MALTING CONDITIONS FOR EVALUATION OF RYE CULTIVARS

MALTING CONDITIONS FOR EVALUATION OF RYE CULTIVARS A Thesis Submitted to the Grdute Fculty of the North Dkot Stte University of Agriculture nd Applied Science By Yujun Wng In Prtil Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE Mjor Deprtment Cerel Science Februry 2017 Frgo, North Dkot

North Dkot Stte University Grdute School Title MALTING CONDITIONS FOR EVALUATION OF RYE CULTIVARS By Yujun Wng The Supervisory Committee certifies tht this disquisition complies with North Dkot Stte University s regultions nd meets the ccepted stndrds for the degree of MASTER OF SCIENCE SUPERVISORY COMMITTEE: Dr. Pul Schwrz Chir Dr. Seny Simsek Dr. Jochum Wiersm Approved: My 10, 2017 Dte Dr. Richrd Horsley Deprtment Chir

ABSTRACT Mlting of rye nd use of rye mlts presents severl chllenges to mltsters nd brewers, like the lck of hull nd dense pcking in steep. While, empiricl evidence shows tht rye genotypes differ in mlting nd brewing performnce nd flvor, there is little published informtion on the mlting of rye or the mlt qulity ttributes of rye genotypes. The objective ws to evlute lbortory micro-mlting conditions tht could be used in qulity screening. Prmeters included germintion time, moisture nd kernel size. Wort rbinoxyln nd phenolic cid content were determined in ddition to stndrd mlt qulity prmeters. In generl, high extrct nd lower viscosity were chieved by mlting for t lest 4 dys t 45-48% moisture. However, some commercil mltsters indicted the difficulty of hndling of germinting rye t high moisture levels. As such, 5 dys of germintion t 45% moisture is recommend for future evlution of rye cultivrs. iii

ACKNOWLEDGEMENTS The whole purpose of eduction is to turn mirrors into windows. -Sydney J. Hrris I would like to express the deepest pprecition to Dr. Pul B. Schwrz, who conveyed spirit of dventure in regrd to reserch. It would not hve been possible to complete this reserch without his continully guidnce, persistence, nd convincingly ptience. I would lso like to thnk my committee members: Dr. Seny Simsek, Deprtment of Cerel Science, North Dkot Stte University, for her suggestions throughout the reserch period; Dr, Jochum J. Wiersm, Deprtment of Agronomy nd Plnt Genetics, University of Minnesot, for his generous help in this reserch nd serving on my grdute committee. In ddition, thnk you to the following people for their help nd support: Dr. Richrd D. Horsley, for his vluble ssistnce nd dvice on the sttisticl nlysis; Dr. Mrk Sorrells, Plnt Breeding nd Genetics Section, Cornell University, for providing smples for this reserch; Mr. Steve Zwinger. NDSU Crrington Reserch Center for providing the smple of ND Dyln (DR02) rye; Ms. Andre Stnley t Vlley Mlt, for her vluble ides for this reserch. Mny thnks to Dr. Jin Zho, Jmes Gillespie, John Brr, Kristin Whitney, nd the professionl technicl stff nd grdute students of the Deprtment of Cerel Science, Plnt Sciences, for their friendship nd ssistnce in the performing of the nlysis for this reserch. Specil thnks to my fmily, for the understnding, precious love, nd morl support throughout my eduction. iv

TABLE OF CONTENTS ABSTRACT... iii ACKNOWLEDGEMENTS... iv LIST OF TABLES... viii LIST OF FIGURES... ix LIST OF APPENDIX TABLES... xi 1. INTRODUCTION... 1 2. LITERATURE REVIEW... 2 2.1. Botny nd Genetics of Rye... 2 2.2. World Production... 2 2.3. Uses... 3 2.3.1. Use in Bking... 3 2.3.2. Forge nd Feed... 3 2.3.3. Cover Crop... 3 2.3.4. Rye in Brewing nd Distilling... 4 2.4. Mlting of Rye... 4 2.5. Chemicl Composition of Rye... 6 2.5.1. Protein... 6 2.5.2. Strch... 7 2.5.3. Cell Wll Polyscchrides... 7 2.5.4. Bet-glucn... 7 2.5.5. Arbinoxylns... 7 2.5.6. Phenolic Acids... 9 2.6. Beer Flvor... 10 3. OBJECTIVE... 12 v

4. EXPERIMENTAL APPROACH... 13 4.1. Mterils... 13 4.2. Methods... 13 4.2.1. Rye Qulity Tests... 13 4.2.1.1. Test Weight... 13 4.2.1.2. 1000 Kernel Weight nd Kernel Assortment... 13 4.2.1.3. Grin Moisture... 14 4.2.1.4. Protein... 14 4.2.1.5. Germintive Cpcity... 14 4.2.1.6. Determintion of Deoxynivlenol (DON)... 14 4.2.1.7. Prehrvest Sprouting... 15 4.2.2. Mlt Anlysis... 16 4.2.2.1. Pilot Mlting... 16 4.2.2.2. Mlt Moisture... 17 4.2.2.3. Mlt Loss... 17 4.2.2.4. Fine Grind Mlt Extrct... 17 4.2.2.5. Mlt Disttic Power... 18 4.2.2.6. Alph-Amylse Activity... 18 4.2.3. Wort Anlysis... 18 4.2.3.1. Wort Soluble Protein... 18 4.2.3.2. Kolbch Index (Soluble /Totl Protein)... 19 4.2.3.3. Wort Viscosity... 19 4.2.3.4. Wort Crbohydrtes... 19 4.2.3.5. Free Amino Nitrogen (FAN)... 19 4.2.3.6. Arbinoxylns (AX)... 19 vi

4.2.3.7. Phenolic Acids... 20 4.2.3.8. Wort Bet-Glucn... 22 4.2.4. Experimentl Design nd Sttisticl Anlysis... 22 5. RESULTS AND DISCUSSION... 24 5.1. Sttisticl Interprettion... 25 5.2. Mlt Loss... 26 5.3. Mlt Extrct... 30 5.4. Mlt Enzymes... 33 5.5. Wort Soluble Protein Content nd Free Amino Nitrogen... 37 5.6. Wort Viscosity, Arbinoxyln, nd Bet-Glucn... 44 5.7. Wort Phenolic Acids... 50 5.8. Wort Crbohydrtes... 57 5.9. Stepwise Liner Regression... 63 5.10. Reltionships between Rye Mlt Qulity Prmeters... 64 6. SUMMARY AND CONCLUSIONS... 69 REFERENCES... 70 APPENDIX... 78 vii

LIST OF TABLES Tble Pge 1. Men of Rye Mlt Qulity Vlues s Affected by Mlting Opertionl Prmeters... 28 2. Anlysis of Vrince for Mlt Loss nd Extrct... 29 3. Anlysis of Vrince for DP nd α-amylse... 39 4. Anlysis of Vrince for Wort Soluble Protein nd Kolbch Index (S/T)... 43 5. Anlysis of Vrince for FAN nd Wort Viscosity... 45 6. Anlysis of Vrince for AX nd β-glucn... 49 7. Men of Phenolic Acids (mg/l) s Affected by Mlting Opertionl Prmeters... 53 8. Anlysis of Vrince for Phenolic Acids: Cffeic Acid nd Ferulic Acid... 54 9. Anlysis of Vrince for Phenolic Acids: p-coumric Acid nd Vnilic Acid... 55 10. Anlysis of Vrince for Totl Phenolic Acids... 56 11. Men of Wort Fermentble Sugrs (g/100ml) s Affected by Mlting Opertionl Prmeters... 60 12. Anlysis of Vrince for Fermentble Sugrs: Fructose nd Glucose... 61 13. Anlysis of Vrince for Fermentble Sugrs: Mltose nd Mltotriose... 62 14. Anlysis of Vrince for Totl Fermentble Sugrs... 63 15. Prtil (Prt.) nd Cumultive (Cum.) R 2 Vlues from Stepwise Regression Anlysis Across Opertionl Prmeters for Mlt Loss, Extrct, DP, α-amylse, Wort Soluble Protein, S/T, Wort Viscosity, FAN, AX, nd Wort β-glucn... 65 16. Prtil (Prt.) nd Cumultive (Cum.) R 2 Vlues from Stepwise Regression Anlysis Across Opertionl Prmeters for Phenolic Acids... 66 17. Prtil (Prt.) nd Cumultive (Cum.) R 2 Vlues from Stepwise Regression Anlysis Across Opertionl Prmeters for Fermentble Sugrs... 67 18. Correltion Coefficients of Rye Wort Chrcteristics (N=96)... 68 viii

LIST OF FIGURES Figure Pge 1. Chemicl Structures of the Min Phenolic Acids in Rye Wort... 9 2. Interction of Germintion Moisture Level nd Germintion Time on Mlt Loss.... 27 3. Interction of Smple nd Germintion Moisture Level on Extrct.... 31 4. Interction of Smple nd Germintion Time on Extrct.... 31 5. Interction on Germintion Moisture nd Germintion Time on Extrct.... 32 6. Interction of Size nd Germintion Time on Extrct.... 32 7. Interction of Size nd Germintion Moisture for Disttic Power (DP).... 34 8. Interction of Smple nd Germintion Moisture for Disttic Power (DP).... 35 9. Interction of Smple nd Germintion Time for Disttic Power (DP).... 35 10. Interction of Grin Size nd Germintion Moisture for -Amylse.... 36 11. Interction of Smple nd Germintion Moisture for -Amylse.... 37 12. Interction of Smple nd Germintion Moisture for Soluble protein.... 40 13. Interction of Grin Size nd Germintion Moisture for Soluble Protein.... 40 14. Interction of Smple nd Germintion Time for Soluble Protein.... 41 15. Interction of Grin Size nd Germintion Time for Soluble Protein.... 41 16. Interction of Smple nd Germintion Moisture for Wort Free Amino Nitrogen (FAN)... 42 17. Interction of Smple nd Grin Size for Wort Viscosity.... 46 18. Interction of Smple nd Germintion Moisture for Wort Viscosity.... 47 19. Interction of Germintion Moisture nd Germintion Time for Wort Viscosity.... 47 20. Interction of Germintion Moisture nd Germintion Time for Wort b-glucn.... 48 21. Interction of Smple nd Germintion Moisture for Wort Ferulic Acid.... 51 22 Interction of Smple nd Germintion Moisture for Wort p-coumric Acid.... 51 ix

23. Interction of Smple nd Grin Size for Wort p-coumric Acid.... 52 24. Interction between Smple nd Germintion Moisture for Wort Totl Fermentble Sugr... 57 25. Impct of Interction between Smple nd Germintion Moisture for Wort Mltotriose... 58 26. Impct of Interction between Smple nd Germintion Moisture for Wort Mltose.... 58 27. Impct of Interction between Smple nd Germintion Moisture for Wort Glucose.... 59 28. Impct of Interction between Germintion Time nd Germintion Moisture for Wort Mltose.... 59 x

LIST OF APPENDIX TABLES Tble Pge A1. 2014 Rye Grin Anlysis... 78 xi

1. INTRODUCTION Rye (Secle cerele L.) is cool seson cerel grss like whet (Triticum estivum L.). It is high yielding crop tht cn grow even in no-till, less productive fields or in low fertility desert regions (Chmielewski et l. 2000). It hs very few insect problems, nd it is strong competitor with weeds (Helm nd Schmeiter, 1991). While rye is trditionlly used in the production of bred, livestock feed, nd spirits (Helm nd Schmeiter, 1991), rye mlt is gining incresed considertion for its bility to dd unique chrcteristics to beer. These include spicy flvor nd stinging mouth feel (Wolfe, 2014). In contrst to ryes gronomic dvntges, the chllenge of mking rye mlt nd rye mlt beer hs been noted by mltsters nd brewers. The viscous rye wort mkes filtrtion extremely slow. The high level of soluble rbinoxyln (AX) in rye cell wlls contributes to the viscosity of rye wort (Hubner et l. 2010). On the other hnd, due to the nked kernel, rye grin presents dense pcking in steep, extreme shrinkge during kilning, nd lso extends the lutering process. Reserch on rye mlts nd use for specilty beers is limited. It is therefore, dvntgeous to define nd qulify the effect of mlting methods on rye mlt qulity. This reserch will evlute process fctors on the rye mlting process. Prticulr emphsis will be plced on extrct, mlt loss, nd wort viscosity. The result of this reserch my give brewers nd mltsters, or other reserchers better insight into how to efficiently hndle rye grin in mlting nd brewing. 1

2. LITERATURE REVIEW 2.1. Botny nd Genetics of Rye Rye belongs to the grss fmily, Grminee, nd the genus Secle. Secle cerele L. is the most common cultivted species. There re both winter rye nd spring ryes. Generlly, winter crops re fll-sown nnul crops. Winter rye hs better winter hrdiness thn winter whet nd winter brley. Spring rye, however, cn be grown in plces where the climte is too severe for winter rye. However, it hs poorer yield nd end-use qulity thn winter rye cultivrs (Bushuk, 2001). Becuse rye is cross-pollinted crop, it is difficult to mintin the genetic purity of cultivrs. Hybrid rye vrieties, tht involve the crossing of inbred lines, were first developed in the 1980 s in Germny. Hybrid rye vrieties hve improved yield, test weight, nd intrinsic qulity when compred to conventionl types (Geiger, 1986; Scoles et l. 2001). 2.2. World Production The winter hrdiness of rye nd its bility to grow under low fertiliztion provide wy to confront globl food shortge. However, when compred with the worldwide production of whet nd other cerels, rye is only of minor importnce (Bushuk, 2001). Rye is historiclly indispensble in the frming nd eting hbits in Northern Europe nd severl former Soviet Union countries. Europe is the min production region in the world, with s much s 90% of totl production from 1993-2013. The top 5 producers in the world re the Russin federtion, Polnd, Germny, Belrus, nd Ukrine (FAOSTAT, 2016). In the USA, the re of rye cultivtion hs decresed in recent decdes (USDA-FAS, 1997), but yields hve incresed due to improvements in gronomic prctices such s the use of fertilizers, crop rottion, nd highyielding cultivrs. The increse in rye yield hs offset the decrese in cultivted re. 2

2.3. Uses 2.3.1. Use in Bking The history of rye bred likely goes bck to the beginning of rye cultivtion. For both historicl nd helth resons, rye is importnt in the diets of northern nd Estern Europe. The bking properties of rye hve been studied by multiple groups, nd were reviewed by Bushuk (2001). The composition of rye presents bkers with severl chllenges, tht include poor gluten strength nd high content of wter soluble rbinoxylns. The rbinoxylns mke rye dough sticky nd difficult to hndle. Rye is lso often infected with ergot. Ergot scerloti re produced by group of fungi in the genus Clviceps, cn contin lkloids tht cuse ergotism in humns nd other mmmls (Schrdl, 2006). The infected grin must be removed before milling or feeding to livestock. 2.3.2. Forge nd Feed Rye is often feed grin or is used for grzing in countries outside of Europe. Winter rye cn be used for extending the grzing seson for livestock frmers in some res. Rye provides forge in lte fll nd erly winter tht reduce the cost of stored feed (hy) (Oelke, Oplinger, Bhri, Durgn, Putnm, Doll, nd Kelling, 1990). 2.3.3. Cover Crop Cover crops re used to provide ground cover, protect ginst soil erosion over the winter time, nd to prevent compction of nnully tilled fields. It is common to use ryegrsses or winter rye s cover crops, s the rye root system cn grow into lte fll. Becuse rye is the most 3

winter-hrdy crop, once it estblished well it will provide ground cover protecting ginst soil erosion (Sullivn, 2002). 2.3.4. Rye in Brewing nd Distilling Rye whiskeys were historiclly populr in North Americ becuse rye ws once more widely cultivted, nd becuse of the chrcteristic spicy or fruity flvor in the finl product. Americ rye whiskeys require use of t lest 51 percent of rye grin s well s rye mlt whiskeys, which re produced from fermented msh of not less thn 51 percent of rye mlt. For historicl resons the term Cndin whiskey is often used synomously with rye whiskey, even though they my contin no rye grin t ll [U.S. Government Publishing Office (GPO), 2008]. Rye les hve thousnds of yers of history in Europe. Stik (2011) reported tht brley, ots, nd rye were ll used in brewing by the Celts in the erly Iron Age through lte Medievl times. However, rye hd lrgely fllen out of fvor s brewing grin, with the notble exception being limited production of roggenbier in Bvri, Germny. Recently, North Americn crft brewers hve sprked somewht of renissnce in rye beers. Drft Mgzine reported just s lmost every brewery produces one whet beer, one ple le nd one stout, now, there s seemingly rye beer on every beer mker s menu. (Stmbor, 2010). The use of rye grin or mlt is ssocited with spicy nd sour-like rye chrcteristics in the beer flvor profile. 2.4. Mlting of Rye Brley is the most common rw mteril in brewing, nd its properties in mlting hve been extensively studied. On the other hnd, there hs been limited reserch on the mlting of 4

rye. Bsed upon the lck of hull (husk) it would pper tht the mlting of rye might be similr to the mlting of whet. Briggs (1998), in fct, mentions in Mlts nd Mlting tht the mlting of rye follows the sme precutions s for whet. It pcks tightly nd is esily dmged in mlting. Older references cited by Briggs stte tht lthough rye steeps rpidly, it is slow to modify, nd requires up to 7 dys of germintion. Pomernz (et l. 1973) reported thn when compred to brley mlt, rye mlts hd higher levels of extrct, soluble protein, nd lph-mylse. Tylor (2000) nd Mule (1998) reported tht compred to brley mlt, rye mlts resulted in higher viscosity in beers. Hübner, et l. (2010) evluted optiml mlting conditions for rye in terms of viscosity nd other prmeters. Vribles included germintion nd steep tempertures, steep moisture nd germintion time. They found tht rbinoxylns (AX) ccumulted during the germintion process nd their extrctbility incresed. The results suggest tht longer germintion periods resulted in n incresed number of AX molecules with lower moleculr mss. High wort viscosity is cused by wter extrctble AX. They indicted the optiml rye mlt qulities within the limits of their study were found for germintion time of 6 dys t 10 C. These conditions resulted in n cceptble FAN levels, with the lowest mesured viscosity. They lso found tht with long germintion periods, mylolytic nd proteolytic enzyme ctivities were incresed, while β-glucnse ctivity wsn't influenced. Totl nd soluble nitrogen content were lso not significntly ffected by the vritions in germintion conditions. Free mino nitrogen (FAN) ws found in higher mounts in worts prepred from rye mlts with long germintion times. Extrct contents were higher in rye mlt thn in the control brley mlt nd could be incresed by fvorble germintion regime while no such impct on wort fermentbility ws found. 5

2.5. Chemicl Composition of Rye Composition nd chemistry hd been studied to understnd the properties of rye. Bsed on former studies rye grin hs been reported to contin 9%-15 % protein, 57-66% strch, 14 to 20% dietry fiber, nd 1.6-2.2% sh (Seibel, 2001; Nilsson, 1997, Bushuk, 2001; Hnsen et l., 2003, 2004). However, the composition of rye my vry between reports becuse of environmentl nd genotypic influences (Hnsen et l., 2004; Békés et l., 2016). 2.5.1. Protein Rye proteins hve been described in numerous studies. Bushuk (2001) divided these proteins into three groups including storge proteins which re locted in protein bodies, storge nitrogen, nd sulfur nd crbon skeletons. The storge proteins found in the endosperm re clled seclins. The 7S globulins re the min storge protein in the embryo nd leurone cells. Another protein group is hydrolytic enzymes nd their inhibitors, like α-mylses nd α-mylses inhibitors, β-mylse, proteinses, esterses, β-glucosidse, nd β-glctosidse. A finl group of proteins re the resistnce-relted proteins, which re involved in defense ginst pests nd pthogens. Although rye hs close reltionship with whet, it hs low gluten content. Rye prolmins re referred to s seclin, which is the mjor storge protein of rye grin. Unlike glidins nd glutenins which cn form gluten in whet, it is difficult to prepre cohesive mss from rye doughs. Seclins lck the cohesiveness of gluten (Gellrich 2003, 2004). 6

2.5.2. Strch Strch is the min source of energy in rye. Totl content nd structure is similr to other cerels like whet. It consists of mylose nd mylopectin, where mylose represents 24-26% of the totl mounts (Hew nd Unru, 1970; Berry et l, 1971). The viscosity of rye strch is similr to bred whet nd durum whet. However, sprouting cn be n issue in rye, nd lower pste viscosities re seen becuse of the presence of α-mylse (Berry, et l., 1971, Klssen nd Hill, 1971). 2.5.3. Cell Wll Polyscchrides The non-strch polyscchrides found in cell wlls re dietry fiber. The whole grin content of rye is bout 17% of dietry fiber, including 3-4% soluble fiber. Arbnoxylns re mjor component, nd were formerly referred to s pentosns (Bushuk, 2001). 2.5.4. Bet-glucn Whole grin rye, contins 2.5% bet-glucn, compred to 1.7% in the endosperm. This indictes high levels of bet-glucn in the brn frction (Bushuk, 2001). Overll numbers re lower thn those for brley nd ots (Rgee et l. 2008). 2.5.5. Arbinoxylns Like bet-glucn, AX levels lso re higher in the brn thn in the endosperm (Rgee et l. 2008). Rye hs considerbly lrger mounts of AX thn brley, especilly wter extrctble rbinoxylns (WEAX), which re minly responsible for the incresed viscosities. Rye AX structure nd its properties hve been studied by Vinkx nd Delcour (1996), nd contribute 7

between 6.5-12.2% of the totl dry weight in rye kernels. AX hve bckbone of β -1 4 linked xylnopyrnose units. Alph rbinofurnose cn be linked to the C (O)-2 nd/or C (O)- 3 positions of the xylose units. Some of the rbinose substituents re esterified with ferulic cid t C (O)-5 (Bengtsson et l., 1992). AX represents lrge prt of the rye s content of dietry fiber nd especilly soluble dietry fiber, which is known to hve helth benefits for the consumer. From technologicl point of view, soluble dietry fiber, mong other substnces like proteins, contribute to body nd mouthfeel of the beer. Lu et l. (2000, 2004) described in two studies on the positive effects of AX on postprndil insulin response in helthy individuls nd improved metbolic control in dibetes ptients. AX brekdown products hve lso been shown to disply prebiotic properties (Cloetens et l.,2008; Courtinetl.,2008; Grootert et l.,2007; VnCreyveld et l.,2008). While there re some positive spects ttributed to AX, they cn cuse problems in beer production. Properties such s moleculr mss, the degree of rbinose brnching, nd the degree of esterifiction with ferulic cid will influence AX properties drmticlly (Li et l. 2005). WEAX increse the viscosity of solutions s they cn bind lrge mounts of wter. Gels cn be formed by oxidtive cross-linking of AX mcromolecules vi the ferulic cid residues. During germintion of rye kernels, AX is degrded by number of enzymes. The most importnt re the endo-xylnses, which form lrger number of AX molecules with shorter chin length by cleving internl linkges in the min xyln chin. β Xylosidses slightly decrese the chin length by relesing xylose residues from the non-reducing end. Arbinose molecules cn be relesed from the mcromolecules by α-l-rbino furnosidse, while the ctions of feruloyl esterse cn hydrolyse the linkge between ferulic cid nd rbinose residues (Grootert et l, 2007). 8

2.5.6. Phenolic Acids Phenolic compounds re usully studied in subgroups, like phenolic cids, flvonoids, isoflvonoids, lignns, stilbenes, nd polyphenols (Dewick, 2001). Rye phenolics were reviewed by Bondi-Pons et l. (2009), s their ntioxidnt ctivities offer potentil helth benefits. However, food processing or other biologicl effects cn mke the ntioxidnt ction become more complex (Frnkel nd Meyer, 2000). Ferulic cid (85%-90%), sinpic cid (9%-10%), nd pr-coumric (3%-5%) hve been reported to be the most bundnt phenolic cids in rye grin (Bondi-Pons, 2009). There is smll mount of syringic, cffeic, vnillic nd prhydroxybenzoic cids present in the rye kernel s well. In terms of milling frctions nd products, the levels of phenolic cids hve been reported to decrese s follows: rye brn> rye grin> rye flour> rye bred. Ktin et l. (2007) fermented both ntive nd germinted rye grins. They found tht, during fermenttion, yest fermenttion plys the mjor roll for 10- fold increse of free ferulic cid in both experimentl groups. However, there were no reports bout phenolic cid bioctivities in rye mlt or wort. Syringic Acid Cffeic Acid Ferulic Acid Sinpinic Acid p-coumric Acid Fig. 1. Chemicl Structures of the Min Phenolic Acids in Rye Wort Vnillic Acid 9

2.6. Beer Flvor Rye mlt contributes distinctive flvor, body nd mouthfeel to the beer. The unique spicy flvor of rye hs drwn ttention in the crft beer mrket. The extent of its contribution to the beer flvor profile is ffected by the mlted grin nd severl other fctors in the brewing process (Pomernz et l. 1973). Roggenbier, is the historicl beer tht ws originlly brewed in Regensburg, Germny [Beer Judge Certifiction Progrm (BJCP), 2015]. Pople often prtition whet nd rye together into Germn Whet nd Rye Beer ctegory s the close reltionship between rye nd whet beer BJCP (2008). BJCP (2008, 2015) introduced Roggenbier in the overll impression s dunkelweizen mde with rye rther thn whet, but with greter body nd light finishing hops. More thn 50% of mlted rye (sometimes up to 60-65%) cn be used during mshing. Ple mlt, Munich mlt, whet mlt, crystl mlt nd/or smll mounts of debittered drk mlts for color djustment were been used s the remining grin. Distinctive bnn esters nd clove phenols chrcter were provided by Weizen yest. Usge of Szer-type hops in bitterness, flvor nd rom is light. Other beverges tht use rye s ingredient re trditionl Scndinvin or Russin beers such s Kvss, Gotlndsdrick, nd Shti (BJCP, 2015). Unlike Roggenbier, Americn rye beer hs less yest flvor nd more hop chrcter (BJCP, 2015). The crft brewers nd homebrewers in the US mke sesonl Rye IPAs, Rye PAs, or RIPAs s their specilty IPA. These rye beers hve drier nd slightly spicier thn n Americn IPA, nd usully hve clen or slightly fruity profile. Americn or New World hops nd Americn or English yest were lso used by crft brewers nd homebrewers, with 15-20% rye mlt nd ple le or 2-row brewers mlt s the bse of the grin bills. 10

People usully chnge the rye beer flvor by djusting the percentge of rye mlt or exchnge the djunct ingredients. There is not much informtion bout the effect of genotype on rye beer flvor. It would not be big problem for homebrewed beer, but it is considerble problem for microbrewers, s the concern of consistency of the finl product between btches. Thus, the flvor profile should be involved in the future studies. 11

3. OBJECTIVE The objective of this reserch ws to evlute lbortory micro-mlting conditions tht could be used in the qulity screening of rye genotypes. Idel conditions should chieve high extrct with miniml mlt loss, nd lower wort viscosity/rbinoxyln content. 12

4. EXPERIMENTAL APPROACH 4.1. Mterils Upon initition of the rye mlt reserch project, smples were obtined from Cornell University, the University of Minnesot-Crookston, the NDSU Crrington reserch extension center, the Oklhom Seed Foundtion (Ardmore, OK), nd severl commercil or frm sources. All smples (n=48) were screened for grin plumpness, germintive cpcity, nd number of other grin qulity prmeters (ppendix tbles A1-A5). 4.2. Methods 4.2.1. Rye Qulity Tests 4.2.1.1. Test Weight Test weight (kg/hl) ws determined on clened smples, fter removl of the dockge [U.S. Deprtment of Agriculture (USDA),2013]. 4.2.1.2. 1000 Kernel Weight nd Kernel Assortment One thousnd kernel weight ws mesured on clened smples, by determining the number of kernels in 10.0g smple. Kernel ssortment ws performed ccording to the Americn Society of Brewing Chemists (ASBC) Method Brley-2C (ASBC, 2009). Kernel size distribution ws determined with 2.8, 2.4, nd 2.0 x 19mm (7 /64, 6 /64, nd 5 /64 x 3 /64 ) sieves on Eurek-Nigr Brley Grder (Silver Creek, NY). Rye grin (100g) ws spred on the top screen nd shken for 2 min. Kernels remining on the 2.8 nd 2.4 mm sieves were considered plump kernels nd kernels retined on 2.0 mm sieve were considered medium kernels. Kernels pssing through the 2.0 mm sieve were considered thin. In this experiment, 13

plump nd medium kernel frctions were sved for further nlysis, while thin kernels were removed. 4.2.1.3. Grin Moisture Grin smples were ground using Perten LM 3600 disc mill (Perten Instruments. Hägersten, Sweden). Grin moisture ws determined with Brbender Moisture Tester (C. W Brbender Corp. Rochelle Prk, NJ) by heting ground smple for 1 hr t 130 o C. 4.2.1.4. Protein Rye grin protein content ws determined using FOSS 1241 NIR (FOSS in North Americ, Eden Pririe, MN) using the clibrtion supplied with the instrument. The ccurcy of results ws crossed-checked, by determining nitrogen on LECO FP 528 nitrogen nlyzer (LECO Corportion, St. Joseph, MI) with the nitrogen fctor of 6.5 (LECO FP 528 Appliction Note, 2017). NIR nlysis ws on whole grin, while smples for combustion nlysis were ground to pss 0.5 mm screen on UDY mill (UDY Corp., Boulder, CO). 4.2.1.5. Germintive Cpcity Germintive cpcity ws determined by ASBC Brley Method-3B (2009). The percent chitted kernels fter 72 hours were recorded s the germintive cpcity. 4.2.1.6. Determintion of Deoxynivlenol (DON) DON ws determined bsed upon the method introduced by Tcke nd Csper (1996). Smples were ground using Perten lbortory mill (model 3600, Perten Instruments. Hägersten, 14

Sweden), nd weighed (2.5g) into 50 ml polypropylene tubes. Extrction ws with20 ml of 84% cetonitrile/wter solution for 60 minutes on horizontl shker. After settling, 2 ml liquot of the superntnt ws trnsferred to column contining 1g of 50/50% C18/lumin. The superntnt (2 ml) ws grvity filtered, trnsferred to 5 ml disposble borosilicte glss culture tube (47729-570, VWR CO), nd dried under nitrogen gs. The dried smple ws derivtized using trimethlysilymidzole (TMSI), trimethylchlorosilne (TMCS) 10:1. The derivtized smples were nlyzed by gs chromtogrphy with electron cpture detection (GC ECD) (model Agilent 6890 GC ECD, Snt Clr, CA). The smples (1 µl) were injected in duplicte onto 5% phenyl methyl siloxne column (30 m 0.25 mm 2 µm) (Agilent HP-5). A polrity dectivted column (1-2 m 0.53 mm) (Restek. Bellefonte, PA) ws ttched s gurd. The system prmeters were s follows: the crrier gs ws helium; flow pressure: 1.38 br; the initil inlet temperture ws 90 C, nd ws then rmped t rte of 20 C/min to 300 C; initil oven temperture 70 C nd incresed to 170 C t rte of 25 C/min, nd then t 5 C/min to 300 C. Detector temperture ws held t 300 C with ArCH 4 t 60mL/min. Mirex (ULTRA Scientific, Kingstown, RI) ws the internl stndrd (0.5 mg/ml). A stndrd curve from 0.1 to 40 ng/µl ws prepred by spiking the stndrd (Biopure, Romer Lb Inc., Union, MO) into DON free brley extrction. DON results re shown in ppendix tble A5. 4.2.1.7. Prehrvest Sprouting Prehrvest sprouting ws determined using the stirring number test on Rpid Visco- Anlyzer (Newport Scientific Pty Ltd, Werriewood, New South Wles, Austrli) ccording to AACC method 22-08.01(2000). Rye ws ground to pss 0.5 mm screen in UDY mill (UDY Corp., Boulder, CO) nd 4.0 g of flour were mixed with 25g distilled wter. The stirring number 15

test ws conducted under the following conditions: 95 o C for 3 minutes stirring t 160 rpm, with the initil high speed for 10 seconds. Stirring number (SN, viscosity cp t 3 minute) iss reported in the ppendix tbles A3 nd A4. 4.2.2. Mlt Anlysis 4.2.2.1. Pilot Mlting The steeping time ech rye smple required to rech 40%, 45%, nd 48% moisture ws determined by pilot-steeping 10 g (dry bsis) smple ccording to the procedures of Bnsik et l. (1955). Smples were steeped t 16 o C in 50 ml perforted round-bottom centrifuge tubes (3122-0050 (Nlgene, Rochester, NY). The smples were removed from steeping fter 24, 48, 72 hr nd centrifuged t 1500 rpm for 2 min to remove surfce moisture, nd then weighed. At ech time intervl, the moisture of steeped smple ws clculted. The time required to rech the moisture level (40%, 45%, nd 48%) ws clculted by plotting log moisture ginst log time. Micro-mlting ws crried out ccording to method described by Krbb et l (1993). Smples (80 g, dry bsis) were steeped for the time determined by pilot-steeping. Steeping includes ertion every four hours for six minutes, nd the wter is drined every 12 hr., nd the smples ir-rested for 1 hr. After steeping, the smples were removed from the steep tnk, spred over pper towels to eliminte surfce wter on the grin. Smples were then weighed, nd djusted to desired weight using distilled wter. Weights for 40, 45, nd 48% moisture levels were 133, 145, nd 154 g, respectively. Smples were plced into 400 ml bekers, germinted for either 3, 4, 5, or 6 dys in the germintion cbinet t 16 o C, nd pproximtely 95% reltive humidity. Smples weight ws 16

djusted every dy by dding distilled wter, nd smples were hd hnd-turned to preventing mtting. After the completion of germintion, ll smples were moved into kiln continers. Kilning ws 24-hr schedule, where the temperture ws sequentilly rmped from 49 to 85 o C. After kilning, the smples were removed from the kiln, cooled to room temperture nd derooted by brding ginst ech other when rubbed by hnds. Clened smples were weighed nd stored t room temperture prior to nlysis. 4.2.2.2. Mlt Moisture Mlt moisture ws determined on corsely ground 10 g smple with semi-utomtic Brbender Moisture Tester (Brbender Corp., Rochell Prk, NJ), heted ccording to ASBC Mlt-3 (2009). 4.2.2.3. Mlt Loss Mlt loss represents the loss of solubles nd CO 2 during germintion, nd the removl of rootlets following kilning. Mlt loss ws clculted s: % Mlting loss = "#$%&' )* +.-../%&#0 1#%* )* +.-. 344% "#$%&' )* +.-. d.b.: Dry bsis wt: Weight 4.2.2.4. Fine Grind Mlt Extrct Fine grind mlt extrct ws determined ccording to modifiction of ASBC Mlt Method (2009), Mlt-4. The mjor modifiction is tht rye smples were centrifuged t 3000 x g 17

prior to filtrtion (20 o C, 15 minutes). 4.2.2.5. Mlt Disttic Power Mlt disttic power ws determined s described in Technicon Industril Method No. 424-76A (Brn nd Luebbe, Inc. Trrytown, NY). Three mlt flour smples of known disttic power were nlyzed with ech set of smples. The stndrd smples were used to prepre plot of Technicon Autonlyzer pek height vs. disttic power. The disttic power of the stndrds mlts were determined through collbortive testing conducted by the Americn Society of Brewing Chemists. 4.2.2.6. Alph-Amylse Activity The lph-mylse ctivity of mlt ws determined ccording to modifiction of the procedure of Bnsik (1971) on Technicon Autonlyzer, s described in Technicon Industril Method NO. 424-76A (Brn nd Luebbe, Inc. Trrytown, NY). Three mlt flour smples of known lph-mylse ctivity were nlyzed with ech set of smples. The stndrd smples were used to prepre plot of Technicon Autonlyzer pek height vs. lph-mylse ctivity. The lph-mylse ctivity of the stndrds ws determined ccording to ASBC Method (2009), Mlt-7. 4.2.3. Wort Anlysis 4.2.3.1. Wort Soluble Protein Wort soluble protein ws determined ccording to ASBC Method (2009), Wort-17. 18

4.2.3.2. Kolbch Index (Soluble /Totl Protein) Kolbch Index ws clculted s: Kolbch Index = 6789 :7;<=;>?879>@A 344% B8C@A D79C;?879>@A Wort soluble protein ws determined ccording to ASBC Method (2009), Wort-17. 4.2.3.3. Wort Viscosity Wort viscosity ws determined t 20 o C ccording to ASBC Method (2009), Wort -13A. 4.2.3.4. Wort Crbohydrtes Wort crbohydrtes were determined by high-performnce liquid chromtogrphy (HPLC) method using n Aminex HPX-87 column (Ctologh No. 125-0095, Bio-Rd Lbortories, Hercules, CA) ccording to ASBC Method (2009), Wort-14B. 4.2.3.5. Free Amino Nitrogen (FAN) FAN ws determined ccording to ASBC Method (2009), Wort-12. 4.2.3.6. Arbinoxylns (AX) The rbinoxyln content of wort ws determined by gs chromtogrphy ccording to the modified method of Crpit nd She (1989). A 100 µl liquot of ech rye wort smple ws derivtized to lditol cettes ccording to the method of Blkeney et l. (1983) with some modifictions. The frozen wort smples were thwed t 25 C nd 100 µl ws dded to screw cp tube (16 x 125mm). The smples were hydrolyzed using 4.17M trifluorocetic cid (1500 µl) 19

by heting t 121 C 1 hr. After hydrolysis, the internl stndrd (inositol) ws dded to ech smple nd the smples were dried t 55 C under nitrogen. The smples were then reduced by dding mmonium hydroxide (1M, 100µl) nd sodium borohydride in DMSO (20mg/ml, 500µl). After heting t 40 C for 90 minutes, 6 drops of glcil cetic cid were dded to ech tube. To cetylte the smples, 100 µl 1-methylimidzol nd 500 µl cetic nhydride were dded to ech tube nd the rection ws stopped fter 10 minutes with the ddition of 4ml of wter. The smples were then prtitioned with 1 ml of methylene chloride two times, nd the methylene chloride frctions were combined nd dried t 45 C under nitrogen. The smples were finlly redisolved in 1 ml of cetone nd plced in2 ml uto-smpler vils (Agilent Technologies) for GC nlysis. The smples were nlyzed with n Agilent 7890 gs chromtogrph (GC) with flme ioniztion detector (Agilent technologies, Snt Clr, CA). The smples (5 µl) were injected in duplicte onto Supelco SP-2380 fused silic cpillry column (30 m 0.25 mm 2 µm) (Supelco Bellefonte, PA, U.S.A.). The system prmeters were s follows: flow rte, 0.8 ml/min; flow pressure, 82,737 P; oven temperture, 100 C; detector temperture, 250 C; nd injector temperture, 230 C. The crrier gs ws helium (Mendis et l. 2013). A stndrd curve ws prepred tht contined monoscchride stndrds in the concentrtions of 250, 500, 750 nd 1000 ng/ul nd inositol ws dded t 750 ng/ul s n internl stndrd. Arbinoxyln content ws clculted s: EFGHIJKLMNGJ (PQ/S) = [ (VWXYZ[\]^_ \`/b + deb]^_ \`/b g. hh)) iggg] igg 4.2.3.7. Phenolic Acids The phenolic cid content of wort ws determined ccording to modifiction of method reported by McMurrough et l. (1984). Congress wort smples (5 ml) were djusted to ph of 20

2.0 by dding 2.0M HCL. Smples were then extrcted twice by shking vigorously for 60 minutes with hexne (5mL). The mixtures were centrifuged for 5 minutes t 3000 x g fter extrction, nd the orgnic phse ws discrded. The remining queous phses were extrcted three times with ethyl ether/cette (1:1, v: v) (5mL). After vigorously shking for 15 minutes, the mixtures were centrifuged (3000 x g) for 5 minutes. The pooled ethyl ether/cette extrcts were dried under nitrogen gs. Acetonitile (0.5 ml) ws dded nd the smples were filtered through Whtmn-40 filter (Whtmn, UK) into 2.0 ml mber (Agilent Technologies,). These concentrted smples were then nlyzed on n Agilient 1290 series liquid chromtogrphy with 6540 UHD Accurte-Mss Qudrupole Time-of-Flight (Q-TOF) LC/MS (Agilent Technologies, Snt Clr, CA). Seprtion ws performed on ZORBAX SB-C18 column (1.8 µm, 2.1 50 mm, Agilent, Snt Clr, CA, USA) t 30. The mobile phse consisted of wter contining 0.1% formic cid (solvent A) nd cetonitrile contining 0.1% formic cid (solvent B). Grdient conditions were s follows: 0-1 min isocrtic with 3% B; then liner increse from 3 to 97% B for 1-10 min; followed by n isocrtic wshout step for 5 min nd shifting bck to initil setting for 2 min. Flow rte ws 0.4 ml/min, nd injection volume ws 2.0 µl. Detection ws (diode rry detector) ws crried out by scnning the bsorption between 250 to 400 nm with step of 2.0 nm. The wvelengths were 260.0 nm, 275.0 nm, 294.0 nm nd 324.0nm with the bnd width of 2.0 nm. The AJS electrospry ioniztion interfce (ESI) interfce ws used in the positive mode, nd the bsorbnce threshold of the 21 entroid dt storge ws 200 (Rel. 0.01%). The stop time ws 15 min nd cycle time 0.5 s. Source prmeters were set s follows: drying gs t the temperture of 300 with the flow rte of 10 L/min; 30 psig nebulizer gs t 300 nd 7 L/min; nd 125 V frgmentor energy. The mss rnge (m/z) of TOF Spectr ws 100-1000, nd the cquisition rte nd time were 2 spectr/s nd 5 21

ms/spectrum, respectively. The m/z of reference msses were 121.0509 nd 922.0098. Ferulic cid, p-coumric cid, vnillic cid, sinpinic cid, cffeic cid, ctechin, syringic cid nd gllic cid were quntitted with their hydrogen dduct of m/z 195.0654, 165.0545, 169.0494, 225.0756, 181.0494, 291.0882, 199.0600 nd 171.0287, respectively. The clibrtion curves were prepred by spiking phenolic cids stndrds into the extrct, nd the response re ws clculted by deducting tht in control extrct. The limit of detection (LOD) nd quntifiction (LOQ) for Ferulic cid, p-coumric cid, Sinpinic cid, Cffeic cid nd Ctechin were 0.02 µg/100 ml (0.02 µg/ml wort) nd 0.1 µg/100 ml (0.1 µg/ml wort), respectively. For Vnillic cid, Syringic cid nd Gllic cid, LOD nd LOQ were 0.05 µg/100 ml (0.05 µg/ml wort) nd 0.2 µg/100 ml (0.2 µg/ml wort), respectively. 4.2.3.8. Wort Bet-Glucn Wort bet-glucn ws determined ccording to djusted ASBC Method (2009), Wort- 18B. Wort smples were plced in n utosmpler (Wters Corportion, Milford, MA). The flow rte ws 3.0 ml/min with Wters 515 HPLC Pump (Wters Corportion, Milford, MA). Wters 474 Scnning Fluorescence Detector (Wters Corportion, Milford, MA) set to 420 nm emission nd 365 nm excittion wvelengths. 4.2.4. Experimentl Design nd Sttisticl Anlysis This study ws designed ccording to rndomized complete block design (RCBD) with five fctors in fctoril rrngement. The fctors were 2 levels of smple, three levels of steep moisture (40%, 45% nd 48%), four levels of germintion dys (3, 4, 5, nd 6 dys), nd 2 levels of kernel size (plump nd medium). Mlting ws replicted (n=2). Dt ws nlyzed by 22

Anlysis of Vrince (ANOVA), performed with procedures of the Sttisticl Anlysis System (version 9.3, SAS Institute, Cry, NC), nd nlyzed using interctions. Min effects nd interctions were evluted using the generl liner models (GLM) procedure. The sources of vrition for germintion dys nd germintion moisture were portioned into single degree of freedom polynomil contrsts. Duncn s multiple rnge test ws used to compre tretment mens. Stepwise regression ws used to evlute how much vribility could be explined by ech independent vrible (e.g. smple, kernel size, germintion moisture, dys of germintion) for the dependent vrible (e.g. mlt loss, extrct, DP, α-mylse, soluble protein, FAN, wort viscosity, AX, β-glucns, fermentble sugrs, nd polyphenolics). 23

5. RESULTS AND DISCUSSION Upon the initition of this project, the first tsk ws to identify rye grin smples for study. As mentioned in the mterils section, smples were obtined from smll grins progrms t Cornell University, North Dkot Stte University, the University of Minnesot, nd the Oklhom Seed Foundtion. In ddition, severl smples were obtined from mltsters nd growers, in coopertion with the North Americn Crft Mltsters Guild. A totl of 48 smples were obtined the initil qulity screening nd results re shown in ppendix tbles A1-A5. Cultivrs included forge types, nd both conventionl nd hybrid grin types. The smples exhibited wide rnge in kernel plumpness (1.2-90.0%, men= 48.2), 1000 kernel weight (15.2-37.7 g men = 26.9g), protein (7.6-19.6%, men = 11.6%), nd germintion (72.5-100%, men=93%). There ws lso rnge in determentl chrcters including deoxynivlenol (0.0-3.0 mg/kg, men= 0.40 mg/kg), nd prehrvest sprouting (stirring number 19.4-172.2 sec, men= 118.5 sec). Exmintion of ppendix tbles A1-A5 shows tht the forge types generlly hd higher protein (men=15.8), lower plumpness (men= 12.5%), nd low 1000 kernel weight (men=20.8 g). As such, grin form the forge types ws considered uncceptble for the production conventionl rye mlt. The current study lso required severl kilogrms of smple when tretments nd replictions were considered. Two rye smples were selected for further study bsed upon germintive cpcity, kernel plumpness, nd dequte smple mount. These were the genotype DR02 (ND Dyln) nd n unidentified genotype of winter rye. DR02 ws grown t the NDSU Crrington reserch extension center in 2014, nd hd cceptble plumpness (63.1%), protein (12.7%), nd germintion (96.5%). The unidentified smple ws from the 2014 crop in Iow, nd ws obtined from Embden Grin in Embden, North Dkot. The plumpness (42.1%) nd protein (10.9%) were cceptble, but germintion ws slightly less thn optiml 24

(91%). As the design of this study cnnot be used to determine vriety effects, DR02 nd the Iow smple will subsequently be referred to s smples A nd B, respectively. 5.1. Sttisticl Interprettion The sttisticl significnce of the prmeters ws determined by ANOVA. Interctions re discussed prior to min effects, s sttisticlly significnt interction indictes the results for certin nlyticl prmeter my not hve responded uniformly cross combintion of fctors (e.g. smple, germintion moisture level, etc.). However, while sttisticlly significnt, some interctions my not be true interction, s the rnk of one fctor my hve remined the sme cross ll levels of tretments. In these cses, the significnce is usully cused by differences in the mgnitude of responses to different tretments. For exmple, while the interction of germintion time germintion for mlt loss, ws sttisticlly significnt, exmintion of Figure 2 shows tht the trends were similr for ech moisture level nd rnks did not chnge. However, the mgnitude of responses cross the germintion moisture times ws different. Not ll interctions re included in the subsequent discussion. Only 2-wy interctions re shown in figures 1-28. The fctor in the ANOVA tble hd been prticipte into seprte single degree of freedom comprisons. The number of comprisons is equl to the number of levels of fctor minus one. For exmple, in tble 2, in the mlt loss nlysis, there were 3 levels of the fctor of germintion moisture, then there were two comprisons, liner nd qudrtic. This nlysis shown the significnce of the liner, qudrtic, nd/or cubic effects for these tretments. 25

5.2. Mlt Loss Mlt loss reflects the mount of mteril lost in converting grin into mlt, nd is n importnt economic considertion to the mltster. When considered on dry bsis, mlt loss cn be ttributed to respirtion losses, loss of soluble mteril in the steep, nd the removl of rootlets. In the current study, mlt loss ws significntly ffected (P 0.05) by ll mlting opertionl prmeters except kernel size (Tble 1, Tble 2). The interction of germintion time germintion moisture ws lso significnt, but s previously discussed this ws due to differences in mgnitude of mlt loss t different moisture levels, when considered cross germintion times (Figure 2). Although the reltive rnk ws the sme cross times, the difference in mlt loss between dys incresed s the germintion moisture level chnged from 40% (3.0%), 45% (6.9%), nd to 48% (7.9%). This indictes the losses increse with time, but becme more pronounced t higher germintion moisture levels. The response of mlt loss to germintion moisture ws not liner, s indicted by the significnt qudrtic contrst (P < 0.0001) (Tble 2). However, response of mlt loss to germintion time ws liner (P<0.0001). Smple ws significnt, but hd reltively smll impct on mlt loss (Tble 1). The verge loss of smple A ws only 0.67% higher thn tht for smple B. 26

MALT LOSS (%) 20 15 10 5 40% Moisture 45% Moisture 48% Moisture d e h gh d c fg c b f c 0 3 4 5 6 DAYS OF GERMINATION Fig. 2. Interction of Germintion Moisture Level nd Germintion Time on Mlt Loss. Columns within ech moisture level denoted by the sme letter re not significntly different (P 0.05). 27

Tble 1. Men of Rye Mlt Qulity Vlues s Affected by Mlting Opertionl Prmeters 28 Smple Mlt Loss (%) Extrct (% Mlt, db) Disttic Power ( o ASBC) Alph- Amylse Wort Soluble Protein (% Mlt, db) S/T (%) Viscosity (cp) FAN (mg/l) AX (mg/l) A/X Rtio (%) β-glucn (mg/l) A 11.92 84.02 131.83 83.20 8.15 64.16 5.23 223.02 3791.50 85.23 68.61 B 11.25 b 87.08 b 115.19 b 88.59 b 7.19 b 65.93 b 4.47 b 221.79 3849.21 88.13 59.43 b Germintion Time Three dy 8.55 84.98 117.75 69.27 7.46 63.41 5.37 217.09 3579.73 88.92 113.18 Four dy 10.41 b 85.69 b 121.46 b 80.62 b 7.68 b 65.15 b 4.98 b 223.41 b 4112.21b 87.65 66.89 b Five dy 12.91 c 85.66 b 129.87 b 93.45 c 7.68 b 65.06 c 4.59 c 227.05 b 3860.13 b 84.43 43.11 c Six dy 14.47 d 85.90 b 124.96 c 100.23 d 7.85 b 66.55 d 4.46 c 222.06 b 3729.34 b 85.72 32.89 d Grin Size Plump 11.52 85.93 123.41 83.24 7.92 67.17 4.83 230.58 3823.27 84.82 63.66 Medium 11.65 85.18 123.61 88.55 b 7.42 b 62.91 b 4.87 214.23 b 3817.43 88.54 64.38 Moisture 40% 5.61 85.87 108.17 71.30 7.61 64.71 b 5.36 215.91 3907.19 85.12 136.13 45% 12.74 b 84.75 108.97 82.81 b 7.79 b 66.10 b 4.76 b 211.03 3834.16 90.51 40.19 b 48% 16.40 c 86.05 b 153.39 b 103.57 c 7.61 64.31 4.43 c 240.28 b 3719.70 84.41 15.74 c Mens followed by the sme letter re not significntly different (P 0.05).

Tble 2. Anlysis of Vrince for Mlt Loss nd Extrct Dependent Vible Source DF Men Squre F vlue Pr > F Mlt Loss Smple (V) 1 10.65 11.28 0.0013 Size (S) 1 0.39 0.41 0.5219 Germintion Moisture (G_M) 2 963.55 1020.53 <0.0001 Liner G_M 1 1862.97 1973.12 <0.0001 Qudrtic G_M 1 64.15 67.94 <0.0001 Germintion Dys (G_D) 3 165.34 175.11 <0.0001 Liner G_D 1 492.53 521.66 <0.0001 Qudrtic G_D 1 0.51 0.54 0.4657 Cubic G_D 1 2.97 3.14 0.0806 V*S 1 0.96 1.02 0.3166 V*G_M 2 0.38 0.40 0.6714 S*G_M 2 0.28 0.30 0.7423 V*G_D 3 0.57 0.60 0.6176 S*G_D 3 0.61 0.64 0.5895 G_M*G_D 6 10.90 11.55 <0.0001 Error 70 1.03 Extrct Smple (V) 1 208.88 122.54 <0.0001 Size (S) 1 12.62 7.41 0.0083 Germintion Moisture (G_M) 2 13.75 8.07 0.0007 Liner G_M 1 0.51 0.30 0.5875 Qudrtic G_M 1 27.00 15.84 0.0002 Germintion Dys (G_D) 3 3.18 1.86 0.1440 Liner G_D 1 7.88 4.63 0.0351 Qudrtic G_D 1 1.24 0.73 0.3969 Cubic G_D 1 1.22 0.72 0.4007 V*S 1 0.19 0.11 0.7366 V*G_M 2 22.26 13.06 <0.0001 S*G_M 2 1.01 0.59 0.5551 V*G_D 3 2.16 1.27 0.2932 S*G_D 3 3.21 1.88 0.1411 G_M*G_D 6 2.67 1.56 0.1714 Error 70 1.70 29

5.3. Mlt Extrct Mlt extrct ws significntly ffected (P 0.05) by cultivr, germintion time, nd germintion moisture level (Tble 1). The interctions of smple germintion moisture, smple germintion time, nd germintion moisture germintion time were lso significnt (Tble 2). The response of extrct to germintion moisture ws non-liner response (Tble 2) nd ppers to be due to the low extrct of smple A t 45% moisture. Figure 3 clerly shows the different responses of two smples cross germintion moisture levels. Compred with smple A, smple B hd reltively high extrct t ech germintion moisture level. However, the extrct of smple A ws significntly lower t 45% thn t either 40% or 48%. These reson for these results is not cler, but experimentl error should not be ruled out. By contrst, the model indictestht the overll response of extrct to germintion ws liner (Tble 2), nd s expected extct incresed with germintion time (Tble 1). However, differences were only significnt between dys 2 nd dys 4-6. The smples lso did not respond uniformly to germintion time (Fig. 4). Incresing germintion time helped smple A chieve higher extrct from 3 to 4 dys, while the differences for smple B cross times were miniml. Exmintion of Figure 5 shows tht high levels of extrct were chieved t four dys of germintion with 40 nd 48% moisture levels, nd did not chnged gretly therefter. Agin the behvior t 45% is unusul. There ws more pronounced effect of incresing time t this moisture level, which suggests some problem with modifiction tht did not exsist t the lower nd higher moisture levels. While the effect of kernel size on extrct ws not significnt (Tble 2), the interction of kernel size germintions time ws significnt t P 0.05. Extrct levels for the medium kernel frction continued to increse cross six dys, while levels for the plump frction peked t 4 dys, nd declined therefter. The decline in extrct is not unexpected in 30

mlts s they become over-modified, but the rte of modifiction in brley is generlly seen to progress fster with smller kernels. EXTRACT (%) 93 90 87 84 81 Smple A Smple B b c b 78 75 40 45 48 MOISTURE (%) Fig. 3. Interction of Smple nd Germintion Moisture Level on Extrct. Columns within ech smple level denoted by the sme letter re not significntly different (P 0.05). 90 88 Smple A Smple B EXTRACT (%) 86 84 c b b b 82 80 3 4 5 6 DAYS OF GERMINATION Fig. 4. Interction of Smple nd Germintion Time on Extrct. Columns within ech smple level denoted by the sme letter re not significntly different (P 0.05). 31