RUMEN EN VITRO S'EUDIES OF ALFALFA Thesis for the Degree 0f M. 8‘ MCHiGAN STATE UNEVERSW JOHN S. SHENK, II 1967 “1- it?" LIB R A R Y Mlcizigd tStatC a Htfllfi \‘ \ \\ \ \ III‘I“ \ “I‘| \ \ \ ‘\I|I \l-In-IIn' ‘l‘l‘l‘ l||!|\|\|.| \ cl‘l‘ll‘l‘ I‘|\I‘\ \‘II ‘\I\I| m \ x M; \ M“ 3 gwwzugm \\\ I University 53“ It '(* ABSTRACT RUMEN IE VITRO STUDIES OF ALFALFA by John S. Shenk, II Four alfalfa populations were analyzed by a six hour i_n v_itr_o technique. The analysis of sixty-four in-place clones of the variety Tuna over a two year period indicated that real differences existed among clone means, although environmental influence and inter- action does occur. Insect and disease resistance did not seem to be related to % DMD (per cent dry matter disappearance) in the MSA-C4 and MSB-C4 populations. The mean % DMD values for these populations were equal to or higher than the susceptible Tuna population. Correlations between yield and % DMD were nonsignif- icant in any of these three populations. A population of selected alfalfa clones, moved from the field toythe greenhouse over the winter and re-evaluated in the field the following year, were inconsistent in their DMD response. Plants placed in abnormal environmental situations may need a John S. Shenk, II period of adjustment before valid DMD comparisons can be made. Chemical or biological treatment of forage is effective in increasing % DMD. NaOH pretreatment gave the greatest over- all response; however, the increase was in solubility and not microbial digestibility. Preliminary screening of large populations of clones with this _i_n E32 technique can be effectively executed with only one determination per sample. RUMEN I_l_\I_ VITRO STUDIES OF ALFALFA By John S. Shenk, II A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Crop Science 1967 ACKNOWLEDGMENTS The author would like to express his sincere apprecia- tion to Dr. F. C. Elliott, whose helpful suggestions and criticisms have been invaluable in preparing this manuscript. I am indebted to the following individuals for helping me develop certain aspects of this thesis: Drs. J. W. Thomas, C. E. Cress, C. M. Harrison, and D. W. Allinson. I wish to express my appreciation to Mrs. V. Satari, who assisted with the laboratory work, and to my wife, Gloria, who prepared the typescript. ii TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION LITERATURE REVIEW . Improvement of alfalfa quality Appraisal of in vitro techniques Digestibility and the lignin- -ce11ulose complex . MATERIALS AND METHODS Plant sources and management . . Analysis technique for 6 hour in vitro DMD Forage treatment and antidote procedure Statistical methods RESULTS Tuna population . . . . MSA- C4 and MSB- C4 populations . Vernal, DuPuits, and Culver clones Forage treatment and DMD . DMD pilot study. DISCUSSION . Unselected Tuna population Disease and insect resistant MSA- C4 and MSB- C4 pOpulation . Selected clone pOpulation iii Page ii vii C500 11 11 12 14 15 17 17 21 22 25 3O 33 34 36 37 Page The solubility and digestibility of treated forage . . . . . . . . . . . . . . . . . . 37 Suggested changes in DMD mass selection technique . . . . . . . . . . . . . . . . 39 SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . 41 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . 43 iv Iill‘lnll ' Table LIST OF TAB LES 1965-66 cutting means and ranges of the % DMD values for 64 unselected Tuna clones Proportional analysis of variance of the % DMD values in an unselected Tuna population Correlation values between characteristics of the Tuna population in 1965-66 Means and ranges for the % DMD of MSA-C4 and MSB-C4 populations. Simple correla- tions between % DMD, dry weight, color, and harvests . . . . . . . Range, mean, and number of clones in each variety represented in the selected pop- ulation. Varieties divided further into sub-groups high and low. Values ex- pressed as % DMD . % DMD for alfalfa forage T46 pretreated 18hours Forage T46 pretreated 12 hours. Solution removed and antidotes added one hour before fermentation. Values expressed as 0/0 DMD . The % DMD, % water soluble, and % microbial digestion of three Tuna forages: T95, T108, and T46. Page 17 19 20 21 25 26 27 28 Table Page 9. The % DMD, % water soluble, and % microbial digestion of three Tuna forages pretreated with NaOH. Calculated values derived by subtracting the untreated values (Table 8) from the treated values . . . . . . . . . . . . . 29 10. The % DMD, % water soluble, and % microbial digestion of three Tuna forages pretreated with NaOH and dried before fermentation. Calculated values derived by subtracting the untreated (Table 8) from the treated values . . . . . . . . . . . . . . . . . . . . . 30 11. Partially nested analysis of variance table of DMD sampling pilot study . . . . . . . . . . . . 31 vi LIST OF FIGURES Figure 1. Histograms of the unselected Tuna clones for three harvests and two years. % DMD values used to measure the population 2. Histograms of the MSA-C4 and MSB-C4 field and greenhouse populations. % DMD values used to measure the population 3. Histograms of the selected clones for three harvests and two years. % DMD values used to measure the pOpulation vii Page 18 23 24 INT RODU C TION The forage breeder, attempting to improve the nutritive value of forage, is beset with many obstacles. He must decide what characteristics contribute the most to nutritive value, devise or modify a selection technique until he is capable of measuring these systems accurately, and then select parental clones, stable in this character, but agronomically acceptable. The lignin-cellulose complex is believed to be one of the plant systems reSponsible for nutritive value, but as yet its expres- sion is not completely understood. Chemical or biological measure- ments of this character are possible; however, since chemical measurements will need to be interpreted in biological terms, i_n m systems have been suggested as a suitable alternative. Bio- assays of this type have only recently been applied to plant selection. The genetic system controlling the nutritive value of perennial forage plants is complex and subject to environmental influence; therefore, parental clones can best be selected by screen- ing large pOpulations of clones over a period of years. Only after relatively stable clones have been positively identified can the plant breeder study this nutritive value character. Preliminary studies were conducted by Allinson (1966). The objective of this research was to determine the relationship between DMD, environment, and agronomic char— acters in four different alfalfa plant populations. A preliminary attempt was also made to increase the % DMD of a forage by chemi- cal or biological treatment. A pilot study was run on the number of sample determinations required for plant selection in large populations. LITERATURE REVIEW Improvement of Alfalfa Quality Alfalfa quality is difficult to define. From a nutritional standpoint it is superior to many forages. The protein level is high and of relatively good quality. Methionine is the only amino acid that may be limiting. It has a high potential energy source, cellulose, but a rather low free sugar and starch content. Adequate minerals are present, especially calcium, and with the exception of B12, it is considered a rich source of vitamins (Bolton, 1952). Excellent references on the chemical composition and feeding value of alfalfa have been published by Morrison (1956), Maynard and Loosli (1956), and Crampton (1956); however, forage composition, as determined by standard chemical procedures, does not necessarily indicate biological activity. Schillinger and Elliott (1966) isolated an antimetabolite substance from individual alfalfa plants. Although these plants were similar in chemical composition to plants without the antimetabolite, their toxicity was demonstrated in three biological assays. Hanson e_t_a_1. (1963), studying the saponin content of four alfalfa varieties (Buffalo, Ranger, Lahontan, and Vernal) from eight locations in the United States, found differences among the varieties. Lahontan had the lowest saponin content and DuPuits the highest. Substances have been extracted from alfalfa which are believed to affect vitamin utilization in growth. Pudelkiewicz and Matterson (1960) indicated that an inhibitor was present in the lipid portion of the separation. Sterol compounds did not appear to effect its reaction nor did the fibrous particles related to the lignin fraction. They postulated the action of the inhibitor was to prevent the utilization of tocopherol, a precurser of vitamin E. Ayala _e_t_a_l_. (1951) extracted an inhibitor from dried alfalfa juice that was not destroyed by heat or chemical treatments. They found that the inhibitor was present in greater concentrations in the dried juice than in the dehydrated meal; moreover, the addi- tion of vitamin B12 was effective in partially overcoming the in- hibitor effect. No two alfalfa plants have the same genotype unless asexually propagated; therefore, the potential for diversity within varieties and species should be great. Elliott (1963) suggested that the unit of selection in alfalfa pOpulations should be the individual clone. Individual plant selection was used by Bolton and Cormack (1953) to develop the varieties Saskatoon, and Viking from Grimm. The quantitative characters of winter hardiness and yield were the criteria used to select these clones from the Grimm popu- lation. An early study by Tisdale it _a_l_. (1950) indicated that inherent differences existed between the methionine content of two alfalfa clones. Singleton 233.1: (1952), evaluating one hundred clones for methionine content, could not establish variety differences, but found high and low clones within the population. The idea of removing toxic substances with a breeding program is not new. Atwood'and Sullivan (1943), using a combina- tion of chemical identification and plant breeding techniques, removed the presence of HCN from a ladino clover strain. Allinson (1966) found that forage from a single alfalfa clone could cause a differential response when analyzed by two entirely different bioassay systems. Any breeding program using chemical or bioassay analysis as a selection tool will be limited by the sensitivity of the assay to mea— sure the inhibitors within the forage; however, the ultimate goal of the breeding program will determine which type of assay or assays should be used. Bioassay techniques presently employed range from small animals to specific strains of microorganisms. Elliott (1963) derived a means of selecting individual alfalfa clones with the meadow vole. Enzyme systems have been used to predict the nutritive value of forages by Donefer 3t :11. (1963) In addition, enzymes are presently being used to measure the biological activity of alfalfa leaf proteins (Gil 3:31. , 1967). Pederson _e_t a_1_. (1963), using Tricoderma and chicks as bioassays, investigated the activity of saponins. Oldfield e: a_l. (1960) used dehydrated alfalfa samples containing the relative poten- cies of 119mg. of coumestrol in lamb feeding trials. Artificial rumen systems have recently been applied by Schillinger (1965), Allinson (1966), and Gil gt a_l. (1967), to evaluate individual alfalfa clones and their progenies. Barnes eta} (1964), El-Shazly e_ta_l. (1960) and Barnes (1965) have published excellent reviews on the different in vitro techniques. Allinson (1966) used a six hour in m technique to screen large numbers of individual alfalfa clones. Clones of dif- ferent nutritive value were differentially susceptible to lignin extrac- tion. Lignin was significantly more extractable from clones of a high nutritive value. Appraisal of In Vitro Techniques Significant correlations have been found between E vitro and i_n vivo digestibility. Bowden and Church (1962b) analyzed forage samples for which live animal digestion coefficients were determined with sheep prior to E vitro digestion. Baumgardt et a1. (1962), using forage evaluated with dairy heifers and sheep, correlated the digestible organic matter, dry matter, and energy, to an .13 vitro system. Highest correlations were obtained with samples analyzed shortly after they were harvested and ground (Bowden and Church, 1962a). Ingalls (1964) found significant correlations between the 6 hour 12% technique and the nutritive value index; 36 hour fermentation was correlated with total animal digestion. Allinson (1966) concluded that the 6 hour fermentation technique was an effective tool for predicting the nutritive value of a forage. Day to day variation was significantly less for a 48 hour fermentation than any other period tested (Baumgardt and Hi Kon Ho, 1964). Baumgardt eta}. (1962) incorporated a standard forage in each fermentation trial to adjust for this variation. El-Shazly _e_t_a_l. (1960) evaluated three different types of apparatus differing in the complexity and mode of action. A simple all glass system devised by Bentley ital. (1954), a semipermeable membrane system suggested by Warner (1956), and a continuous flow system modified from Warner' s were tested. Only minor differences existed in the values obtained; therefore, the all glass system was preferred because of its simplicity. Simple, but efficient, techniques were advocated by Walker (1959). He stated that complexity was no criteria for obtaining agreement between digestibility values in in vitro and in vivo determinations. The small quantity of forage, and the large number of samples that can be included in a single fermentation are examples of the practical application of this system; however, standardization of grinding within the laboratory is important (Baumgardt and Hi Kon Ho, 1964). The i_n m digestibility of grasses was increased to a greater extent by fine grinding (60 mesh vs. 40 mesh) than alfalfa digestibility (Baumgardt and Hi Kon Ho, 1964). One fine— ness of grind had no advantage over any other, nevertheless, error occurred if forage samples ground to pass through a 40 mesh sieve were compared to a 60 mesh. The important role of solubility was pointed out by Donefer it _a_l_. (1963). As much as 29. 3% of mature alfalfa forage was water soluble; therefore, solubility differences among forages may be erroneously interpreted as differences in digestibility. Microbial digestion may also be influenced by the rate at which carbohydrates become available to the fermentation media (Allin- son, 1966). Digestibility and the Lignin- Cellulose Comjlex Forbes and Garrigus (1949), conducting a series of seven digestion trials with steers and wethers, found a straight line relationship between lignin content and dry matter digestibility. Highly significant negative correlations between lignin content and digestibility were shown by Sullivan (1955). This relationship was found between digestion coefficients obtained from sheep feces and chemical lignin analysis of the same forage. Tomlin f1 31; (1965) concluded that lignin content was negatively correlated with i_r_1_ vitro cellulose digestibility at the 12 hour period for both grasses and legumes; moreover, the regression equations were significantly different. As early as 1939 Crampton suggested that physical encrustation of the cellulose structure by lignin resulted in reduced digestibility. Dehority et 31. (1962) suggested a lignin carbohy- drate compound of plant cell wall material. It is known that lignin forms a dense membranous layer around the cell and even pene— trates within the microfibrils, chemically combining with the crystalline cellulose. Bolkei (1963) reported the first physical evidence that acetal or hemiacetal bonds existed between the car- bonyl groups of lignin and the hydroxyl groups of some portions of the holocellulose in wood and pulp lignin. Van Soest (1962) demonstrated that alfalfa had a higher lignin content than grasses of equal digestibility; however, grass lignin is more easily dissolved. Allinson (1966), using ultra- violet spectrophotometry, found different absorption Spectra for alfalfa clones of high and low nutritive value. The lignin polymer isolated from leaf and stem portions of timothy has different 10 functional groups (Stafford, 1962). Further studies by Stafford (1964) indicated that lignin in grasses was of two general types: One, an acid lignin polymer of ferulic acid and, two, a classical lignin polymer of coniferyl aldehyde or alcohol. Pigden and Heinrichs (1957), studying the lignin con- tent of six clonal lines of wheat grass, Agropyron intermedium, concluded that environmental factors affect lignin formation in plant tissue. Moisture was postulated as the primary environ- mental factor. Allinson (1966) found variability, both within and between genotypes, expressed as % DMD. MATERIALS AND METHODS The alfalfa clones used in this study were obtained from six populations. Clones with a different germplasm source and different types of management were grown from 1964 to 1966. Clones were harvested at the 1/2 bloom stage. Plant Sources and Management The unselected Tuna population consisted of 64 space- planted clones, randomly chosen from a larger pOpulation of 110 clones. These clones were grown in the field from 1964 through 1966. Only five were collected for testing from the first cutting, June 19, 1966. The second and third harvest of all 64 clones was made July 30 and September 14, 1966. Seeds from two unrelated insect and disease resistant pOpulations, MSA-C4 and MSB—C4, which had undergone fourteen cycles of recurrent phenotypic selection in eastern United States, were planted in the greenhouse April, 1966. 1 On June 19 all seed- lings were flowering. Forage from each plant was harvested 1MSA-C4 and MSB-C4 germplasm pools were released in 1965 by the Crops and Research Division, U. S. Dept. of Agri- culture, and the North Carolina Agricultural Experiment Station. 11 12 separately and then the plants were transplanted to the field. The first field harvest was made September 19, 1966. Ten Vernal, five DuPuits and eight Culver clones con- stituted the selected population. High and low % DMD clones were selected by Allinson in 1965 from a large source nursery of vari- eties. The selected plants were moved into the greenhouse during the winter of 1965-66. All clones were replanted in June, 1966. Harvests were made July 19 and September 13, 1966. Analysis Technique for Six Hour In Vitro DMD Forage samples were analyzed by the six hour in vitro technique outlined by Allinson (1966). This procedure was based on the work of Bowden and Church (1962a), Baumgardt _e_t a_1. (1962), and modifications by Ingalls (1964). Individual alfalfa clones were harvested by hand, the forage dried at 44 C. for 48 hours, and ground with a Wiley mill (40 mesh screen). Samples were analyzed immediately or stored in plastic bags in a cold chamber at 4 to 5 C. One g. duplicate samples from each clone were weighed into 125 ml. Erlenmeyer flasks to prepare for the fermentation. This was done one day before the actual test. The moisture con- tent of the sample was taken by normal procedure. 13 To each flask was added 20 ml. of a phosphate-urea- carbonate buffer. The buffer contained 2, 000 ml. water, 8. 2 g. monobasic potassium phosphate, 17. 4 g. dibasic sodium phOSphate, 4.0 g. urea, and 7. 4 g. monohydrate sodium carbonate. Adjust- ment was sometimes necessary with carbon dioxide to attain the 6. 8-6. 9 pH required. During this one hour time interval, rumen inoculum was collected from the cow in the following manner. One hour after morning feeding time and one hour before collection the Holstein cow was denied free access to food and water. Rumen fluid was removed by hand and strained through three layers of cheesecloth into a preheated thermos jug. The inoculum was placed in one liter vacuum flasks in the laboratory and maintained in a 39 C. water bath for 30 minutes. During the 30 minute interval, the fluid separated into two layers. The lower layer, containing the active portion of the mixture, was siphoned off and bubbled with carbon dioxide for three minutes. Fermentation was initiated by adding 24 ml. of rumen inoculum to each buffered forage sample. Carbon dioxide was immediately added to the flask and the flask was sealed with a Bunsen valve stopper. Flask, sample, buffer, and inoculum were placed into a 39 C. water bath for the fermentation period. The addition of four drops of 20% thymol, six hours later, stopped all microbial action. Flasks were refrigerated until the following day. 14 Filtration was accomplished with predried and weighed crucibles fitted with frittered glass and a layer of Solka Floc. Crucibles were dried in a forced draft oven for 36 hours at 80 C. and then weighed to determine the dry weight of forage remaining. The % DMD was determined by correcting for inoculum error and moisture. Forage Treatment and Antidote Procedure 1. Duplicate samples of forage T46 were treated four different ways 18 hours before fermentation. The treatments were applied to 1 g. forage samples as follows: (a) 6 ml. of 0. 5 N NaOH for 18 hours then neutralized before fermentation with HCL, (b) 6 ml. of 0. 5 N NaOH neutralized with HCL and 25 mg. of trypsin and 19 mg. pepsin added before fermentation, (c) 5 ml. of water plus 0. 1 ml. cellulase. The untreated (none) sample was analyzed by the standard DMD procedure. 2. Forage T46 pretreated 12 hours. Solutions were re- moved and antidotes added one hour before fermentation. This test consisted of treating 10 g. of T46 forage 12 hours before fer- mentation with (1) 100 ml. 0.5 N NaOH, (2) 60 mg. each of trypsin and pepsin plus 100 ml. of water, (3) 0. 15 ml. of cellulase and 100 ml. of water, (4) 100 ml. water. At the end of the pre- treatment period each sample was filtered and washed with distilled 15 water, dried at 60 C. , reweighed in 1 g. duplicate samples 24 hours later, and antidoted. The antidotes consisted of 50 mg. hydrostate casein, 50mg. of vitamin trace mineral mix, and 50 mg. of urea. Each of these substances was added immediately before the inoculum fluid. 3. Response of three Tuna forages differing in % DMD to two different NaOH pretreatments. The experiment consisted of two parts carried out simultaneously. All duplicate samples in the first group were pretreated for their Specified time period and then inoculated with rumen fluid. (1) 0. 5 N NaOH was added to forage samples of clones T95 and T108 and allowed to stand for 12 hours at room temperature before neutral- izing with HCL. Forage T46 was pretreated similarly for 18 hours. (2) These samples were pretreated the same as (1) except the neutralized samples were dried at 60 C. for 24 hours before fermen- tation. (3) No pretreatment employed, standard DMD procedure. Duplicate samples in the second group were pretreated in the same manner as the first group; however, no inoculum was added. This method was used to establish a correction factor to derive the actual % digested by the rumen fluid. Statistical Methods The fermentation variation from day to day and week to week was estimated in the DMD technique by including a control 16 sample in each fermentation. The mean value for these control samples was used to adjust the % DMD of each forage sample (Spragg, 1920). A partially nested analysis of variance was used to estimate the standard error of the mean of four, two, and one determinations per sample. Forage from five clones of three varieties was used for this analysis. On July 21, 1966, duplicate samples of the forage from these 15 clones were analyzed by the DMD procedure. Like- wise, August 4, 1966, the same forage samples were re-evaluated in a similar manner, giving a total of four determination values for each of the 15 clones. The standard error of the mean for four determina- tions was calculated, and estimates of the standard error of the mean of two and one determinations were obtained by changing the denomination. The proportional analysis of variance used on the Tuna population was explained by Snedecor (1956). This technique was employed because data was only available from one cutting in 1965 and two cuttings in 1966. The interaction term, years x clone (y) was used to estimate the environmental error. RESULTS Tuna Population Table 1 contains the means and ranges for the % DMD of each cutting and year. The % DMD range decreased from 1965 to 1966 and decreased further from July to September, 1966. A major % DMD increase occurred in the clone population between 1965 and 1966. This increase was evident in the June, 1966, harvest also, but only five clones of the total population were sampled. Another smaller increase in % DMD was observed between July and Septem- ber, 1966. The histograms, Figure 1, indicate the normality of the population and the decrease in % DMD range, and the increase in % DMD pOpulation mean over the three harvests. Table 1. -- 1965-66 cutting means and ranges of the % DMD values for 64 unselected Tuna clones. Harvest Date Variety No.a June, 1965b July 30, 1966 Sept. 14, 1966 Range Mean Range Mean Range Mean Tuna 64 22.8-40.6 32.3 33.3-50.5 40.6 37.7-49.2 42.7 aNumber of clones in study. b% DMD values by Allinson (1966). 17 18 % DMD June, 1965 H [\‘J I Number of Clones 0 /\ I I I I % DMD July 30, 1966 12— Number of Clones l I I % DMD September 14, 1966 12- Number of Clones o I I I I J I 20 25 30 35 4O 45 50 % DMD Figure 1. —- Histograms of the unselected Tuna clones for three harvests and two years. % DMD values used to measure the population. 19 Proportional analysis of variance was applied to this population. The average response for the two years was quite dif- ferent although no appropriate error term for making a test exists in this data. The relevant question is whether the clones tend to be ranked the same from year to year. Table 2 shows that the sums of squares for the year variance were very large. The F statistic calculated for the clone effects was significant at 1%, indicating real differences existed among the clonal means. The year x clone (y) interaction was significant at 6%. Variation in the % DMD of a given clone was influenced to some extent by factors causing this interaction. For example, of the four clones in the tOp 5% of the % DMD distribution in 1965, two were still in this group at the end of 1966; moreover, of the four clones in the bottom 5% of the dis- tribution, one clone was still in this group at the end of 1966. Table 2. -— PrOportional analysis of variance of the % DMD values in an unselected Tuna population. Source of Variance df SS MS F Total 191 5, 749. 61 Year 1 3, 753. 74 Cuttings (y) 1 141. 96 Clone 63 952. 08 15.11 2. sea Year x clone (y) 63 544. 41 8. 64 1. 52b Error 63 357. 42 5. 67 aSignificant at 1% level. bSignificant at 6% level. 20 Correlations between the % DMD for each of the Tuna clones, forage dry weight, and clone color are represented in Table 3. Clonal dry weight was not significantly correlated with % DMD. Significant positive correlations were obtained between the clonal dry weight of the July and September cuttings, and clone color and dry weight of the September cutting. Table 3. -- Correlation values between characteristics of the Tuna population in 1965-66. Correlations r Value June, 1965, Harvest % DMD vs. Dry Wt.a 0. 0211 July 30, 1966, Harvest % DMD vs. Dry Wt. 0.0971 September 14, 1966, Harvest % DMD vs. Dry Wt.° 0.1129 % DMD vs. Colorc’ d 0. 2201 Dry Wt. vs. Colorc’ d 0. 7569e July 30 Harvest and September 14 Harvest Dry Wt. vs. Dry Wt.b 0. 6195‘e a110 Clone comparisons (Allinson, 1966). b64 Clone comparisons. C20 Clone comparisons. dColor rating 10 green, 1 yellow 63Significant at 1% level. 21 MSA-C4 and MSB-C4 Populations These two populations were chosen for evaluation be- cause of their insect and disease resistant qualities and‘their exceptional dark green foliage. Table 4 contains the means and ranges for the % DMD of the plants for both harvests, as well as simple correlations between % DMD, dry weight, and color. Table 4. -- Means and ranges for the % DMD of MSA-C4 and MSB-C4 populations. Simple correlations between % DMD, dry weight, color, and harvests. Population Number of Clones Mean Range June Greenhouse Harvest MSA-C4 40 35. 9 30. 0-44. 2 MSB-C4 54 36. 6 28. 1-46. 3 September Field Harvest MSA-C4 46 41. 3 35. 5-47. 9 MSB-C4 70 45.3 38.7-51.6 POpulation Correlations r Value MSA-C4 % DMD and Forage Weight -0. 2278 % DMD and Forage Colora -0. 2545 % DMD Greenhouse and Field 0.1924 MSB-C4 % DMD and Forage Weight 0.0075 % DMD and Forage Colora 0. 1669 % DMD Greenhouse and Field 0. 1078 aColor rating-~10 green, 1 yellow. 22 The populations were near normal in their distribu- tions, Figure 2, but differed in the range and mean of their % DMD from greenhouse to field; nevertheless, the % DMD values for the September field cutting of both populations were similar to those of the Tuna population harvested in September. Even though the green- house plants were only twelve to sixteen inches high when harvested, with a total dry weight of a few grams, their mean % DMD was 8. 7-5. 4% lower. All correlations were non-significant, but the correla- tion values calculated between % DMD, dry weight, and foliage color differed in sign from population to population. The insignificance of their value obtained between the % DMD of the greenhouse seed- lings and the field clones needs further verification in 1967. Vernal, DuPuits, and Culver Clones This population consisted of 10 Vernal, 8 Culver, and 5 DuPuits clones. Table 5 contains the range and the mean % DMD values for the selected high and low clones of each variety. Changes in the mean values and ranges from year to year and cutting to cutting are evident; however, the clones in the low selected group made the greatest change from 1965 to 1966. This is illustrated graphically in Figure 3. The % DMD values of all clones in 1966 were grouped together in the high range; in addition, the range m % DMD MSB-C4 Field Cutting September 19, 1966 g 12 _. 2 U 9 — 90-! o r. 6 _ o '2 :s 3 5' Z 0 l I l l m % DMD MSB-C4 Greenhouse Cutting June 19, 1966 8 12 - o 5 9 _ c... o s 6 - <0 .0 E 3 — 3 Z O m 12 % DMD MSA- C4 Field Cutting Septembler 19,1966 0:) .— 2 o 9 _ Q4 0 h 6 _ /\/ <0 ‘2? :3 3 ’ Z 0 l I I m 12 % DMD MSA- C4 Greenhouse Cutting June 19,1966 91’ 2 o 9 - CH 0 5.. 6 - k <0 2 5 3 - . Z 0 I L I I 25 30 35 40 45 % DMD Figure 2. -— Histograms of the MSA-C4 and MSB—C4 field and greenhouse populations. % DMD values used to measure the population. U1 03 00 II}- Number of Clones [\3 030 H 0" Number of Clones Number of Clones 24 A‘ % DMD June, 1965 IVE M m . . % DMD July 19, 1966 I I I I I I I % DMD September 13, 1966 . I I I I I I I LA I 25 35 40 45 50 55 % DMD Figure 3. -- Histograms of the selected clones for three harvests and two years. % DMD values used to measure the popula- tion. among the clones decreases. 25 Clonal changes between cuttings in 1966 were rather small in comparison to the year to year changes. Table 5. -- Range, mean, and number of clones in each variety Varieties divided further into sub-groups high and low. Values expressed as % DMD. represented in the selected population. Harvest Date Variety No.a June, 1965b July 19, 1966 Sept. 13, 1966 Range Mean Range Mean Range Mean Vernal High 6 34.3-40. 38.4 43.5-50. 46.5 41.3-49.6 46.1 Low 4 23.4-26. 25.3 43.3-46. 44.5 38.1-47.4 42.0 DuPuits High 2 44.2-45. 44.6 40.2-42. 41.3 44.7-42.6 43.7 Low 27.0-29. 27.9 40.0-40. 40.3 40.6-43.5 41.9 Culver High 3 37.9-40. 38.8 41.2-44. 42.9 44.1-45.8 43.8 Low 5 23.6-25. 24.4 42.5-44. 43.4 42.6-48.7 46.2 POpulation 23.4-45. 33.2 40.0-50. 43.2 38.1-49.6 44.0 aNumber of clones in each sub-group. b% DMD values by Allinson (1966). Forage Treatment and DMD Forage from alfalfa clone T46 was treated with two enzyme preparations: (1) cellulase or (2) a combination of trypsin 26 and pepsin. The % DMD was increased 3. 1% and 4. 8% respectively, Table 6. NaOH alone gave an 8. 0% increase. Another slight in- crease was attained by adding trypsin and pepsin immediately before fermentation. Table 6. -— % DMD for alfalfa forage T46 pretreated 18 hours. Pretreatmenta % DMD % Increase in DMD NaOH 0.5N 47.4 8.0 NaOH.O'5N . 48.1 8.7 Trypsm, Pepsm Cellulase 42. 5 3. 1 Trypsin, Pepsin 44.2 4. 8 Noneb 39. 4 -—- aConcentrations and procedures listed in materials and methods. b No pretreatment. To determine the relative digestibility of the insoluble portion of forage T46 it was again pretreated 12 hours with .5N NaOH, cellulase, trypsin, pepsin, and water. All soluble sub- stances were removed by filtration and the remaining portion was dried before rumen fermentation. Table 7 contains the results. The organic and inorganic nitrogen sources gave near equal increases to the residue of each respective pretreatment. 27 The vitamin, trace mineral mix lowered the % DMD in every case except the NaOH pretreated sample; in addition, the NaOH pre- treated sample had the lowest % DMD without the antidotes, 10. 5%. When urea was added the % DMD increased to 19. 6%. This is higher than any of the water soluble residue values with or without antidotes. The insoluble portion of the trypsin-pepsin pretreat- ment had the highest % DMD values, with only minor antidote effects. Table 7. —— Forage T46 pretreated 12 hours. Solution removed and antidotes added one hour before fermentation. Values expressed as % DMD. 12 Hour Pretreatment Solution Removed Antidote (0. 5N NaOH) Try p51“ Cellulase H 0 Average Pepsm 2 Hydr98tat 18.0 27.6 21.1 18.4 21.3 Caseln Vltamm. 11.4 23. 3 15.8 15.0 16.4 Trace Mln. Urea 19.6 26.6 21.7 18.3 21.3 Nonea 10. 5 26. 8 20. 2 17.9 18. 9 Average 14.9 26.3 19.7 17.4 ---- aNo antidote added. The NaOH pretreatment was selected for further study on three forage samples with different untreated % DMD values. 28 The untreated % DMD values, the % soluble in water, and the actual % digested by the rumen microorganisms are found in Table 8. These three forages not only differed-in their % DMD but in the other values. Forage T95 had the highest water soluble portion and T46 the lowest, but T108, the intermediate clone, had the highest percent of its insoluble portion digested by the rumen inoculum. Table 8. -- The % DMD, % water soluble, and % microbial digestion of three Tuna forages: T95, T108, and T46. Clone % DMD % Soluble % Digesteda T95 49.2 34.6 14.6 T108 45.7 28.3 17.4 T46 39.4 27.0 12.4 3% DMD minus % water soluble. The effect of NaOH pretreatment on these three Tuna forages is found in the first half of Table 9. The second portion of the table was calculated by subtracting the untreated values for each forage (Table 8) from the treated (Table 9). All samples increased in % DMD and % soluble when treated with NaOH. The calculated % DMD increased 6. 5 to 8. 0 as the treated forage % DMD decreased 55. 7 to 47. 4; moreover, as the calculated solubility 29 values increased 6. 2 to 14. 9, the calculated percent digested by the rumen inoculum decreased 0. 3 to -6. 9. Forage T46 had the lowest treated % DMD, the greatest increase in calculated solu- bility, and the greatest decrease in calculated digestibility. Clone T95 was exactly the opposite and T108 was intermediate. Table 9. -- The % DMD, % water soluble, and % microbial diges- tion of three Tuna forages pretreated with NaOH. Calculated values derived by subtracting the untreated values (Table 8) from the treated values. Calculated Values Treated Values Treated Minus Untreated Clone % DMD % Soluble % Digesteda % DMDb % Solublec % Digestedd T95 55.7 40.8 14.9 6.5 6.2 0.3 T108 53.2 36.9 16.3 7.5 8.6 -1.1 T46 47.4 41.9 5.5 8.0 14.9 -6.9 aTreated % DMD minus treated % soluble. bTreated % DMD minus untreated % DMD. CTreated % soluble minus untreated % soluble. d . . . Remainder of b mlnus remalnder of c. A somewhat different trend occurred when the sample treated with NaOH was dried and then exposed to fermentation, Table 10. Another increase in % DMD occurred for each sample with a corresponding increase in solubility; however, both the 30 treated in % digested and the calculated were lower than in the undried NaOH treatments. Table 10. -- The % DMD, % water soluble, and % microbial diges- tion of three Tuna forages pretreated with NaOH and dried before fermentation. Calculated values derived by subtracting the untreated (Table 8) from the treated values. Calculated Values Treated Values Treated Minus Untreated Clone % DMD % Soluble % Digesteda % DMDb % SolubleC % Digestedd T95 62.4 55.3 7.1 13.2 20.7 -7.5 T108 58.5 48.4 10.1 12.8 20.1 -7.3 T46 49.3 ---- ---- 9.9 -——— ———- aTreated % DMD minus treated % soluble. bTreated % DMD minus untreated % DMD. CTreated % soluble minus untreated % soluble. Remainder of b minus remainder of c. DMD Pilot Study To determine the relative importance of clone and determination error variance, a partially nested analysis of vari- ance was employed, Table 11. The variety main effect was not significant. The mean square value for the interaction term (block x clone [v]) was very small in comparison to the clone 31 mean square; moreover, it was noticeably smaller than the deter- mination mean square. Legitimate pooling of the sums of squares for the interaction term and the determination was therefore pos- sible, Formula (1). Table 11. -- Partially nested analysis of variance table of DMD sampling pilot study. Source of Variance df SS MS F Total 59 289. 010 Block 1 3. 504 Variety 2 85. 317 42.569 4.022 Block x Variety 2 21. 212 10. 606 Clone (v) 12 155. 051 12. 921 22. 6683 Time x Clone (v) 12 2. 291 0. 570 (1) Determinations 30 21. 635 aSignificant at 1%. SS1 + SS 2 p df1 + df2 (1) S (2) s- a/_9-_5_7£_ z 0. x 4 (3) s- =\/__9_-_5_7£_ = o, x 2 (4) S- a/fl = o, x 1 0.570 378 534 755 32 The calculated pooled variance was 0. 570 with 42 degrees of freedom. The standard error of the clone pooled mean for four determinations per clone was calculated to be 0. 378, Formula (2). The estimated standard error of the mean of two determinations was 0. 534 and one determination 0. 755, Formula 3 and 4. Apply- ing these calculated values to Tukey' s multiple range test, any two means differing by more than 1. 90% would be judged significant at the 5% level with four determinations, in contrast to 3. 77% if only one determination was made. DISCUSSION The four populations of alfalfa clones analyzed by the E1 vitro DMD technique approached normal distribution. Curves of this type are representative of quantitative genetic characters, with continuous variation. Since this bioassay technique is responsive to the solubility and digestibility of the lignin-cellulose complex (Halliwell and Bryant, 1963), the physiological develop- ment of this structural arrangement is possibly a multi—gene reaction. Hopefully, some degree of genetic control over this system may be possible which will permit an improvement in the overall nutritional value of alfalfa forage. In order to make genetic advance by selection, two population parameters must be known or estimated: (1) The amount of genetic variability in a population and (2) the masking effect of environment or the interaction components of Variation. Agronomic factors which may be directly or indirectly affected by selection and development of plants with i_n vitro tech- nique should also be considered. Alfalfa varieties with high quality combined with low forage yield would be of little use agriculturally; 33 34 likewise, low forage quality together with high insect and disease resistance would certainly not be agronomically acceptable. Unselected Tuna Population The DMD analysis of the Tuna population demonstrated three things: (1) Some environmental interaction does occur and must be considered; (2) Alfalfa clones should be analyzed over a period of at least two years with more than one cutting per year; (3) Yield and % DMD are not necessarily related. The environmental influence was clearly shown by the increase in all % DMD values from 1965-1966. This effect is prob- ably a combination of developments. A number of hypotheses can be offered to explain the reaction. Even though the same Holstein cow provided the rumen inoculum for all forage samples during both years of study, some physiological or microbial change within the cow could have occurred between the years. This view is supported to some degree in the literature, Baumgardt and Hi Kon Ho (1964). The second hypothesis, changes in the physiological de- velopment of the plant, is more likely. Pigden and Heinrichs (1957), working with lignin content in wheatgrass, concluded that environ- mental changes could alter the formation of the lignin polymer. Allinson (1966) demonstrated that the % DMD can even vary within plants of the same genotype; therefore, with the potential buffering 35 capacity of the heterozygous genotype of alfalfa plants, it is entirely possible that the response of a clone to DMD analysis could vary considerably from one environmental situation to another. Pigden and Heinrichs (1957) concluded further that moisture was the primary factor reSponsible for the environmental influence. Differences did exist in amount and distribution of rain- fall from 1965 to 1966. Low moisture and dry conditions existed in 1965, with correSponding low % DMD values, and in 1966, with more normal moisture, clones had higher % DMD values. Assuming the difference measured in % DMD for a clone over the two years and three cuttings was due primarily to environmental influence, the magnitude of its importance was estimated by the significance of the year x clone (y) interaction component, Table 2. The F. Value of the 1. 53 was within .01 units of being significant at the 5% level, whereas the main effect of clone differences was significant at 1% level. The lack of a strong year x clone interaction leads to the conclusion that the clone means are a good indication of the genetic population for % DMD. Selection of quantitative characteristics in perennial CI‘OpS is best conducted over a period of years and a number of cuttings within a year. The additional information of years and cuttings is needed to more accurately estimate the plant' 8 inherent stability under different environmental conditions. 36 The agronomic character of yield was compared to the % DMD values for the two cuttings in 1966. Allinson (1966) found a significant negative correlation between % DMD and yield of some varieties. The correlations in this study were low and approached zero; therefore, the relationship of % DMD to yield was small or nonexistent. Disease and Insect Resistant MSA-C4 and MSB-C4 Population The possibility of combining insect and disease resis- tance with high nutritional value was the primary reason for evalu— ating the MSA-C4 and MSB-C4 populations. The first analysis of these clones indicates that they are equal and perhaps superior to present varieties. The mean % DMD for the MSA-C4 clones was higher than any other group of clones tested. Any relationship between yield and % DMD could not be demonstrated in either pop- ulation. The attempt to predict field % DMD from greenhouse % DMD data shows little promise within the same year. Due to the low values achieved for these young greenhouse seedlings and the apparent distortion of the % DMD response of the twenty-three selected clones, DMD evaluation of alfalfa plants under greenhouse conditions must be viewed with caution. Further study is needed to gain a better understanding of the phenomena. 37 Selected Clone Population The twenty-three clones representing three varieties in the selected pOpulation were separated into two distinct groups in 1965, Figure 3. After placing these clones under an abnormal winter environment in the greenhouse (18 hour photoperiod and 70-80 F. temperature), the population distribution curves for the July and September, 1966, field harvest again approached normality. Such results may occur because of the alteration of the physiological processes Within the plant distorting any re-evalua- tion or comparison the following year. As many as 63% of the clones could have been falsely classified, since their selection was on a single harvest basis. The Solubility and Digestibility of Treated Forage Chemical and biological treatment of forage has not been used to increase its efficiency by ruminants. If accomplished, it would have the advantage of immediate application, providing beneficial animal response could be economically demonstrated. Additional benefits from this approach would be knowledge about the factors in forage which limit or stimulate animal usage. Forages pretreated with enzymes would be expected to increase forage breakdown, thereby increasing its % DMD. 38 Likewise NaOH, which has had extensive use as a lignin solvent (Stafford, 1962), should increase the % DMD of forage. These facts were verified in the initial experiment, Table 6. Values similar to these were reported by Donefer e_t 11. (1963). Antidoting the insoluble portion of the forage with organic and inorganic nitro— ‘gen increased the digestibility in every case. The most impressive increase occurred when urea was added to the insoluble portion of the NaOH pretreatment, Table 7. Forage solubility and microbial digestibility are com- bined in the % DMD values. Emphasis in most DMD studies is placed on microbial digestibility; however, the importance of the solubility factor can be seen in Table 8. Forage from clone T95 had the highest % DMD value but not the highest microbial diges- tion. Treatment with NaOH (Table 9) caused increases in all % DMD values in proportion to the untreated values; however, this increase was again due to solubility since the percent digested actually de- creased. The greatest digestion reduction was exhibited by the forage with the lowest untreated % DMD value. Evaporation of the water from the NaOH treated sample to resemble a dried forage caused further changes in solubility and digestion. The solubility again increased in all samples with a corresponding decrease in microbial digestion. 39 Increasing forage solubility may be important in overall animal nutrition even though rumen digestion by the microorganism may be reduced. If the increased availability of these nutrient sub- stances would increase their assimilation into the animal' 8 digestive system, pretreatment to increase solubility could be beneficial. Suggested Changes in the DMD Mass Selection Technique The application of the six hour DMD technique to mass selection of forage may be a valuable innovation. However, with large numbers of samples requiring analysis, any reduction in the sample number would be a worthwhile improvement. Allinson (1966) suggested that since the variance between fermentations was greater than the variance within fermentation, samples should be evaluated in as many fermentations as possible. Duplication would be eliminated within fermentation. This would increase the error variance accuracy for a clone, but it would not reduce the number of samples. The standard error of a clone mean using two determina- tions was 0.534. This value is comparable to those reported by Barnes (1965), and Allinson (1966). With the relatively small standard deviation for a single determination, 0. 755, and the range in % DMD values reported in the preceding studies, identification of the upper 5% or 10% of the clones could be adequately III I All-ll. .IIII'III 40 accomplished by one determination. The number of clones evaluated in a mass selection study is of much greater importance than the slight misclassification of a few. This would be eSpecially valuable in analyzing perennials with multiple harvests. SUMMARY AND CONCLUSIONS Four alfalfa populations have been evaluated by the six hour .19. 11119 technique. The unselected Tuna population yielded information about the environmental influence upon in-place field clones. The MSA-C4 and MSB-C4 populations indicated the possi- bilities of selecting superior clones within insect and disease limi- tations, and the selected population demonstrated the effects of abnormal environment on DMD re-evaluation. Studies were conducted on the solubility and digesti— bility of chemically or biologically treated forage. Finally, a change was suggested in the application of the six hour in vitro procedure to a preliminary forage screening technique. The fol- lowing conclusions may be stated. 1. Environmental influence and interaction does alter the DMD values of a Space planted alfalfa population; how— ever, selection of superior clones should be possible if the population is analyzed for more than one year and one cutting per year. 2. The agronomic characters of yield and disease and insect resistance are not necessarily related to DMD 41 42 in all populations. The genetic system responsible for these characters appears to be independent. Alfalfa clones moved from the field to the greenhouse over the winter and re-evaluated in the field the fol- lowing year were inconsistent in their DMD response. Abnormal environmental stress placed upon the plant for a prolonged period of time may require a period of adjustment before valid DMD comparisons can be made. Forage solubility can be increased by chemical or biological treatment; however, its importance in animal nutrition has not been established. Preliminary screening of large numbers of individual plants can be effectively carried out with the DMD technique by making only one determination per sample. This allows more clones, cuttings, and years to be evaluated with the same resources. BIB LIOG RAPHY Allinson, D. W. 1966. Factors affecting the nutritive value of forages. Ph.D. Thesis, Michigan State University. Atwood, S. S., and J. T. Sullivan. 1943. Inheritance ofa cyano- genetic glucoside and its hydrolyzing enzyme in Trifolium repens. J. Hered. 34:311-320. Ayala, E., and E. L. Johnson. 1951. Vitamin B12 and growth inhibiting properties of dried juice of alfalfa. Poul. Sci. 30:893-899. Barnes, R. F., G. O. Mott, L. V. Packett, and M. P. Plumlee. 1964. Comparison of i_n vitro fermentation methods. J. Ani. Sci. 23:1061-1065. Barnes, R. F. 1965. Use of i_n vitro fermentation techniques for estimating forage digestibility and intake. Agron. J. 57:213-216. Baumgardt, B. R., J. L. Cason, and M. W. Taylor. 1962. Evaluation of forages in the laboratory. 1. Compara- tive accuracy of several methods. J. Dairy Sci. 45:59- 61. Baumgardt, B. R., and D. Smith. 1962. Changes in estimated nutritive value of herbage of alfalfa, medium red clover, ladino clover, and bromegrass. Ag. Exp. Sta., Univ. Wisc., Res. Rep. 10. Baumgardt, B. R. , and Hi Kon Ho. 1964. Evaluation of forages in the laboratory. IV. Within and among trial variability of the Wisconsin artificial rumen procedure. J. Dairy Sci. 47:263-266. Bentley, O. G., R. R. Johnson, S. Vanecko, and C. H. Hunt. 1954. Studies on factors needed by rumen microorgan- isms for cellulose digestion i_r_l vitro. J. Ani. Sci. 13:581-593. 43 44 Bolker, H. I. 1963. A lignin carbohydrate bond as revealed by infra-red spectrosc0py. Nature 197:489-490. Bolton, J. L. 1962. Alfalfa. Interscience Pub., Inc, N. Y. Bolton, J. J., and M. W. Cormack. 1953. Breeding alfalfa in Canada. Emp. J. Exp. Agri. 21:242-254. Bowden, D. M., and D. C. Church. 1962. Artificial rumen investigations. 1. Variability of dry matter cellulose digestibility and production of volatile fatty acids. J. Dairy Sci. 45:972-979. 1962. Artificial rumen investigation. 11. Correla- tions between 1_n_ vitro and in vivo measures of digesti- bility and chemical comporTe-nts of forages. J. Dairy Sci. 45:980-985. Crampton, E. W. 1956. Applied animal nutrition; the use of feed- stuffs in the formulation of livestock rations. W. H. Freeman, San Francisco, 458pp. Crampton, E. W., and L. A. Maynard. 1938. The relation of cellulose and lignin content to the nutrition value of animal feeds. J. Nutri. 15:383-395. Dehority, B. A., R. R. Johnson, and H. R. Conrad. 1962. Digestibility of forage hemicellulose and pectin by rumen bacteria in vitro and the effects of lignification thereon. J. Dairy—S-ci. 45:508-512. Donefer, E., P. J. Niemann, E. W. Crampton, and L. E. Lloyd. 1963. Dry matter disappearances by enzyme and aqueous solutions to predict the nutritive value of for- ages. J. Dairy Sci. 46:965-970. Elliott, F. C. 1963. The meadow vole (Microtus pennsylvanicus) as a bioassay test organism for individual forage plants. Quart. Bull. Mich. Agric. Exp. Sta. 46:58-72. El-Shazly, D., B. A. Dehority, and R. R. Johnson. 1960. A comparison of the all-glass semipermeable membrane and continuous flow types of apparatus for in vitro rumen fermentations. J. Dairy Sci. 43:1445-145T. 45 Forbes, R. M., and W. P. Garrigus. 1948. Application ofa lignin ratio technique to the determination of the nutrient intake of grazing animals. J. Ani. Sci. 7:373-382. Gil, H. D., R. L. Davis, and R. F. Barnes. 1967. Inheritance of i_n- vitro digestibility and associated characters in Medicago sativa L. Crop Science 7:19-21. Halliwell, G., and M. P. Bryant. 1963. The cellulolytic activity of pure strains of bacteria from the rumen of cattle. J. Gen. Microbiol. 32:441-448. Hanson, C. H., G. O. Kohler, J. M. Dudley, E. L. Sorenson, G. R. Vonatta, K. W. Taylor, M. W. Pederson, H. L. Carnahan, C. P. Wilsie, W. R. Kehr, G. C. Lowe, E. W. Stanford, and J. A. Yungen. 1963. Saponin content of alfalfa as related to location, cutting, and variety. ARS, USDA, ARS. 34-44. Hartman, G. H., W. R. Akeson, and M. A. Stahmann. 1967. Leaf protein concentrate prepared by Spray-drying. Agri. Food Chem. 15:74-79. Hershberger, T. V., T. A. Land, E. W. Hartsook, and R. W. Swift. 1959. Use of artificial rumen technique to estimate the nutritive value of forages. J. Ani. Sci. 18:770-779. Ingalls, J. R. 1964. Nutritive value of several forage Species as measured by i_n vitro and i_n vivo methods. Ph. D. Thesis, Michigan State University. Jacobson, D. R., W. M. Miller, D. M. Leath, S. G. Yates, H. L. Tookey, and I. A. Wolf. 1963. Nature of fescue toxicity and progress toward identification of the toxic entity. J. Dairy Sci. 46:416. Maynard, L. A., and J. K. Loosli. 1956. Animal nutrition. McGraw-Hill, N. Y., London, and Toronto. 484pp. Morrison, F. B. 1956. Feeds and feeding 22nd edn. Morrison Publ., Ithaca, N. Y. 1165pp. 46 Oldfield, J. C., C. W. Fox and E. M. Bickoff. 1960. Effects of estrogenic activity in alfalfa on growing lambs. J. Ani. Sci. 19:1281. Pedersen, M. W., D. E. Zimmer, J. 0. Anderson, and C. F. McQuire. 1963. A comparison of saponin from DuPuits, Lahontan, Ranger, and Uinta alfalfas. Pigden, W. J., and D. H. Heinrichs. 1957. The lignin content of six clonal lines of intermediate wheatgrass. J. Ani. Sci. 37:128-131. Pudelkiewicz, W. J., and L. D. Matterson. 1960. A fat soluble material in alfalfa that reduces the biological availability of toc0pherol. J. Nutri. 71:143-148. Schillinger, J. A., and F. C. Elliott. 1966. Bioassays for nutri- tive value of individual alfalfa plants. Quart. Bull Mich. Agri. Exp. Sta. 48:580-590. Singleton, V. L., E. T. Mertz, and R. L. Davis. 1952. The hydrolysis of amino acid assay of alfalfa, and the methi- onine range in 100 selections. Agron. J. 44:346-348. Snedecor, G. W. 1956. Statistical methods. Iowa State College Press, Ames, Iowa. Spragg, F. A. 1920. The coefficient of yield. Agron. J. 12:168- 174. Stafford, H. A. 1962. Histochemical and biochemical differences between lignin-like material in Phleum pratense L. Plant Physiol. 37:643-649. 1964. Comparison of lignin-like products found natu- rally or induced in tissue of Phleum, Elodea, and Coleus, and in a paper peroxidase system. Plant Physiology 39:350-360. Sullivan, J T. 1955. Cellulose and lignin in forage grasses and their digestion coefficients. J. Ani. Sci. 14:710-717. Tisdale, S. L., R. L. Davis, A. F. Kingsly, and E. T. Mertz. 1950. Methionine and cystine content of two strains of alfalfa as influenced by different concentrations of sul- phate ion. Agron. J. 42:221-225. Towlin, D. 47 C., R. R. Johnson, and B. A. Dehority. 1956. Relationship of lignification to in vitro cellulose digestibility of grasses and legu_mes. J. Ani. Sci. 24:161-165. Van Soest, P. J. 1962. The estimation of forage protein digesti- Walker, D . Warner, A. bility and determination of the effects of heat-drying upon forages by means of the nitrogen content of acid- detergent fiber. Am. Dairy Sci. Assoc. , Paper 3. M. 1959. The in Vitro digestion of roughage dry mat- ter. J. Agr. Sci. 53:192. C. I. 1956. Criteria for establishing the validity of i_n vitro studies with rumen microorganisms in so- called artificial rumen systems. J. Gen. Microbiol. 14:733. "IIIIIIIIIIIIIIIIIIIIII