UM H H l | E M THE RELAHDNSHEP €35 Si’lKE DRYING RAT‘TE AM) GRAEN COLOR IN BARLEY Thesis for the E‘segree of NE. 8. fii‘éflfiéfim STE-3E UWER‘SHY 10H? Bfafifihm 1970 THEG‘S LIBRARY ‘ ' Michigan Start“ 1L113i171tt1iit37 J. rw’rv ‘: BIND'ING BY ‘23; _- me a. my if WHEEMEUEE' ABSTRACT THE RELATIONSHIP OF SPIKE DRYING RATE AND GRAIN COLOR IN BARLEY BY John Barnard Sampling of twelve barley lines indicated varietal differences in grain color at harvest. The two-row lines gave the better colored grain. Within the six-row lines the lax spikes, i.e., the spikes with the longer internodes, tended to produce less discolored grain. Six types of barley spike, selected for visual differences in shape and form, were collected and dried in the laboratory at room temperature. Computed 'drying indices' were tested for association with grain color as determined from field collections of the same lines taken at harvest. A significant correlation was observed between final grain color and drying index. Spikes from the two- row barleys were observed to possess the lower drying indices, indicating that they dried faster than the six- row. In one of the six-row lines, drying rate and inter- node length were correlated. John Barnard It was concluded that spike drying rate was assoc— iated with structure and that the faster drying spike tended to produce the less discolored grain. THE RELATIONSHIP OF SPIKE DRYING RATE AND GRAIN COLOR IN BARLEY BY John Barnard A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Crop and Soil Sciences 1970 ACKNOWLEDGMENTS The author expresses his appreciation to Dr. J. E. Grafius who suggested the present project and who guided the progress, and to Drs. C. E. Cress, D. D. Harpstead, C. M. Harrison and D. H. Smith for their thorough reading and criticism of the manuscript. The author also recognizes the invaluable contact with Drs. Chung Lee and R. L. Thomas during the execution of this thesis. Finally, the author thanks Dr. Noel N. Standridge of the U.S.D.A. Barley and Malt Laboratory, Madison, Wis- consin, who consented to perform agtron determinations on the barley samples collected for this thesis. ii TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES LIST OF FIGURES . . . . . . . . . . . . LIST OF PLATES INTRODUCTION LITERATURE REVIEW . 1. Grain color formation 2. The relevance of color to quality. 3. Moisture loss in small grain spikes EXPERIMENTAL SECTION 1. Field survey Materials and methods Results . . . . . . . . 2. Drying experiment . . . . . . . Materials and methods Results DISCUSSION . . . . . . . . . . . . . . BIBLIOGRAPHY . . . . . . . . . . . . . iii Page ii vi LIST OF TABLES Table Page 1. Simple correlations between a visual color score and numbers of various microbes found on and in barley grains from 84 different sources . . 4 2. Analysis of variance for the East Lansing field survey. Agtron value is the dependent variable . . . . . . . . . . . . . . . . . . . 14 3. Analysis of variance for the Hillsdale County field survey. Agtron value is the dependent variable . . . . . . . . . . . . . . . . . . . 14 4. The relationship of agtron color at harvest to internode length and spike length . . . . . . 15 5. Regression statistics for the drying experiment . . . . . . . . . . . . . . . . . . 23 6. Test for heterogeneity of within line regression coefficients (drying indices) in the drying experiment . . . . . . . . . . . . . . . . . 25 7. Maturity scores, untransformed and transformed drying indices for individual lines . . . . . 29 8. Correlations of drying index with structural features of the spike . . . . . . . . . . . . 32 iv LIST OF FIGURES Figure Page 1. The change in agtron value with time at East Lansing. The five lines giving the highest agtron values at harvest . . . . . . . . . . 10 2. The change in agtron value with time at East Lansing. The five lines giving the lowest agtron values at harvest . . . . . . . . . . ll 3. The change in agtron value with time at Hillsdale County. The four lines giving the highest agtron values at harvest . . . . 12 4. The change in agtron value with time at Hillsdale County. The four lines giving the lowest agtron values at harvest . . . . . 13 S. The change in the ratio seed #/3gm with time . 28 6. The relationship of drying index, b, and agtron value at harvest . . . . . . . . . . . 30 7. The relationship of transformed drying index, b', and agtron value at harvest . . . 31 8. Factors influencing barley grain color . . . . 34 LIST OF PLATES Plate Page 182. The change in agtron value with time illustrated by sequential samples from one replicate at East Lansing . . . . . . . . 16 3. The drying experiment . . . . . . . . . . . . . 19 4. Detail of the assembly in the drying experiment . . . . . . . . . . . . . . . . 20 vi INTRODUCTION Grain color is an important quality in malting barley since it is an indicator of microfloral contamina- tion. Microbial contamination has been shown to be respon— sible for a number of undesirable qualities in malt and malt products. The relationship of poor barley color, variously termed as 'staining,‘ 'weathering' and 'scabbing,' to high humidity environments is apparent from the literature. It has been suggested that spike morphology might be one of the factors determining the moisture environment of un- threshed grain. In particular, the more streamlined two- row spike might dry faster than the aerodynamically more complex six-row. If differences in spike drying do occur, then it is possible that such differences will explain, to some degree, intervarietal differences in grain color at harvest. In the present thesis, spike morphology and its effect on drying is studied. The postulated effect of spike drying rate on grain color is examined. LITERATURE REVIEW The literature pertaining to the etiology of grain color was reviewed and the industrial significance of dis- colored barley was briefly examined. The importance of high humidity as an environmental prerequisite of grain discoloration led to a consideration of factors affecting moisture loss in the small grain spike. 1. Grain color formation Grain color is a function of genotype, climate under which the grain is grown and stored, and the activity of microflora in association with the grain. Varietal differences in kernel pigmentation were described by Harlan (1914). Differences were attributed to the varying presence, absence or combination of blue anthocyanin in the aleurone layer, red anthocyanin in the glumes and pericarp, and the degree of melanization. Hence, the 'blue grains' of the brewer can be accounted for by a blue aleurone pigmentation together with a clearing of overlying tissue during steeping. Mullick gt 31. (1958) discuss the genetic aspects of anthocyanin pigmentation. Of more immediate interest as a factor in the for- mation of color, is the 'weathering' discoloration of barley due to staining and surface molding during the latter part of the growing season. This discoloration is especially noticeable in wet seasons. Weathering discoloration appears to be a complex phenomenon that usually involves microfloral activity. Pepper (1960) has undertaken an extensive review of the literature concerning grain discoloration in wheat and barley and the reader is directed to this work for complete literature references. Two of Pepper's own experiments will be mentioned here. In one experiment Pepper plated grains onto agar plates inoculated with bacteria and produced severely dis- colored grain. Varietal differences in the location and intensity of staining were observed. In an experiment with grain samples from diverse locations within the United States, Pepper cultured micro- flora present on and in the grain and recorded the rela- tionship of a visual color score with numbers of specific organisms. Correlations calculated from Pepper's data are given in table 1. The color association with Alternaria is striking although the presence of most other microbe groups was also correlated with discoloration. There is evidence that not all staining involves microflora. Broadfoot and Robertson (1933) isolated Alternaria, Cladosporium and Macrosporium from glumes of Reward wheat but concluded that staining was mainly due to the interaction of light with color factors present in certain strains of the wheat. Table l.--Simple correlations between a visual color score and numbers of various microbes found on and in barley grains from 84 different sources. Organism r Alternaria .83 ** Helminthosporium .49 ** Fusarium .47 ** Cladosporium .36 ** Bacteria .19 Yeasts .28 ** Storage fungi .13 Others .10 ** sig. at P=.01 Hagborg (1936) discussed the influence of environ- mental factors in the browning process generally associated with certain pathogens. Johnson and Hagborg (1944) in a greenhouse exper- iment with wheat, concluded that melanism could be induced by environmental conditions alone, especially high tempera- tures with high humidities. In their experiment, Johnson and Hagborg proceeded to demonstrate the absence of path- ogens that could have been implicated as causing the browning. Pepper (1960) discusses a possible biochemical model for the discoloration process. It is suggested that a material in the husks may be oxidized or polymerized by the environment or by interaction with certain microflora. The phenolic nature of the staining and its possible con- nection with disease resistance mechanisms is discussed. From the literature, Alternaria emerges as an organism commonly associated with grain discoloration. The reduced infection by Alternaria in drier climates has been reported by Whitehead (1949). Pepper's data (1960) bears out the association of drier climates with reduced Alternaria infection and reduced grain discoloration. Johnson and Hagborg (1944), mentioned above, indicate the importance of humidity in the non-microbial discoloration process. In conversations with S. T. Dexter it was deduced that spike morphology might affect staining by influencing the humidity of the microclimate. Differences in spike morphology might explain, to some extent, the differential staining observed between lines of barley. Grafius (1969) suggested that the success of the morphologically distinct two-row barley in Europe may, in part, be due to its more rapid drying. 2. The relevance of color to quality The importance of grain color as a quality feature of malting barley is due to its close association with the microfloral burden on the grain. The following listing of some of the principle effects of high levels of microflora on the brewing process is taken from a review by Anderson, Gjertsen and Trolle (1967). Some of the important effects are: (1) More rapid uptake of water on steeping (2) High protein modification resulting in high nitrogen wort and beer (3) Increase in beer color (4) Abnormal beer taste and aroma (5) Beer from weathered barley was susceptible to gushing Kneen (1963) in an interesting demonstration, com- pared a bright barley sample with two weathered ones. One of the weathered samples came from a field source; the other had weathering stimulated by treating the steeped barley with a large dose of microbial spores. Both natural and synthetic weathering gave similar results, viz., lower agtron values, increases in nitrogen modification and beer color, decrease in haze stability, and poor flavor. 3. Moisture loss in smallgrain spikes Tull (1733) was probably the first to mention dif- ferences in Spike drying characteristics when he remarked that awnless wheat ”does not hold the drops of rain so long as the bearded (or cone) wheat." Harlan (1923) in his examination of the water relations of barley kernels, suggested that loss of water would be impeded in those kernels which were covered by awns or overlain by other kernels. Miller, Gauch and Gries (1944) refer to work by a number of European researchers pertaining to the role of awns in transpiration. Zoebl and Mikosch (1892) found that removal of awns from barley reduced spike transpiration by 75%. Vasilyev (1897), working with wheat, barley, rye and Stipa capillata, confirmed the conclusions of Zoebl and Mikosch. Schmid (1898) found that removal of wheat awns lowered transpiration rate by 10 to 30%. Pool and Patterson (1958) discussed moisture rela- tions in certain wheats. Varietal differences in rate of drying were observed after rains or dews. In a laboratory study, awns were shown to increase the magnitude of moisture losses and gains in ripe wheat stands. It was suggested that selection for awns in a breeding program would be an efficient method of obtaining faster drying varieties for humid climates. The presence of waxy glumes was shown to slow moisture changes in the grain. EXPERIMENTAL SECTION 1. Field survey The field survey was conducted in order to determine the degree of change of grain color, as determined by the agtron reflectance meter, with time, and to assess interline differences in color. Materials and methods Samples were collected from the 1969 winter barley variety trials at Hillsdale County, Michigan, and at East Lansing, Michigan. Two replicates were sampled at each location. The lines sampled were selected to obtain a range of spike conformations. Samples were taken at four weekly intervals with the fourth coinciding with harvest. The samples were dried and then submitted to the U.S.D.A. Barley and Malt Laboratory, Madison, Wisconsin, for color measurement using the agtron reflectance meter. Eight six-row lines were taken from the plots at Hillsdale County. The same six-row lines were collected at East Lansing and in addition two two-row types were taken. The first collection of samples from Hillsdale was of in- sufficient size for agtron determination. Results Figures 1 to 4 show the time trends of color ob- served in the experiment. In all cases, agtron value declined with time indicating reduction in color quality. In the variety Wong and in the two-row lines some recovery was observed at the last collection. Such recovery could be real, perhaps due to environmental bleaching, but the possibility of sampling error remains. Tables 2 and 3 show the analyses of variance for the experiment. As expected, line and collection effects were very highly significant. The significance of the interaction at East Lansing reflects the relative recovery of agtron value in Wong and the two-row barleys at the last sampling. Within each location, lines were ranked according to final agtron value and examined for visual correlation with Spike character. The results are shown in table 4. The samples from East Lansing demonstrate a rela- tionship between spike density, as indexed by internode length, and color. The samples from the last collection at Hillsdale were at a physiologically later stage, were all badly discolored, and no pattern was readily discernible. At East Lansing the two-row lines produced the better colored grain. Cass, a variety currently being examined for possible malting barley production in south- western Michigan, was the superior barley with respect to 100? A 62-420-21}2_r0w Agtron A 62-414-17 color . Cass v 62-445-6 }6-row A o 62-445-110 90- 80" O V A 70’ 0 A 0 60- \\\\\x O 50“ 40- 'L l l J June June July July 19 26 3 11 Date of collection Figure l.--The change in agtron value with time at East Lansing for the five lines giving the highest agtron values at harvest. 11 A 62-434-5 v 62-433-123 90 - 0 Lakeland all 6-row A. 62-449—22 Agtron . Wong color 80 ~ 70 F 60 — SO _ O A 40__ \ O 30 _ L I 1 I June June July July 19 26 3 11 Date of collection Figure 2.--The change in agtron value with time at East Lansing for the five lines giving the lowest agtron values at harvest. 12 A 62-434-5 A 62-449-22 o 62-433-123 all (”row V’ Lakeland Agtron color 70’ A 60 ' ‘V 50 — A 40 - 30 - ‘V 20 L I 4 I June July July 25 l 10 Date of collection Figure 3.--The change in agtron value with time at Hillsdale County for the four lines giving the highest agtron values at harvest. 13 V Cass A 62-445-6 ‘ Wong all 6-row o 62-445-110 Agtron color 70r 60* A 50- V’ 40 _ V' A. 30 _ A L 20 __fl6 I 1 I June July July 25 l 10 Date of collection Figure 4.--The change in agtron value with time at Hillsdale County for the four lines giving the lowest agtron values at harvest. 14 Table 2.--Analysis of variance for the East Lansing field survey. Agtron value is the dependent variable. Source df SS MS F Reps 1 25.3500 25.3500 1.05 Lines 9 1558.1062 173.1229 7319 *** Collections 3 8496.0128 2832.3042 117.65 *** L x c 27 1725.7548 63.9168 2.66 ** Residual 29 698.1499 24.0741 Total 69 12503.3737 Table 3.--Analysis of variance for the Hillsdale County field survey. Agtron value is the dependent variable. Source df SS MS F Reps 1 50.3809 50.3809 2.75 Lines 7 767.3809 109.6259 5.98 *** Collections 2 5256.3809 2628.3519 143.38 *** L x C 14 518.0889 37.0063 2.02 Residual 20 366.6190 18.3309 Total 44 6959.1734 *** sig at P=.001 ** sig at P=.Ol 15 color. The lines with the more compact spikes, i.e., those with shorter internodes, were observed to collect at the bottom of table 4. Wong, a compact-spiked, six-row, and the only awnless barley included in the experiment, pro- duced the poorest colored grain. Table 4.--The relationship of agtron color at harvest to internode length and spike length. Agtron Av. internode Av. spike East Line length (cm) length (cm) Lansing Hillsdale 62-420-21 59.0 - ]2-row 62-414—17 51.5 - Cass .81 6.5 50.0 23.0 62-445-6 .85 6.3 46.5 21.5 62-445—110 .76 6.3 46.5 19.0 62-434-5 .76 5.6 43.5 31.0 62-433-123 .76 4.7 40.5 24.5 Lakeland .56 4.4 40.5 23.5 62-449—22 .68 4.3 38.0 25.0 Wong .54 4.3 35.0 20.0 2. Drying experiment The objectives of the drying experiment were to establish whether or not differences in spike drying existed between the lines examined, and to determine whether any such differences could be attributed to structural variation. 16 Plates 1 and 2.--The change in agtron value with time illustrated by sequential samples from one replicate at East Lansing. 18 The drying experiment and the field survey (exper- iment l) were tied together with a correlation analysis to ascertain whether any association existed between spike drying and grain color. Materials and methods Twelve spikes from each of six varieties were col- lected from the 1969 winter barley trials at East Lansing, Michigan. The six varieties used were among those repre- sented in the field survey (experiment 1). The spikes were detached with 8 centimeters of culm retained with the spike. The specimens were immediately transferred to the laboratory and supported on modified paper clips, affixed to a table top, to ensure free drying. The spikes were arranged in a twelve-replicate randomized block design. Additional border blocks, around the margins of the table, eliminated any edge effect. The assembly is shown in plates 3 and 4. At intervals, individual spikes were 'unplugged' from the paper clips and weighed. A systematic order of weighing was followed to ensure that the same time interval elapsed between successive weighings of the individual spikes. Weights were taken over a period of 84 hours. The following measurements were taken on individual spikes: spike length, number of nodes on the spike, diam- eter across the Spike (lateral floret to lateral floret), diameter through the spike (central floret to central floret), and average awn length. m 19 Plate 3.--The drying experiment. “I 20 Plate 4.--Detail of assembly in the drying experiment. 21 Results Exponential drying functions of the following form were assumed and applied to the data. M1: = M0 exp(-bt) where Mt = moisture % of spike at time t M0 = moisture % at collection (time [mug zero) % t = time elapsed since collection E b = drying index :1 b being the parameter of particular interest in this study. Br?“ The above exponential function reflects the progres- sive reduction in rate of moisture loss as the spike dries out. The equation can be derived as follows. It is as- sumed that rate of moisture loss at any given time is directly proportional to the moisture content at that time. Symbolically it follows, dM=_ a? bM 0 where M = moisture content, 6 dM _ . . . d? - rate of change of m01sture content with tlme b = proportionality constant (minus because mois- ture % is decreasing) rearranging and integrating over time elapsed Since collec- tion, i.e., since t 22 1: Mt dM _ _ ’5 M—-- b dt MO 0 1n Mt - ln M0 = -bt discarding logarithms and rearranging, Mt = M0 exp(-bt) (0.5.0.) By taking natural logarithms of both sides of the above equation, the statistical mechanics of curve fitting "9.- 0 become simply that of linearly regressing the logarithm of moisture percentage on time, i.e., 1n Mt = In M -bt 0 ln MO appears as the overall constant, a, of the familiar Y = a - bx Curves were fitted to individual spikes and to spikes pooled together into varieties. The regression statistics are given in tables 5 and 6. When the spikes were collected, differences in maturity were observed between varieties. In particular, the variety Wong was found to be in the dough stage while the other five lines were in various stages of milkiness. 23 uooo Fourc-- chem. mmcfic.- comm. mmmao.- wmmm. mmnao.- Nwmm. n¢mao.- wfimm. mmNHo.- memo. wmmfio.- ammo. oomao.- ammo. mmmao.- ovmm. wuoao.- Nmmm. movao.- mmmm. omqao.- mooo. omooo. wwwao.- mcoz Hmom. owvao.- comm. ommao.- Hwom. mmnao.- comm. mmvao.- mwvm. Howao.- memo. mfivao.- Nwmm. mmmHo.- mama. owvao.- mmmm. omwfio.- mafia. ommao.- comm. omeao.- momm. HomHo.- mooo. wmooo. mm¢ao.- ma-¢av-mo Homo. vqoao.- mama. owoao.- Nnmm. moqao.- comm. NmHNo.- owmm. mmwao.u Hmmm. onmao.- wmwm. mmoao.- ammo. vvmao.- nmmm. mefio.- mmmm. vafio.- Hawm. mNmHo.- omwm. wm5H0.- mooo. wmooo. mmwao.- HN-ON¢-N© Na o E o Wm who o... o oofl mafia m wo moxfimm ofimcflm afinpfiz mcoflmmohwom modfla :flcuwz coammohmom :0 p2 CH wo mcoflmmo .ocHH sumo mo moxflmm oamcflm :pflz can mocHH cfinpflz oEfip Hmom .ucoefluomxo mcflxuw ogu How moflumflpmpm coflmmohmom--.m ofinme 24 cowmmopmou kn How woucsooom oHnmmum> pcopcoaow map mo mopmscm mo meow mo comuHomomm one mm mm ucofiommmooo commmonmon may wo yoguo whmvcmpm one mm pm vamp so a: :H mo pcofiommwooo commmoumoh on» mm a emmm. HmmHe.- Hmmm. mmmme.- Hmmm. eemme.- emmm. meeme.- mNmm. NmNHe.- eemm. ememe.- mmmm. meeme.- Nmem. eeNHe.- emmm. hemme.- emmm. ememe.- emem. Nemee.- memm. mmeme.- meee. Nmeee. memme.- NN-mee-Ne Heem. mmeme.- Hmmm. mmeme.- mmmm. eNeme.- mmem. Hmmme.- Nemm. ememe.- meem. meeme.- emNm. mmmee.- Nmmm. Neeme.- mmmm. mmmme.- mmem. Hemme.- mmmm. ememe.- Nemm. mmmme.- meee. Nemee. Hemme.- m-eme-Ne meem. emmme.- eeem. emmme.- emem. ~m~me.- Nmem. eemme.- Nmmm. Nemme.- meem. mmmme.- emem. Nmmme.- mmem. emmme.- mNmm. mHmNo.- omom. NMNHO... macs. wmooo. awmao.: Ucmavxmd 25 Table 6.--Test for heterogeneity of within line regression coefficients (drying indices) in the drying experiment. Line df Residual SS 62-420-21 176 7.9605 62-414-17 175 2.6958 Wong 177 3.0420 Lakeland 178 14.6874 62-434-5 175 60.1515 62-449-22 177 12.4347 1058 100.9719 Single Regression 1063 114.5665 Heterogeneity of slopes 5 13.5946 13.5946 1058 5 100.9719 2.851, differences existing between line drying indices are significant at P=.025 26 In order to obtain a fairer comparison of drying between lines, a set of transformed drying indices was calculated from those determined directly from the experiment, by removing the effects of maturity. Differences between the ratios of seed number/3gm sample from field plot samples taken at the time of the experiment and at harvest, were used as measures of maturity, i.e., m = 81 - S4 84 where m = maturity score 81 = seed #/3gm at time spikes were collected for the experiment S4 = seed #/3gm at harvest Correlations between the maturity scores and the drying indices were removed by the method discussed by Rao (1952) using the maturity scores as the base vector. Thus we have, vector of maturity scores 2 .3 o H o a u b = vector of drying indices b'= vector of transformed drying indices 27 cov(b,m) = covariance of maturity scores and drying indices 0 2 m variance of drying indices e.g., for line 62-420-21, -.02083 = -.0184 - Ef§§§§§§ .84 where b = -.0184 m = .84 cov(b,m) = .00006 om2 = .02099 The coefficient of m_in the equation is chosen such that the covariance (b',m) is set to zero, i.e., cov(b',m) = cov(b,m) - (cov(b,m)) omz = 0 o m The progression of seed number/3gm with time is shown in figure 5. Flattening of the curves indicate the points in time when grain dry matter accumulation has apparently ceased. Since only the initial and terminal points on figure 5 were used in calculating the vector m, non-linearity of the ratio over time was ignored for the present purpose. The transformed drying indices, together with ma- turity scores for the six lines are given in table 7. The two-row lines possessed the higher drying indices. Seed number per 3gms 200 150 100 Figure 5.--The change in the ratio 28 A 62-434-5 B Wong C Lakeland D 62-449-22 E 62-420-21 I— 1 J June June July 19 26 11 Date of collection seed number 3 gms Flattening of the curves demonstrates reduction in dry matter accretion. with time. 29 Transformation effected a greater degree of separation of the two-row types and the six-row types. Table 7.--Maturity scores, untransformed and transformed drying indices for individual lines. Line m b b' 62-420-21 .84 -.0184 -.02083 62-414-17 1.00 -.0149 -.Ol779 Wong .68 -.0149 -.01687 Lakeland .99 -.0139 -.01676 62-434-5 .76 -.0136 -.01580 62-449-22 1.03 —.0135 -.01648 The relationship of the drying indices to grain color at harvest is shown in figures 6 and 7. Comment on these figures will be reserved for the discussion. Regressions of individual spike drying indices on a number of morphological measurements were made. The measurements used were: 1. Average awn length ii. Spike length iii. Internode length = spike len th number 0 nodes iv. Average diameter = diameter across spike + diameter through Spike Drying index 30 -.019 ' r = -.8021 Sig at P=.10 62—420—21 -.018 h -.017 f —.016 - -.015 _ _+_ ‘R— Wong 62-414-17 -.014 _ ‘+‘ Lakeland + + 62-434-5 62—449-22 -.013 - 30 40 50 60 Agtron color at harvest Figure 6.--The relationship of drying index, b, and agtron value at harvest. Transformed drying index 31 -.021’ _+- r = -.8423 Sig at P=.05 62-420-21 -.020' -.019 ' -.018 ‘ -+' 62-414-17 '0017" + Wong _+—Lakeland ‘+’ 62-449—22 ‘0016 '- -+' 62-434-5 -.015 - L 1 . 4. 30 4O 50 60 Agtron color at harvest Figure 7.--The relationship of transformed drying index, b', and agtron value at harvest. 32 v. 'Volume' = nrzh where h spike length r half the average diameter vi. 'Density' = weight of spike Volume' Average diameter may also be interpreted as volume- surface ratio, differing from it by a scaling factor. i.e., Volume = nrzh = diameter Surface anh Regression over all lines together failed to produce any significant results but a number of within line corre- lations were significant. These are given in table 8. Table 8.--Corre1ations of drying index with structural features of the spike Variables correlated with drying index, Line r Significant at P=.05 or less 62-420-21 Awn length -.85 62-414-17 Diameter x internode -.85 Wong Internode length -.81 62-449-22 [Diameter - 97 [Volume -:84 DISCUSSION A suggested model combining the various environ- mental and genotypic factors influencing grain color in barley is given in figure 8. Grain discoloration is most commonly the result of microbe-genotype interaction. Field climatic conditions during the period following spike emergence will influence the build-up of microflora on barley spikes. Humidity, in particular, would appear to be an important factor. It has been suggested that morphological features of the barley spike play an important role in determining the rate of moisture loss during the in-field curing pro- cess. The slower drying spike might be suspected of retaining a higher moisture regime, advantageous for the establishment and maintenance of a microfloral population, and being prone to grain discoloration. Influence of Spike form on any purely environmental discoloration process is also a possibility. The present work attempts to link grain color, as determined by the agtron reflectance meter, with spike drying. 33 34 Field conditions \ \ 3 Mir \. Lod in nu . \\ - \\ 8 g ,' ' fl n- AMicroflora I Staining4’ process(es)' Base p1gmenta:::H\\\\\\\\\\\\‘ COLOR Genotype Figure 8.--Factors influencing barley grain color. The-figure relates the genotypic, environmental and micro- floral factors contributing to the formation of grain color. Arrows indicate paths of influence. The broken arrow indi- qates the postulated effect of spike conformation on micro- floral population. 3S Drying indices, b and b', were computed from a laboratory experiment. It was hypothesized that the indices might be an indication of drying during the period from grain set to harvest. Correlation analysis suggested a number of rela- tionships between spike structure and drying index. Al- though no significant overall relationship was found, within line correlation produced some interesting results (table 8). In the two-row line 62-420-21, drying index was found to be negatively* correlated with average awn length. This result is compatible with the conclusions of Pool and Patterson (1958) who observed that awns increased moisture losses in wheat. In the variety Wong, internode length was negatively correlated with drying index. This relationship is partic- ularly interesting in view of the data in table 4, which demonstrates an association between higher agtron values and longer internodes. Figures 6 and 7 show the association between final agtron readings, i.e., readings from samples taken at har- vest, and drying indices. Figure 6 demonstrates the results for the drying indices calculated directly from the , . *Confusion of Signs should be avoided. It should be remembered that a 'fast' drying index is a larger neg- ative magnitude hence the correlations actually come out negative. 36 experiment. Figure 7 shows the results for the transformed indices described earlier in this thesis. Transformation brought about a greater degree of differentiation of two- row and six-row lines with respect to drying index, and improved the correlation of drying index with final agtron. One deficiency in the concept of drying index is that it does not take into account any drastic morphological r—n changes that may occur as the season progresses. For exam- ple, in some lines fracture or loss of awns occurred during 3 the season. Following such reduction, the drying charac- 9 teristics of the spike might be considerably altered. If 9'” the latter is the case then final grain color could not be expected to be too closely related to an index determined with the awns intact. An experiment which included removal of awns from test lines would help to clarify the position. Two other phenomena that are modifying influences on spike drying are necking and lodging. Necking, i.e., the bending over of the spike that occurs Shortly before maturity, would be expected to induce Shedding of more of the incident rainfall. Lodging, on the other hand, will cause many spikes to be covered with straw and while rain may penetrate the resulting mat, evapOration from under- lying grain will be reduced. In the consequent humid con- ditions the formation of discolored grain would tend to be severe . 37 Grafius (1969) has proposed that better grain color might be one of the benefits to be gained by introducing the two-row character into Michigan winter malting barley. The research herein confirms the relatively rapid drying properties of two randomly chosen two-row lines and their superiority over six-row types with respect to color at harvest. A comparison among the six-row lines further suggests that the lax Spike will tend to produce better colored grain than the compact. From the foregoing it is inferred that drying is most rapid in those spikes which are physically disposed such as to permit a greater degree of air flow through the Spike. In figure 8 a direct path has been inserted between genotype and the staining processes. It is suggested that variation in these processes occurs pep s3 and may be the ultimate factor determining differences in discoloration between lines of similar spike type, lodging resistance and initial pigmentation. The physiological inclination to stain may, in some cases, over-ride the morphological and 'ontological' features discussed in this thesis. It is strongly suspected that in case of the variety Wong, such an over-riding factor was in operation for although a relatively 'fast' drying index was determined for this variety, poor colored grain was consistently produced both in 1969 and in previous years. BIBLIOGRAPHY Anderson, K., Gjertsen, P. and Trolle, B. 1967. The microflora of barley and its effect on wort and beer. Brewers Digest, August. Broadfoot, W. C. and Robertson, H. T. 1933. Pseudo-black chaff of Reward wheat. Sci. Agr. 13: 512-514. Grafius, J. E. 1969. Malting quality in winter barley: a progress report. (unpublished) Hagborg, W. A. F. 1936. Black chaff, a composite disease. Can. J. Res. 14C: 347-359. Harlan, H. V. 1914. Some distinctions in our cultivated barleys with reference to their use in plant breed- ing. U.S.D.A. Bulletin 137: 1-38. 1923. Water content of barley kernels during growth and maturation. J. Agr. Res. 23: 333-360. Johnson, T. and Hagborg, W. A. F. 1944. Melanism in wheat induced by high temperature and humidity. Can. J. Res. 22C: 7-10. Kneen, E. 1963. Proc. Irish Maltsters' conference, 51 [cited in Anderson, Gjertsen and Trolle (1967)]. Miller, E. C., Gauch, H. G. and Gries, A. G. 1944. A study of the morphological nature and physiological function of the awns of winter wheat. Kansas Agr. Exp. Sta. Tech. Bulletin 57. Mullick, D. B., Faris, D. G., Brink, V. C. and Acheson, R. M. 1958. Anthocyanins and anthocyanidins of the barley pericarp and aleurone tissues. Can. J. Plt. Sci. 38: 445-456. Pepper, E. H. 1960. The microflora of barley, their iso- lation, characterization, etiology and effects on barley, malt and malt products. Ph.D. thesis, Michigan State University. 38 39 Pool, M. and Patterson, F. L. 1958. Moisture relations in soft red winter wheat, I and II. Agron. J. 50: 153-157, 158-160. Rao, C. R. 1952. Advanced statistical methods in bio- metrical research. Wiley 6 Sons, N. Y. Schmid, B. 1898. Bau und funcktionen der grannen unserer getreidearten. Bot. Centrbl. 76: 1-9, 36-51, 70- 76, 118-128, 156-166, 212-221, 264-270, 301-307, 328-334 [cited in Miller, Gauch and Gries (1944)]. Tull, Jethro. 1733. The horse-hoing husbandry: or, an essay on the principles of tillage and vegetation. Strahan, London. Vasilyev, N. I. 1897. On the role of awns of gramineae. Zap. Novo-Alexandri Inst. Selsk. Khoz. i Lyesov 10: 119-168, Exp. Sta. Rec. 10: 718 [cited in Miller, Gauch and Gries (1944)]. Whitehead, M. D. 1949. Studies on some seed-borne micro- organisms of cereals. Ph.D. thesis, University of Wisconsin. [cited in Pepper (1960)] Zoebl, A. and Mikosch, C. 1892. Die function der grannen der gerstennahr. Sitzungsber. d. k. Akad. d. Wissensch in Wien. Math.-Naturw. Classe 101: 1033- 1060 Ann. Agron. 21: 143-144 Bot. Centrbl. 54: 240 [cited in Miller, Gauch and Gries (1944)].