Effect of Lint on Whole Cottonseed Passage and Digestibility and Diet Choice on Intake of Whole Cottonseed by Holstein Cows1

Effect of Lint on Whole Cottonseed Passage and Digestibility and Diet Choice on Intake of Whole Cottonseed by Holstein Cows 1 C, E. COPPOCK, J. R. MOYA, J. W. WEST, D. H. NAVE, and J. M. LABORE Department of Animal Science C, E. GATES Institute of Statistics Texas A&M University College Station 77843

ABSTRACT

Effects of rolling whole linted cottonseed and whole acid-delinted cottonseed on whole seed passage and digestibility were determined using 24 lactating Holstein cows. Whole seed passage averaged .74% in all cows fed whole linted seed during the standardization period and .45% in 6 cows fed whole linted seed during a comparison period, contrasted to 11.3% in 6 cows fed acid-delinted seed. Digestibility of ether extract was less in the 6 cows fed whole acid-delinted cottonseed. No advantage was for rolling whole linted cottonseed before feeding, but roiling was beneficial for whole aciddelinted cottonseed. Individual consumption of whole linted cottonseed by 18 nonpregnant, dry Holstein cows varied from .04 to 5.05 kg/day when two-choice options were given for cottonseed versus chopped coastal bermudagrass hay or corn silage, or a complete ration of 75% corn silage and 25% concentrate. No indication of gossypol toxicity was seen, but the large variation in each choice situation suggests it is desirable to blend cottonseed with other ingredients to ensure more uniform consumption. INTRODUCTION

Whole cottonseed (WCS) has been used as a cattle feed for many years, particularly in areas

Received September 10, 1984. 1Technical article 19869 by the Texas Agricultural Experiment Station, College Station 77843. 1985 J Daily Sci 68:1198--1206

where cotton is grown. Early work (16) compared the feeding value of WCS with an equal amount of a mixture of 57% corn and 43% cottonseed meal. This comparison equated only crude protein, but WCS contains 20% more net energy of lactation (NE I) than cottonseed meal and 10% more than corn (14). Its high energy and highly digestible fiber apparently were not considered earlier, probably because lowproducing cows did not respond to additional energy, and milk fat maintenance was not an issue. More recent resports (17, 18) of increases of milk yield and milk fat test have renewed interest by both researchers and dairy producers, resulting in shipment of WCS to dairies in northeast US and eastern Canada. Whole linted cottonseed (WLCS) is a paradox among feedstuffs, having both high energy and high fiber. The fiber is composed in part of lint (about 10% by weight of the whole seed), which is nearly pure cellulose and highly digestible. The energy reflects the oil in the seed, and if oil (fat) also has a low heat increment in ruminants, WCS would be more valuable in hot weather when feed consumption, milk yield, and milk fat test may be depressed. Palmquist and Jenkins (15) suggested 3 to 5% added fat to increase energy density and energy intake by high-producing cows without reducing fiber in the diet. Use of WCS to add oil increases both energy and fiber. Most dairy producers do not process WCS prior to feeding. Morrison (11) recommended grinding whole shelled corn for dairy cattle to avoid from 18 to 35% of the whole grain passing through the digestive tract undigested. Grinding is the most widely used method of processing feed grains, and nearly all grains are processed prior to feeding to dairy cows (21). Many dairy producers continue use of conven-

1198

WHOLE COTTONSEED PASSAGE AND DIGESTIBILITY tional feeding systems of hay and pasture with concentrate fed primarily in the milking parlor in lieu of newer systems such as the complete ration (CR). The latter is better adapted to forage programs based on silage. For those with CR systems, WCS can be added to the mixture; for those with conventional systems, the issue is how to feed WCS. It would be convenient to feed it separately and free choice if intakes were predictable and safe. Feed dealers have the problem of incorporating a light lint-coated seed that may not require processing but which does not flow readily or remain dispersed with conventional ingredients. Processing also may increase susceptibility to rancidity in hot weather. MATERIALS AND METHODS Experiment 1

Twenty-eight multiparous lactating Holstein cows in the first half of lactation were assigned to a 14-day standardization period (SP) and fed a control diet with dry matter (DM) of 35% corn silage (CS), 15% WLCS, and 50% concentrate (Table 1). The ration was fed as a CR, which was similar to the herd diet they had been eating. Cows were fed individually twice daily (4 h access ad libitum each feeding) and at other times were handled as one group in a dry lot with water and free stalls except for

TABLE 1. Ingredient composition of concentrate mixture, Experiment 1. Ingredients (%) Ground corn Cottonseed meal Defluorinated phosphate Chromium oxide premix I Limestone Plain salt Trace element premix Vitamin A premix 2

64.30 30.00 .40 1.00 3.70 .45 .10 .05 100.00

1Composed of 75% ground shelled corn and 25% Cr20~. Added 10 million IU of vitamin A/metric ton.

1199

TABLE 2. Schedule for fecal grab sampling, Experiment 1. Days 12 26

13 27

14 28

(h) 2400 0600 1200 1800

0200 0800 1400 2000

0400 1000 1600 2200

milking twice daily in a milking parlor. Body weights were taken on days 6 and 18. During days 1 through 11, feed was offered ad libitum and individual feed consumption was measured. One hundred percent of the average feed eaten during days 7 through 11 was offered on days 12, 13, and 14, in which four fecal grab samples per day were taken at 6-h intervals with a 2-h advance at the beginning of days 13 and 14 (Table 2). The 12 fecal grab samples were dried at 55°C to constant weight, air equilibrated to atmospheric moisture, and combined on an equal weight basis as representative of the fecal excretion during this period. All cows were continued on the control diet ad libitum through days 15 to 18, and blocked according to intake of feed DM as a percentage of body weight into quartets and assigned randomly within blocks to one of four treatments. Four cows were dropped at this stage, three for mastitis, and one from off-feed. Treatments were four forms of cottonseed: 1) WLCS, 2) rolled linted cottonseed (RLCS), 3) whole acid-delinted cottonseed (WACS), and 4) rolled acid-delinted cottonseed (RACS), all fed at 15% of diet DM of complete rations. The same concentrate mixture (Table 1) fed in the SP was used in all diets and it contained .25% chromic oxide as an indigestible marker (Table 1) for estimating digestibility. The diets were fed for ad libitum intake from days 19 through 25. Feed offered on days 26 to 28 was restricted to 100% of that eaten during days 21 to 25. Fecal grab samples were taken during days 26 to 28 according to Table 2 and processed as described for the SP. Individual feed consumption was measured during days 7 to 28. Individual feeds and complete rations were sampled on days 8 to 12 of Journal of Dairy Science Vol. 68, No. 5, 1985

1200

COPPOCK ET AL.

the SP and days 22 to 26 of the treatment comparison period. Samples were dried as described for feces and composited by period prior to grinding and chemical analyses. Chemical analyses of feces and feeds included crude protein by Kjeldahl and Technicon, and fat as ether extracted for 16 h for feces and feeds without cottonseed, and 96 h for cottonseeds and complete rations. Acid detergent fiber (ADF) was analyzed according to Goering and Van Soest (7). For starch analyses we used .2 g samples mixed with 20 ml distilled water and autoclaved for 2.5 h to gelatinize the starch. After dilution with 25 ml .1 N sodium acetate buffer (pH 4.5) and addition of I ml Diazyme L100, samples were incubated 30 min at 60 C to hydrolyze starch to glucose. Samples were diluted to 50 ml and analyzed by Technicon Autoanalyzer with bound hexokinase (19). Calcium and magnesium were by atomic absorption (1) and chromium was b y the spectrophotometric method as modified in (8). Eight additional mineral elements were analyzed by the New York Dairy Herd Improvement (NYDHIC) Forage Testing Laboratory (3). During days 12 to 14, one 5-kg sample of wet feces from each cow was screened (3-mm screen) to separate whole seeds, which were counted, dried, weighed, and processed as feed samples. The same screening procedure was used for 5-kg wet fecal samples from the 6 cows in treatment groups 1 and 3 during days 26 to 28. The model used to compare digestibilities was

I

i

.e

8 "-d 8

:

Yijk =/~ + Bi + Tj + BiSijk + eijk, where/~ represented a mean, B, blocks; T, treatment; BiSijk, a covariate, and e, the residual term. ~a

Experiment 2

Eighteen dry, nonpregnant multiparious Holstein cows were taken from pasture and fed in two 4-h periods per day in a tie-stall feeding barn where mangers were fitted with partitions to allow a simultaneous choice of two feeds. The cows were assigned randomly to 1 of 3 diet choices (treatments) in a 3 × 3 Latin square arrangement of six treatment sequences. The diet choices were: 1) WCS and chopped Journal of Dairy Science Vol. 68, No. 5, 1985

t¢ :ff o

,,q

I

WHOLE COTTONSEED PASSAGE AND DIGESTIBILITY Coastal bermudagrass hay (CCBH), 2) WCS and CS; and 3) WCS and a C R of 75% corn silage and 25% concentrate, dry basis. Feeds were offered for ad Iibitum intake and positions o f the two feeds in the t w o - c o m p a r t m e n t manger were rotated each day to balance cue effects. Observations. Periods were 14 days with the first 7 days an adjustment to diet interval, and the last 7 days, an observation period. Individual feed c o n s u m p t i o n was measured daily. Cows were weighed at the end o f each period. Feeds were sampled daily during the 7 days of each observation period and processed as in E x p e r i m e n t 1. Analyses included crude protein, A D F , and estimated total digestible nutrients (TDN) by the N Y D H I C Forage Testing Laboratory (3). The m o d e l was: Yijkl = /~ + TSi + C(TS)j + WK k + T 1 + eijkl where TS was t r e a t m e n t sequence; C(TS), cow within t r e a t m e n t sequence; WK, e x p e r i m e n t week; T, t r e a t m e n t ; and e, the residual term. RESULTS A N D DISCUSSION Experiment 1

Nutrient c o m p o s i t i o n of feeds is in Table 3. Ether extract o f the WLCS was considerably

1201

lower than in the National Research Council (NRC) report o f 1982 (13). R e c e n t analyses by U S D A of c o t t o n s e e d p r o d u c e d in Texas (2) show values in t h e range of 17 to 18% ether extract similar to ours (Table 3). The WACS had sharply higher concentrations of ether extract and crude protein, reflecting the absence of lint, but the t w o seed classes also were f r o m different sources, although of the same year. The digestibility estimations were based on the nutrient c o n t e n t of the c o m p l e t e rations. Passage estimates of WCS are in Table 4. During the standardization, 24 cows e x c r e t e d .74% of the WLCS eaten, d e m o n s t r a t i n g little advantage f r o m rolling. During the CP w h o l e seeds f o u n d in the feces were only .4% of t h o s e eaten by 6 cows fed WLCS c o m p a r e d to 11.3% by 6 cows fed WACS (Table 4). We assume the WLCS stratified in the r u m e n c o n t e n t s so t h e y were regurgitated with the forage and were chewed during rumination. Conversely, we assume a significant p o r t i o n o f t h e WACS (11.3%) did not stratify in a m a n n e r to cause regurgitation with corresponding chewing. F r o m t h e ' 2 4 cows fed WLCS during the SP, the whole seed screened was a b o u t 50% lint-free and a b o u t 50% retained a distinct a m o u n t of lint. Bath et al. (4) n o t e d that m u c h of the oil of WLCS appeared to escape microbial digestion in the r u m e n because o f its encapsulation by

TABLE 4. Whole seed passage in cows fed whole-linted or whole acid-delinted cottonseed, Experiment 1. Standardization period

Comparison period

WLCS 1

Dry matter eaten, kg/day Dry seeds eaten/day, kg Dry matter digestibility, % Dry seeds in 5 kg wet feces, % Fecal dry matter, % Fecal dry matter excreted/day, kg Fecal dry matter/5 kg wet feces, kg Dry seeds/kg dry feces, g Dry seeds excreted/day, g Seeds excreted as a percentage of seeds eaten, %

WLCS 2

WACS2

SE

X

SE

X

SE

20.90 3.13 58.1 3.0 22.8 8.74 1.14 2.68 24.41

.55 .08 .9 .5 .4 .28 .02 .46 4.42

18.71 2.80 55.8 1.7 22.4 8.18 1.12 1.56 11.90

1.67 .25 1.7 .4 1.9 .56 .09 .43 2.28

21.17 3.18 56.4 43.6 22.7 9.16 1.14 38.48 347.22

1.08 .16 2.0 5.9 1.4 .32 .07 5.31 40.60

.7

.1

.4

.1

11.3

1.7

1Whole linted cottonseed fed to 24 cows during the standardization period. 2Whole linted cottonseed and whole acid-delinted cottonseed each fed to 6 cows during the comparison period. All cottonseed was fed at 15% of the dry matter of complete rations. Journal of Dairy Science Vol. 68, No. 5, 1985

e~

b~

t7

5 t~ ¢D

< O

TABLE 5. Effect of seed type and physical form of whole cottonseed on digestibility of n u t r i e n t s - m e a n s ~ and standard errors. Standardi-

Z .o

Comparison period

zation period

Whole acid delinted cottonseed

Rolled linted cottonseed

Whole linted cottonseed

Rolled acid delinted cottonseed

0o

(%) SE Dry matter Starch Ether extract Crude protein 2 Calcium Magnesium 3

R

SE

55.8 90.5

1.19 .17 .05 .01 1.14 2. O0

58.2 90.8 89.4

.92 .56 .64

90.9 a

61.79

.65 1.73 1.38

60.2 a 31.9 25.2 ac

25.4 26.5

a'b'C'dMeans with the same superscript are not significantly different (P>.05). 1 Predicted means after adjustment for covariance. 2Whole vs. rolled are significantly different (P<.01). 3 Linted vs. delinted and whole vs. rolled are significantly different (P<.05),

R 60.9 89.8 90.9 a 63.6 b 24.8 33.0 bc

© SE .75 .24 .07 .01 .48 1.30

R

SE

R

SE

55.7

.80 .18 .07 .01 .54 .47

60.6

.52 .13 .05 .Ol .73 .66

89.9

85.3 b 58.9 a 22.1 21.2 ad

92.9 90.9 a

64.4 b 26.6 26.7 bd

C~

~" .t-

1203

WHOLE COTTONSEED PASSAGE AND DIGESTIBILITY TABLE 6. Nutrient composition of feeds, Experiment 2.

Corn silage Chopped coastal berrnudagrass hay Concentrate Whole linted cottonseed Complete ration

Dry matter

Crude protein

(%)

~

34.0 86.7 88.2 89. 5 42.1

8.1 7.9 22.2 20.9 12.4

the seed coat. Smith and others (18) reported that transfer of oil from WLCS resulted in a twofold increase in yields of stearic and oleic acids of milk fat. It was noted also that rumen hydrogenation of WLCS fatty acids produced about a fourfold increase of trans-oleic acid in the resulting milk fat. Heifers have been shown to reduce feed particle size by chewing to less than .3 mm (average) for chopped alfalfa hay (20). We do not know whether this applies to our diets or if the resulting particle size of WLCS after chewing would result in significant protection for the entrapped fat. Bath et al. (4) suggested that oil in WLCS escapes hydrogenation when the seed is digested in the abomasum and intestine. It seems unlikely that the acid of the abomasum is strong enough to rupture WCS, because commercial grade sulfuric acid (93 to 96%) is used to delint WLCS without rupturing or destroying the integrity of the seed coat (Rogers Delinted Cottonseed Co., Waco, TX, personal communication). Nor would our

Acid detergent fiber (% of dry matter) 27.4 41.3 10.4 37.0 22.6

results, showing 11.3% of the WACS passing through the cow intact, suggest digestion of the whole seed in the abomasum or intestine. If abomasal-intestinal digestion of the whole seed occurred, it also would suggest that gossypol would escape detoxification in the rumen, but gossypol toxicity was not a problem in cows fed twice (30% of diet DM) the amount used here (17). It is not apparent how unsaturated fatty acids from WLCS escape hydrogenation to appear in milk fat. The seed passage of 11.3% from WACS compares with 18 to 35% for shelled corn, but different conditions were present (11). Processing WACS prior to feeding should be beneficial as noted for other feed grains (21), but because of the small amount of WLCS passed, no advantage from processing would be predicted, and in hot weather, an increase in rancidity might occur. Brown et al. (5) compared long staple cottonseed (no lint), the same seed cracked, and WLCS; no significant differences in milk

TABLE 7. Analysis of variance for expressions of feed consumed, Experiment 2.

Source

Treatment sequence Cow (treatment sequence) Experiment week Treatment R-square

df

5 i2 2 2

WCS1

Option feed

Total feed

.0001 .0001 .1332 .0001 .81

(% body weight) PR > F 2 .0002 .0240 .0554 .0331 .1075 .0473 .0001 .0001 .89 .85

WCS

.0001 .0001 .0100 .0001 .89

aWhole cottonseed. 2Probability greater than F. Journal of Dairy Science Vol. 68, No. 5, 1985

4~ V-

t~ TABLE 8. Effect of diet choice (treatment) on c o n s u m p t i o n of whole cottonseed (WCS), the option feed, total diet dry m a t t e r (DM), and the percentage of WCS eaten, Experiment 2. <

Diet choice (treatments)~

O Ox

Wk

1

Z

3

Mean

2

"X Range X Range X Range

4 6

Mean

.39 .13--.87 .48 .29--.69 .67 .56--.90

.29 .11--.61 .32 .12--.58 .43 .25--.68

.29 .16--.49 .31 .06-.62 .14 .01-.57

.51 a

.35 b

.24 c

X Range X Range X Range

4 6

Mean

1.57 1.06--1.87 1.35 .80--1.74 1.53 1.32--1.69

1.74 1.52--2.06 1.83 .81--2.26 1.84 1.67--1.96

2.51 2.27--2.65 2.30 1.63--3.03 2.83 2.27--3.24

1.49 c

1.80 b

2.55 a

2

3

Mean

Option feed DM (% of b o d y weight) .32 a .37 a .41 a

1.18 .85--1.60 .88 .53-1.35 .86 .68-1.02 .97c

1.44 1.14--1.95 1.50 .67-1.80 1.41 1.21-1.60

2.22 1.95--2.46 1.99 1.54-2.80 2.69 1.89-3.18

1.45b

2"30a

1.62 a 1.4~ a 1.66 a

1.94 ab 1.83 ab' 2.07 a

24 9-51 35 22--46 43 36-53

17 5-35 17 6-26 23 14-36

11 7-20 14 3-29 5 .2-23

34 a

19 b

10 c

a'b,CMeans with different superscripts differ (P<.05). T r e a t m e n t s were diet choices of: 1) WCS and chopped coastal bermudagrass hay; 2) WCS and corn silage; and 3) WCS and a complete ration.

C3 © O

>

WCS (% of total feed)

Total diet DM (% of b o d y weight) 2

......

1

WCS DM (% of b o d y weight)

O

00 ol

2

18 b 22 a 24 a

WHOLE COTTONSEED PASSAGE AND DIGESTIBILITY yield were apparent when all cottonseeds were fed as 15% of the diet. We have no explanation for these results, which differ from ours. Digestibility data in Table 5 show lower than predicted (from NRC) values for DM, although the chromic oxide analyses of the complete rations (Table 3) showed values very near the amount added (.125%). We assume the bias was related to fecal sampling that resulted in low chromic oxide recovery but was uniform across cows and treatments. Others (9) have observed problems of erratic excretion patterns of chromic oxide resulting in low recoveries, which in our case caused low DM digestibilities (Table 5). The significantly lower digestibilities of ether extract and crude protein by cows fed the WACS reflects the loss of those nutrients in the intact seeds. Our analyses of the whole clean seeds passed by the 24 cows in the standardization period showed 21.5% ether extract and 23.6% crude protein, which suggests little digestion occurred during transit through the digestive tract. If no value can be shown for processing WLCS before feeding, how should dairy producers without silage or CR feeding systems feed WLCS? Experiment 2

Nutrient content of the feeds is in Table 6. As single feeds, the CS and CCBH are low in percentage of protein for nonpregnant cows (12). However, the CR had ample percentage of protein and exceeded energy needed (12). The average consumption of WCS when CCBH and CS were the option feeds provided more crude protein than necessary for maintenance of nonpregnant cows. Body weights averaged 607 kg and increased from 568 to 641 kg during the 6-wk experiment. The sharp increase in weight probably reflects fill as well as tissue accumulation as the cows were brought from pasture to high-energy diets. The statistical analysis summary in Table 7 shows treatments were significant for each expression of feed consumed. As the energy density of the option feed increased from CS to CCBH to the CR, the amount of WCS eaten decreased, the amount of the option feed and the total diet eaten increased, and the percentage of the total diet eaten as WCS decreased (Table 8). The increase of the option feed consumed with increasing energy density was expected, based

1205

on work of Montgomery and Baumgardt (10). However, WCS has a greater energy density than any of the option feeds but it was never the preferred feed (Table 8). An important finding was the large range among cows in WCS and the option feed eaten in every choice (Table 8). Nevens (14) described this phenomenon over 50 years ago when he noted that cows differed greatly in their preference for a feed and no two cows exhibited the same preference. Wide preference also was seen when two excellent forages were offered to lactating Holstein cows (6). In this study, the range of WCS eaten varied from 38 to 5.05 kg/day. As a result, topdressing or feeding WCS separate from other ingredients is contraindicated. No indications of health disorders were observed. Maximum consumption of 5.05 kg/day of WCS did not induce any clinical symptoms of gossypol toxicity; this amount is less than previously fed to lactating cows without causing obvious toxicity signs (17). Under these conditions, offering WCS for ad libitum intake in a two-choice option did not result in sufficient consumption to cause any apparent toxicity in these cows. ACKNOWLEDGMENTS

The writers express special appreciation to student workers, Ann Bauer, Anne Woelfe], and Shelton Wasilewsky, for help with sample preparation and data entry.

REFERENCES

1 Allan, J. A. 1969. The preparation of agricultural samples for analysis by atomic absorption spectroscopy. Varian Aerograph Co., Walnut Creek, CA. 2 Anonymous. 1983. Cottonseed quality. Crop of 1982. AMS, Cotton Div., US Dep. Agric., Memphis, TN. 3 Anonymous. 1983. Forage analysis report. Forage testing laboratory. New York Dairy Herd Improv. Coop., Ithaca, NY. 4 Bath, D. L., J. R. Dunbar, J. M. King, S. L. Berry, R. O. Leonard, and S. E. Olbrich. 1980. Byproducts and unusual feedstuffs in livestock rations. West. Reg. Ext. Publ. 39. 5 Brown, M., M. DeLung, O. Lough, and R. Swingle. 1982. Long staple cottonseed compares favorably with short staple seed in University of Arizona dairy feeding trial. Arizona Dairy Newsl., Sept. :3. 6 Coppock, C. E., R. W. Everett, N. E. Smith, S. T. Slack, and J. P. Harner. 1974. Variation in forage preference in dairy cattle. J. Anita. Sci. 39:1170. 7 Goering, H. K., and P. J. Van Soest. 1970. Forage Journal of Dairy Science Vol. 68, No. 5, 1985

1206

8

9

10

11 12

13 14

COPPOCK ET AL.

fiber analysis. Agric. Handbook 370. Agric. Res. Serv., US Dep. Agric., Washington, DC. Hill, F. W., and D. L. Anderson. 1958. Comparison of metabolizable energy and productive energy determinations with growing chicks. J. Nutr. 64: 587. Kiesling, H. E., H. A. Barry, A. B. Nelson, and C. H. Herbel. 1969. Recovery o f chromic oxide administered in paper to grazing steers. J. Anim. Sci. 29:361. Montgomery, M. J., and B. R. Baumgardt. 1965. Regulation o f feed intake in ruminants. 1. Pelleted rations varying in energy concentration. J. Dairy Sci. 48: 569. Morrison, F. B. 1951. Feeds and feeding. Morrison Publ. Co., Ithaca, NY. National Research Council. 1978. Nutrient requirements of dairy cattle. 5th rev. ed. Natl. Acad. Sci., Washington, DC. National Research Council. 1982. United StatesCanadian tables of feed composition. 3rd rev. ecl. Natl. Acad. Sci., Washington, DC. Nevens, W. B. 1927. Experiments in the self-

Journal o f Dairy Science Vol. 68, No. 5, 1985

15 16

17

18

19 20 21

feeding o f dairy cows. Illinois Agr. Exp. Sta. Bull. 289:427. Palmquist, D. L., and T. C. Jenkins. 1980. Fat in lactation rations: review. J. Dairy Sci. 63:1. Ramsey, D. S., and J. L. Miles. 1953. Cottonseed vs. cottonseed meal and corn as a protein source in a concentrate mixture for dairy cows. J. Dairy Sci. 36:1308. Smith, N. E., and L. S. Collar. 1980. Whole cottonseed and extruded soybeans for lactating cows. Page 33 in Proc. California Dairy Carrie Day. Smith, N. E., L. S. Collar, D. L. Bath, W. L. Dunkley, and A. A. Franke. 1981. Digestibility and effects of whole cottonseed fed to lactating cows. J. Dairy Sci. 64:2209. Technicon SMA. 1976. Industrial method 33474A. Technicon Ind. Syst., Tarreytown, NY. Van Soest, P. J. 1982. Nutritional ecology of the ruminant. O. & B. Books, Inc., Corvallis, OR. Williamson, J. L. 1973. Effect of grain processing for dairy cattle and other animals. Page 349 in Effect of processing on the nutritional value o f feeds. Natl. Acad. Sci., Washington, DC.