundant mineral in lentil seed was K(＞10,000 μg g?1).Significant differences between means were detected for B,Ca,Cu,Fe,and Mg.The breeding line LC07600591R had the highest B concentration,significantly higher than those of Avondale and CDC Richlea.The breeding line LC07600524L had the highest concentration of Cu,but was not significantly different from Merrit.The breeding line LC07600376R had the highest Ca concentration but was not significantly different from any of the three cultivars.LC07600151R had the highest Fe concentration,which was significantly greater than those of all other entries.Avondale had the highest concentration of Mg,but was not significantly different from Merrit or CDC Richlea.

3.3.Correlations between mineral concentrations,yield,100-seed weight,and days to 50%flowering

Pairwise correlations between different chickpea seed mineral concentrations,yield,HSW and DF are presented in Table 9.Significant correlations were observed for the majority of pairwise combinations.The highest positive correlations between mineral concentrations were observed between P and Zn,P and K,and Fe and Mg.The highest negative correlations were between Ni and Zn,Ni and P and B and K.Seed concentrations of Fe,B,and Ca had the highest positive correlations with yield,while the highest negative correlations between yield and mineral concentration were observed for P,K,and Zn.Ni,Fe,and Cu had the highest positive correlations with HSW.K,P,Zn,and Mg had the highest negative correlations with HSW.K,P and Zn had the highest positive correlations with DF.Ni,Fe,and Mn had the highest negative correlations with DF.

Pairwise correlations between different lentil seed mineral concentrations and between mineral concentrations and agronomic traits including yield,seed size,and DF are presented in Table 10.Significant correlations were observed for the majority of pairwise combinations.The highest positive correlations between mineral concentrations were observed between P and K,P and Zn,and K and Zn.The highest negative correlations were between K and Mn,Zn and Mn,and P and Ca.Seed concentrations of Mn,Ca,and Fe had the highest positive correlations with yield,while the highest negative correlations between yield and mineral concentration were observed for K,P,and Zn.Mn,Ni,and B had the highest positive correlations with HSW.K,Zn,and Mg had the highest negative correlations with HSW.K,P,and Zn had the highest positive correlations with DF.In contrast,Mn,Ni,and Ca had the highest negative correlations with DF.

3.4.Comparisons between chickpeas and lentils

Comparisons of mineral concentrations between lentils and chickpeas grown in adjacent plots are presented in Table 11.Chickpeas had higher concentrations of Ca,Mg,and Mn than lentils,whereas lentils were higher in Cu,Fe,P,and Zn.Concentrations of B,Ni,K,and S were very similar for both species,which had equal seed concentrations of Se.

Table 4–Meanaseed concentrations of boron(B),calcium(Ca),iron(Fe),potassium(K),and magnesium(Mg)for chickpea breeding lines and cultivars grown at two locations in Washington and one location in Idaho in both 2009 and 2010.

Table 5–Meanaseed concentrations of manganese(Mn),nickel(Ni),phosphorus(P),sulfur(S),selenium(Se),and zinc(Zn)for chickpea breeding lines and cultivars grown at two locations in Washington and one location in Idaho in both 2009 and 2010.

4.Discussion

Genotype effects,although significant in chickpeas and lentils for all minerals except Se,tended to be minimal compared to other sources of variance.The chickpea and lentil entries were included based on agronomic performance and were not developed with any selection for mineral concentrations.The minor genotype effects observed for chickpeas are likely due to limited genetic variation for seed mineral concentrations among the breeding lines and cultivars(Table 1),which included 10 entries that share CA9783142 as a parent.The higher magnitude of environmental effects compared to genetic effects indicates that only limited gains in seed mineral concentrations can be made through selections ofprogeny derived from crosses among these elite chickpea breeding lines and cultivars.Several of our observations are consistent with results for chickpeas grown in Saskatchewan,Canada[11],where location,year,and their interaction effects were the predominant sources of variance for several minerals,including Mg,Fe,Cu,and Zn.Among the minerals evaluated in chickpeas by Ray et al.[11],genotype effects were highest for Ca,which was also the case in our study.

Table 6–Mean squares of combined ANOVA and coefficient of variation(CV)for seed mineral concentrations of lentil cultivars and breeding lines grown in Washingtona.

Table 7–Meanaseed concentrations of boron(B),calcium(Ca),iron(Fe),potassium(K),and magnesium(Mg)for lentil breeding lines and cultivars grown at two locations in Washington in both 2009 and 2010.

Genotype effects in lentils were highest for Fe(Table 6).This finding suggests that these adapted lentil materials can be used to develop new cultivars with higher Fe concentrations,an approach that has been proposed as a means of increasing Fe uptake in human diets[7,13].However,genetic effects for Zn concentration,another mineral of global dietary importance,were minor in lentils.This finding implies that limited genetic variation for Zn concentration is present in the lentil entries,which are all medium or large lentils with yellow cotyledons.More genetic variation for this trait may be detected by examining lentil genotypes representative of other market classes,such as lentils that have orange or red cotyledons.No significant genotype effects on Se concentration were observed in either chickpeas or lentils.Significant genotype effects were observed for Se concentration in both chickpea[11]and lentils[11,12]grown in Saskatchewan.However,in each of these previous studies location and year effects were larger than genotype effects.Differences in Se concentrations among lentils grown in different environments are proposed to be due in part to differences in soil Se concentrations[22].

A consistent finding among this study and others is the detection of significant location and location×year effects on seed mineral concentrations.This finding suggests that better understanding of differences between locations in environmentalconditions and historical management practices will promote the production of more nutritious pulse crops.Several variables have been observed to affect seed mineral concentration of pulse crops,including soil moisture[11,23],harvest timing[24],intercropping with small grain crops[25],and colonization by arbuscular mycorrhizal fungi[26].Knowledge of how seed mineral concentrations are affected by soil mineral availability,the microbiological profile of the rhizosphere,and management practices can complement breeding efforts to develop pulse crops with elevated concentrations of dietary minerals.

Table 8–Meanaseed concentrations of manganese(Mn),nickel(Ni),phosphorus(P),sulfur(S),selenium(Se),and zinc(Zn)for lentil breeding lines and cultivars grown at two locations in Washington in both 2009 and 2010.

Table 9 –Pairwise correlations between seed mineral concentrationsa,yield(kg ha?1),100-seed weight(g;HSW),and days to 50%flowering(DF)for 18 ARS chickpea breeding lines and four check cultivars(Dwelley,Sierra,Dylan,and Sawyer)across two years(2010 and 2011)with three locations per year(Pullman,WA;Genesee,ID;and Kendrick,ID).

Mean mineral concentrations presented in Tables 4 and 5 can be used to compare chickpeas grown in Washington and Idaho with chickpeas grown in North Dakota[27]and Saskatchewan[11].Fe concentration in this study was at the low end of means reported for North Dakota(60 μg g?1)and Saskatchewan(52.1 μg g?1).Zn concentration was between values reported for North Dakota(53 μg g?1)and Saskatchewan(25.2 μg g?1).Ca concentration was between values reported for North Dakota(1320 μg g?1)and Saskatchewan(472 μg g?1).Mg concentration was similar to those in North Dakota(1440 μg g?1)and Saskatchewan (1699 μg g?1).Mn concentration was higher than that of chickpea grown in Saskatchewan(23.90 μg g?1).Cu concentration was similar to those in North Dakota(8.1 μg g?1)and Saskatchewan(7.3 μg g-?1).Se concentration was similar to those in North Dakota(0.33 μg g?1)and less than Saskatchewan(0.73 μg g?1).

Similarly,mineral concentrations of lentils grown in Washington(Tables 7–8)can be compared to those of lentilsgrown in Saskatchewan [11].Concentrationsofseveral minerals,including Fe,Zn,Mg,K,Cu,and Mn were similar for both locations,not differing by＞20%.However,Ca concentration was 75%higher in Washington lentils than in Saskatchewan lentils(322 μg g?1).In contrast,Se concentration in Saskatchewan lentils(1.18 μg g?1)was nearly 175%higher than that in Washington lentils.

Table 10 –Pairwise correlations between seed mineral concentrationsa,yield(kg ha?1),100-seed weight(g;HSW),and days to 50%flowering(DF)for 13 ARS lentil breeding lines and three check cultivars(Avondale,Merrit,and CDC Richlea)across two years(2010 and 2011)with two locations per year(Fairfield,WA and Pullman,WA).

Table 11–Mineral concentrations(mean±standard error)for all entries combined for 2010 and 2011 at Pullman,WAa.

Mineral concentrations observed for chickpeas(Tables 4–5)and lentils(Tables 7–8)can be compared to concentrations recently reported for dry beans based on field evaluations of the Middle American Diversity Panel,which consists of 277 genotypes of races Durango and Mesoamerica[28].Seed concentrations in dry beans of B(11.98 μg g?1)and K(14,690 μg g?1)were＞40%higher than mean concentrations observed for chickpeas and lentils.The concentration of Ca in dry beans(1780 μg g?1)was＞70%and 200%higher than concentrations observed for chickpeas and lentils,respectively.Fe concentration in dry beans(61.40 μg g?1)was between values observed for chickpeas and lentils.Ni concentration in dry beans(1.91 μg g?1)was ＜50%of concentrations observed for chickpeas and lentils.S concentration in dry beans(2240 μg g?1)was similar to concentrations observed for chickpeas and lentils.Zn concentration in dry beans(30.87 μg g?1)was similar to that in chickpea but only 60%of that in lentil.Heritability of seed mineral concentrations were high(H2＞0.60)for the Middle American Diversity Panel,suggesting that an examination of a more diverse collection of chickpea and lentil materials may detect more genetic variation for these traits.

Correlation analysis of seed mineral concentrations and selected field traits were conducted for chickpeas(Table 7)and lentils(Table 8).Correlations between P and K,P and Zn,and K and Zn were among the five highest positive correlations between minerals for both chickpeas and lentils.These results are consistent with those of other studies that detected significant positive correlations among these three minerals in lentils grown in Saskatchewan[29]and Turkey[15].A significant positive correlation(r=0.55)between seed concentrations of K and Zn was also recently reported for dry beans[28].We detected significant positive correlations for chickpeas between Ca and Fe,and Ca and Mg,but a significant negative correlation between Ca and Zn.These results are consistent with correlations detected among these minerals for chickpeas grown in Saskatchewan[10].

The direction and magnitude of correlations between mineral concentration and yield,and mineral concentration and DF was similar for chickpeas and lentils across the majority of minerals,including Ca,Fe,K,Mn,Ni,P,and Zn.These results are likely due to the similarities of the two crops in their physiological processes for mineral uptake and partitioning as well as of the effects of location and genotype×location interactions on their seed mineral concentrations.Concentrations of three minerals,K,P,and Zn,had high negative correlations with yield and high positive correlations with DF.A significant negative correlation between yield and K concentration has been observed for dry beans[28].Negative correlations between yield and seed concentrations of K,P,and Zn have also been reported for soybeans(Glycine max(L.)Merr.)[30]and peas[31].These negative correlations suggest a dilution effect,whereby the whole-plant uptake or internal pool of certain minerals may be limited and unable to meet the genetic potential of a higher seed mineral load.Breeding for increased uptake of certain minerals may be needed to sustain higher mineral concentrations in higher-yielding lines[32].

5.Conclusion

In this study,minimal genotype effects on seed mineral concentrations were detected in chickpeas.Examination of a more diverse panel of genotypes is needed to identify useful genetic variation for seed mineral concentrations in chickpeas grown using locations and management practices typical of the U.S.Pacific Northwest.The lentils examined in this study hold promise for developing new cultivars with increased concentrations of Fe.Correlations between seed mineral concentrations and selected agronomic traits suggest that chickpeas and lentils share similar physiological processes for uptake and partitioning of several minerals.Marked negative correlations between yield and Zn concentration in both chickpea and lentil indicate that deleterious linkages between genes conditioning the two traits must be broken to facilitate development of cultivars that both are high-yielding and have elevated seed concentrations of Zn.Plant genotypes that are more efficient at obtaining minerals from growing environments will be useful as parental materials to develop improved chickpea and lentil cultivars that have high yield potential coupled with high seed mineral concentrations.

Acknowledgments

This work was conducted with the support of the United States Department of Agriculture-Agricultural Research Service(2090-21000-029-00D).The contents of this publication do not necessarily reflect the views or policies of the U.S.Department of Agriculture,nor does mention of trade names,commercial products,or organizations imply endorsement by the US Government.

[1]R.L.Bailey,K.P.West Jr,R.E.Black,The epidemiology of global micronutrient deficiencies,Ann.Nutr.Metab.66(Suppl.2)(2015)22–33.

[2]B.de Benoist,E.McLean,I.Egll,M.Cogswell,Worldwide prevalence of anaemia 1993–2005,WHO Global Database on Anaemia,WHO Press,Geneva,Switzerland,2008.

[3]J.C.King,Zinc:an essential but elusive nutrient,Am.J.Clin.Nutr.94(2011)679S–684S.

[4]K.R.Wessells,K.H.Brown,Estimating the global prevalence of zinc deficiency:results based on zinc availability in national food supplies and the prevalence of stunting,PLoS One 7(2012),e50568..

[5]G.F.Combs,Selenium in global food systems,Br.J.Nutr.85(2001)517–547.

[6]Q.Wu,M.P.Rayman,H.L.Schomburg,B.Cui,C.Gao,P.Chen,G.Zhuang,Z.Zhang,X.Peng,Low population selenium status is associated with increased prevalence of thyroid disease,J.Clin.Endocrinol.Metab.100(2015)4037–4047.

[7]D.D.Miller,R.M.Welch,Food system strategies for preventing micronutrient malnutrition,Food Policy 42(2013)115–128.

[8]R.Yip,U.Ramakrishnan,Experiences and challenges in developing countries,J.Nutr.132(2002)827S–830S.

[9]P.J.White,M.R.Broadley,Biofortifying crops with essential mineral elements,Trends Plant Sci.10(2005)586–593.

[10]R.A.Bueckert,D.Thavarajah,J.Pritchard,Phytic acid and mineral micronutrients in field-grown chickpea(Cicer arietinum L.)cultivars from western Canada,Eur.Food Res.Technol.233(2011)203–212.

[11]H.Ray,K.Bett,B.Tar'an,A.Vandenberg,D.Thavarajah,T.Warkentin,Mineral micronutrient content of cultivars of field pea,chickpea,common bean,and lentil grown in Saskatchewan,Canada,Crop Sci.54(2014)1698–1708.

[12]D.Thavarajah,J.Ruszkowski,A.Vandenberg,High potential for selenium biofortification of lentils(Lens culinaris L.),J.Agric.Food Chem.56(2008)10747–10753.

[13]D.Thavarajah,P.Thavarajah,A.Sarker,A.Vandenberg,Lentils(Lens culinaris Medikus subspecies culinaris):a whole food for increased iron and zinc intake,J.Agric.Food Chem.57(2009)5413–5419.

[14]D.Thavarajah,A.Wejesuriya,M.Rutzke,R.P.Glahn,G.F.Combs,A.Vandenberg,The potential of lentil(Lens culinaris L.)as a whole food for increased selenium,iron,and zinc intake:preliminary results from a 3 year study,Euphytica 180(2011)123–128.

[15]T.Karak?y,H.Erdem,F.S.Baloch,F.Toklu,S.Eker,B.Kilian,H.?zkan,Diversity of macro-and micronutrients in the seeds of lentil landraces,Sci.World J.2012(2012)710412.

[16]FAOSTAT,Food and Agriculture Organization,http://www.fao.org/faostat/en/#data/QC(Verified 9/25/17).

[17]National Agriculture Statistics Service(NASS),United States Department of Agriculture,https://www.nass.usda.gov/Statistics_by_Subject/index.php?sector=CROPS(Verified 9/25/17).

[18]C.Toker,A note on the evolution of kabuli chickpeas as shown by induced mutations in Cicer reticulatum Ladizinsky,Genet.Resour.Crop.Evol.56(2009)7–12.

[19]G.J.Vandemark,M.A.Brick,J.M.Osorno,J.D.Kelly,C.A.Urrea,Edible grain legumes,in:S.Smith,B.Diers,J.Specht,B.Carver(Eds.),Yield Gains Major U.S.Field Crops,American Society of Agronomy,Inc.,Crop Science Society of America,Inc.,and Soil Science Society of America,Inc.,Madison,WI,USA 2014,pp.87–124.

[20]M.W.Farnham,A.P.Keinath,M.A.Grusak,Mineral concentration of broccoli florets in relation to year of cultivar release,Crop Sci.51(2011)2721–2727.

[21]R.M.Walpole,R.Myers,Probability and Statistics for Engineers and Scientists,McMillan,New York,USA,1978.

[22]D.Thavarajah,A.Sarker,M.Materne,G.Vandemark,R.Shrestha,O.Idrissi,O.Hacikamiloglu,B.Bucak,A.Vandenberg,A global survey of effects of genotype and environment on selenium concentration in lentils(Lens culinaris L.):implications for nutritional fortification strategies,Food Chem.125(2011)72–76.

[23]A.Gunes,E.G.Bacgi,T.Erol,D.J.Pilbeam,A.Inal,M.S.Adak,M.Alpasian,Mineral nutrition of wheat,chickpea,and lentil as affected by mixed cropping and soil moisture,Nutr.Cycl.Agroecosyst.78(2007)83–96.

[24]M.S.Adak,N.Kayan,A.Gunes,A.Inal,M.Alpaslan,N.Cicek,T.Guzelordu,Effect of harvest timing on yield and mineral nutritional value of kabuli type chickpea seeds,J.Plant Nutr.30(2007)1397–1407.

[25]Y.Zuo,F.Zhang,Iron and zinc biofortification strategies in dicot plants by intercropping with gramineous species.A review,Agron.Sustain.Dev.29(2009)63–71.

[26]E.Pellegrino,S.Bedini,Enhancing ecosystem services in sustainable agriculture:biofertilization and biofortification of chickpea(Cicer arietinum L.)by arbuscular mycorrhizal fungi,Soil Biol.Biochem.68(2014)429–439.

[27]D.Thavarajah,Evaluation of chickpea(Cicer arietinum L.)micronutrient composition:biofortification opportunities to combat global micronutrient malnutrition,Food Res.Int.49(2012)99–104.

[28]P.E.Mc Clean,S.M.Moghaddam,A.F.López-Millán,M.A.Brick,J.D.Kelly,P.N.Miklas,J.Osorno,T.G.Porch,C.A.Urrea,A.Soltani,M.A.Grusak,Phenotypic diversity for seed mineral concentration in North American dry bean germplasm of Middle American Ancestry,Crop Sci.57(2017)3129–3144.

[29]N.Wang,J.K.Daun,Determination of cooking times of pulses using an automated Mattson cooker apparatus,J.Sci.Food Agric.85(2005)1631–1635.

[30]L.R.Gibson,R.E.Mullen,Mineral concentrations in soybean seed produced under high day and night temperature,Can.J.Plant Sci.81(2001)595–600.

[31]Y.Ma,C.J.Coyne,M.A.Grusak,M.Mazourek,P.Cheng,D.Main,R.J.McGee,Genome-wide SNP identification,linkage map construction and QTL mapping for seed mineral concentrations and contents in pea(Pisum sativum L.),BMC Plant Biol.17(2017)43.

[32]R.P.Sankaran,M.A.Grusak,Whole shoot mineral partitioning and accumulation in pea(Pisum sativum),Front.Plant Sci.5(2014)149.