Underutilized Resources as Animal Feedstuffs (1983)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Forest Residues INTRODUCTION Wood residues such as wood pulp and bark have been used as energy sources for ruminants during periods of critical feed shortages, but they have never been generally recognized as alternatives for conventional feedstuffs under normal economic conditions (Scott et al., 19691. Although more than 1.5 million tons sulfate and sulfite pulps from spruce, pine, and fir were fed to cattle and horses in the Scandinavian countries during World War II when feed supplies were limited (Hvidsten and Homb, 1951; Nordfeldt, 1951), the feeding of wood pulp to livestock ceased when conventional feedstuffs became available. Feeding of wood-derived feedstuffs in North America has been largely experimental, with the ex- ception of isolated situations in which wood residues have been fed on a commercial scale. WHOLE-TREE RESIDUE AND FRACTIONS OF WHOLE TREES Quantity Historically there has been little need for forest inventory information on quantities of forest biomass beyond information on raw material needed by the forest products industry. As a result, inventories estimated the volume of forest biomass as merchantable boles of commercially important tree species measured from a minimum of 13 cm diameter at breast height 69

70 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS (DBH) to a minimum top diameter of 10 cm. There have been recent attempts to estimate the total annual forest biomass available for use as feed, fuel, and chemical raw material. Some of these estimates assume current commercial forestland area and forest management practices while other estimates assume maximum forest biomass production from fully stocked, intensively managed forestland. A report by the Society of Amer- ican Foresters (1979) indicates that the annual forest biomass production from present commercial forestland area could double by the year 2035. The estimates of annual forest biomass potential are shown in Table 7. In addition, the report indicates that if 10 percent of the arable land that is currently private forest, pasture, range, and hay land were used for intensive production, up to 240 million dry tons additional annual forest biomass production could be available. If the additional land were managed under short-rotation, intensive-culture biomass farms, the increase could be nearly 450 million dry tons annually (Inman, 19771. Some studies have been made to measure and to develop methods to estimate the weight of forest biomass for various species, based on DBH TABLE 7 Estimates of U.S. Aboveground Forest Biomass Potential (millions dry tons) 2035 to 2085 (With Full 2010 to 2035 Stocking and (With Full Intensive Source19701976a2002a Stocking) Management)b Net growth from commercial forest408481544 726 1,088 MortalityC1 091 091 09 1 09 1 09 Other sources100100100 100 100 Harvest for conventional product uses-17?- 181- 236 - 236 - 290 Total440509517 699 1,007 NOTE: Estimates are derived from data published by Zerbe (1977) and U.S. Department of Agriculture (1977). Moderate industrial demand is projected (Society of American For- esters, 1979). aNet growth estimates from Thomas H. Ellis, USDA Forest Service, Oct. 20, 1978. bModerate industrial demand is projected to the year 2020, and it is also assumed that intensive management will double growth on 50 percent of commercial forest. CAssumes mortality is recoverable. Intensified management may reduce mortality, but an equivalent volume would be available as live wood. Land clearing, noncommercial lands, urban tree removal, and urban wastes.  Reprinted courtesy of the Society of American Foresters.

Forest Residues 71 and height measurements or on DBH alone. Whole-tree weight tables for 23 species growing in New York have been published (Monteith, 19791. The results are tabulated by species and DBH, showing fresh and dry weights of the whole tree (aboveground portion), of the entire bole, and of the bole cut off at various top diameters. Monteith also derived pre- diction equations for green and dry weight of the various components for 10 species and made recommendations for application of these equations to other species. Eleven species of puckerbrush ranging from 1.0 cm to 16.5 cm DBH were sampled and the total fresh and dry weights measured, along with the fresh and dry weights of leaves, branches, and stems (Ribe, 19734. Prediction equations were also developed to calculate the fresh and dry weight, the total, and each component for various DBH of each species. Ek and Dawson (1976) have published the dry-weight yields of Populus "tristis #1" grown under short-rotation intensive culture. The results are expressed in terms of dry-weight yields at four spacings of from 0.23 x 0.23 cm to 1.22 x 1.22 m at up to 5 years of growth. After 5 years they reported total aboveground weights at 1.22 x 1.22 m spacings of 49.2 tons/hectare. This total weight yield consisted of 10.76 tons wood, 1.88 tons bark, 24.72 tons wood and bark, 6.32 tons twigs, 5.22 tons leaves, and 0.35 tons other. Since aspen (Populus tremuloides Michx., P. grandidentata Michx., and P. balsamifera L.) is the species that has most potential as a livestock foodstuff, it should be considered separately. Aspen is the most widespread species in North America. The range is controlled by adequate moisture levels and cool summer temperatures and stretches from Mexico to the Arctic Ocean and from Maine to Alaska. Residue from aspen utilization is usually 100 percent because when aspen is commercially utilized, it is often the only species processed. Commercially important concentrations of aspen exist in the Northeastern, Great Lakes, and Central Rocky Moun- tain areas of the United States, and the central portions of Canada. The growing stocks in the Rocky Mountain and Lake States areas of the United States are over 100 and 200 million m3, respectively (U.S. Department of Agriculture, 1976~. Collectibility Not all of the annual forest biomass potential shown in Table 7 is collec- tible. Some will be on lands where it is too costly to harvest because of terrain or distance from markets. Some will not be harvested because of environmental, aesthetic, wildlife, and soil fertility considerations. Har- vesting as much as possible of the annual growth from forests under the current forest management practices, and especially under intensive forest

72 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS management practices, requires new and innovative forest harvesting equipment and schemes. Whole-tree harvesting equipment is currently used to obtain pulp-chip-quality wood and a feed-quality by-product. This equipment was developed during the early 1970s to harvest whole-tree chips for the pulp and paper industry. Currently, the main emphasis is to develop equipment to harvest forest biomass for fuel. Although limited quantities of forest biomass will be available in the future from primary and secondary wood processing plants, biomass avail- ability for livestock feed will most likely be obtained directly from the forest. This material could be a by-product of the system for the harvesting of whole-tree pulp or fuel chips. The by-product would consist mainly of the leaves or needles, twigs, buds, and wood and bark fines. If veneer, pulp, and sawmill logs were harvested, the potential feedstuff would be a by-product of harvesting the tree tops and branches, and it would also contain leaves, needles, twigs, buds, and wood and bark fines. In the case of short-rotation, intensively managed tree farms, the feedstuff could consist of the whole tree, but most likely would be a by-product after separation of stemwood, which would be used for pulp and fuel chips. The collectibility of aspen depends upon the region of the country. In the Lake States, northeastern areas of the United States, and central Can- ada, aspen residues from primary wood processing plants are available, as well as whole-tree or portions of whole-tree chips. In the Rocky Moun- tain areas where less aspen is harvested, only modest amounts of residues from primary processing are available. Whole-tree or portions of whole- tree chips are not available in the Rocky Mountain area because this harvesting method is not yet practiced there. The primary markets for aspen in the Lake States, northeastern United States, and central Canada are for pulpwood, sawlogs, and composition board. Residue from these uses are bark, which contains up to 50 percent wood, and sawdust that may contain some bark. These residues are col- lected and, if not currently used, constitute a disposal problem. For a large operation, such as a pulpmill, a sawmill with dry kilns, or a com- position-board plant, the bark and other wood residues are usually burned to supply process and space heat. This can require all of the residue wood except in the case of a sawmill with dry kilns. A sawmill usually requires only about one-half of the sawdust and bark to kiln-dry the lumber. The remainder is often used for landfill or sold as cattle bedding, mulch, or fuel. Where whole-tree harvesting is practiced for pulp chips, the segre- gated fines and bark are put back on the soil. This residue accounts for up to 20 percent of the harvest, and some of it may have value as a feedstuff. If the whole tree is harvested for fuel without further processing, the fines could be separated at the burning site for feedstuff use. If whole

Forest Residues 73 tree aspen is to be processed to fuel pellets (6 mm diameter x 15 mm long), the fines would not be available because they are fuel for the drier; but the fuel-pellet product could be suitable as a feedstuff. In the Rocky Mountain area, considerable aspen is available. It is not often harvested, however, because it is in scattered stands, usually at elevations of between 2,100 and 3,300 m, and long distances from pro- cessing plants and markets. Physical Characteristics Many forms of residues are produced from wood processing plants. For instance, sawmill residue consists of sawdust, planer shavings, lumber edge and end trim, slabs cut from the outer portions of the log, and shredded bark. Usually the larger pieces are processed to pulp chips or used for fuelwood or charcoal. The shredded bark, sawdust, and shavings frequently have no markets, but these residues are used increasingly as fuel by the forest products industry. The particle size of sawdust, planer shavings, and shredded bark de- pends upon the processing equipment used. Particle size of sawdust is usually less than 6 mm. A sawmill cutting lumber from bark-free aspen logs in the Lake States will produce sawdust with about 10 percent, by weight, of the particles greater than 6 mm, about 40 percent between 6 mm and 10 mm, and 50 percent less than 10 mm. Planer shavings are usually curls of wood 20 to 50 mm long and of various widths. The particle size of shredded bark also depends upon the species and age of the tree. Some barks are stringy, and older barks tend to break into small pieces. Sometimes, especially in the spring, long strips of bark come free of the log. Bark must be processed to achieve a smaller and more uniform particle size for feeding. This can usually be done quite economically with a hammermill. Forest residues include treetops, branches, and short lengths of logs. These residues can be reduced in size with portable chippers, such as those used to make whole-tree chips. These chips would have to be further reduced in size if they were to be used for livestock feed. The foliage can be separated from wood and bark by air classification at the whole-tree chipping site. The bulk density of green sawdust is in the range of 250 to 350 kg/ m3. The moisture content of sawdust from a fresh log depends upon species, time of year, and relative amounts of heartwood and sapwood. Usually fresh or green sawdust will contain 40 to 50 percent water. The bulk density of planer shavings can be as low as 100 kg/m3 for uncom- pacted material and up to 250 kg/m3 for compacted shavings. The moisture

74 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS content of planer shavings usually ranges from 7 to 15 percent because the shavings are usually cut from dried lumber. The bulk density of shredded bark is in the range of 250 to 350 kg/m3 if processed through a hammermill. The moisture content of fresh bark is usually 40 to 50 percent. The bulk density of fresh whole-tree chips, uncompacted, ranges from 300 to 400 kg/m3. Storage of the residues and whole-tree materials is a problem when the materials are fresh or have a moisture content above 17 percent. Numerous studies have been made on the storage of fresh pulp chips and residue for fuel use because of the decomposition and spontaneous combustion that have occurred in large piles (Springer et al., ] 978; White, 19791. Among the variables of storage are the geographic area, species, size of pile, time of year, and particle sizes of materials in the pile. It would be advisable not to store fresh materials intended for use as a livestock foodstuff. If they are stored, it should be in piles less than 6 m high. If they are to be stored in higher piles for more than a few weeks, provisions should be made to monitor the temperature within the pile. The pile should not be covered with plastic, canvas, or heavy snow because they trap heat and cause more rapid internal heating in the pile. A low, uncovered pile will ventilate and maintain low internal temperatures. Storing fresh wood in anaerobic conditions such as a silo would be a suitable method to prevent heating and decay organisms. Storing fresh bark, whole-tree chips, and foliage is more difficult than storing wood in the form of sawdust, chips, and shavings because of the higher amount of nutrients and biologically active materials. Nutritive Value Most untreated woods are quite indigestible. Using an in vitro technique, Millett et al. (1970) determined the relative digestibility of 27 species of trees. A summary of the results is shown in Table 8. All of the hardwood species examined showed some degree of digestibility, ranging from a low of 2 percent to a high of 37 percent. Aspen was most digestible, followed by soft maple and black ash. All of the softwoods were essentially indigestible. These results are in general agreement with those Stranks (1959) obtained with a pure culture of Ruminococcus flavefaciens. He reported that aspen and ash were relatively digestible in vitro, whereas elm, birch, and basswood required alkali or chlorite pretreatment before they were appreciably digested. Nehring and Schramm (1951a,b) also reported that ash, aspen, maple, and elm were superior roughage sources for sheep, compared to oak and birch. There is a positive correlation between the digestibility of the bark and the wood of a given species, with the bark usually being more digestible

Forest Residues 75 (see Table 81. The rather high content of ether extract in some bark might contribute to its higher digestibility. As shown in Table 8, maple twigs and buds are more digestible than the stemwood because they contain less lignin and have a higher percentage of digestible protein and ether extract (Nehr~ng and Schutte, 1950~. The soft maple buds and twigs had digestion coefficients of about 36 percent, compared to 20 percent for the wood. It is well known that buds and twigs are preferred brouse for a number of wild animal species. TABLE 8 In Vitro Dry-Matter Digestibility of Various Woods and Barks Digestibility'' Substrate Wood (%) Bark (%) Hardwoods Red alder Trembling aspen 33 50 Trembling aspen (groundwood fiber) 37 Bigtooth aspen 31 Black ash 17 45 American basswood 5 25 Yellow birch 6 16 White birch ~ Eastern cottonwood 4 American elm 8 27 Sweetgum 2 Shagbark hickory 5 Soft maple 20 Soft maple buds 36 Soft maple small twigs 37 Sugar maple 7 14 Red oak 3 White oak 4 Softwoods Douglas fir 5 Western hemlock 0 Western larch 3 7 Lodgepole pine 0 Ponderosa pine 4 Slash pine 0 Redwood 3 Sitka spruce 1  White spruce 0 a96-hour in vitro rumen digestibility. For comparison, the 96-hour digestibility of cotton liners was 90 percent and of a reference alfalfa, 61 percent. SOURCE: Millett et al. (1970). Courtesy of Journal of Animal Science.

76 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS Table 9, taken from Millett et al. (1970), shows the water solubility of various hardwood barks and the in vitro digestibility of the various barks. Significant amounts of hydrolyzable oligosaccharides or phenolic glyco- sides are present in the extracts of aspen and black ash barks. Approxi- mately one-half of the total weight of material in a hardwood-bark water extract of these barks appears to be carbohydrate that could be of nutritional value to the ruminant. It is assumed the remaining portion of the material consists primarily of phenolic compounds. It is clear that there are large differences in digestibility between tree species. Of the important tree species found in North America, aspen (Populus spp.) is by far the most promising as a ruminant [eedstuff; thus the emphasis below is on aspen. Aspen Bark Chemical Composition The chemical composition of aspen bark can be quite variable. One of the more complete reports on chemical analysis was prepared by Enzmann et al. (1969) (see Appendix Table 11. Crude protein content of aspen bark is usually less than 3 percent on a dry basis. From 5 to 10 percentile ether-extractable. Crude fiber is usually in the range of 40 to 55 percent, and ash can be highly variable, depending upon the amount of soil contamination. Nutrient Utilization Enzmann et al. (1969) ground and ensiled aspen bark containing 59 percent dry matter. Lambs were fed the ensiled bark TABLE 9 Water Solubility of Various Hardwood Barks and Extent of Carbohydrate Dissolution Carbohydrates in Water Extract Weight Free Sugars After Dry-Matter Loss Sugars Hydrolysis Digestibility Species (%) (%) (%)a (%)b Aspen 16.0 3.3 6.2 50 Black ash 22.1 6.5 10.0 45 Yellow birch 11.5 5.0 4.8 16 American elm 10.4 5.0 3.6 27 Sugar maple 5.5 1.7 1.9 14 al-hour hydrolysis at 121°C with 4 percent H2SO4 followed by neutralization with lime. b96-hour in vitro rumen digestibility of unextracted bark. SOURCE: Millett et al. (1970). Courtesy of Journal of Animal Science.

Forest Residues 77 in three forms: with no additions; as a mixture of 87 percent bark and 13 percent ground barley; and as a mixture of 87 percent bark, 13 percent ground barley, and 0.025 percent Aspergillus oryzae fermentation product on a wet basis. Digestible dry-matter content of the ensiled bark was 31.7 percent, and total digestible nutrient (TDN) content was 36.7 percent on a dry basis. This compares with the TDN content of about 45 to 50 percent for barley, rye, wheat, and oat straw. The mixtures of bark and barley grain had a TDN content of 44 to 48 percent. Mellenberger et al. (1971) offered goats diets containing 15, 30, 45, or 60 percent of air-dried bark. By regression they determined that the bark had a digestibility of about 50 percent. This experiment used smooth, green bark from freshly cut trees. Bark from older trees may not be as digestible. In a series of unpublished studies from this same laboratory, aspen bark from other sources appeared to have a dry-matter digestibility closer to 30 percent (L. D. Satter, University of Wisconsin, Madison, 1980, personal communcation). Similar estimates of digestibility of aspen bark have been reported by Gharib et al. (19751. This study was designed to test the effect of bark particle size on digestibility. Lambs were provided diets that contained 20 percent ground corn, 20 percent soybean meal, and 60 percent bark, plus a mineral and vitamin supplement. Dry-matter digestibility was 27.4, 25.7, and 30.3 percent for bark ground through hammermill screens of 0.32, 0.95, and 1.59 cm, respectively. Animal Performance Relatively little information is available on animal productivity when aspen bark is fed. Enzmann et al. (1969) offered diets containing approximately 37, 53, or 68 percent ensiled aspen bark (dry basis) to 40 kg wethers. The balance of the diet was a mixture of soybean meal and oats. Wethers receiving diets containing 37 and 53 percent aspen bark gained approximately 0.04 kg/day during the 48-day trial. Wethers receiving the 68 percent bark diet lost about 0.04 kg/day. The authors suggested that other feedstuffs, such as poor quality hay or straw, may be better sources of energy during emergency periods than aspen bark. Fritschel et al. (1976) fed ewes a diet containing approximately 67 to 70 percent aspen bark (dry-matter basis); 8 percent dehydrated alfalfa; 20 percent a mixture of ground shelled corn, oats, and soybean meal; and the balance as mineral, salt, and vitamin mix. The ewes were on the experimental diet for approximately 11 months, and during this time they lambed, lactated, and were bred. The ewes readily consumed the feed and had moderate gains in body weight. Consumption of the diet ranged from about 1.7 kg to 3.0 kg/day, dry-matter basis. During lactation the ewes received an additional 0.5 kg of a grain mix. An average of 1.3 lambs were weaned per ewe. The ewes appeared normal in all respects, and their

78 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS TABLE 10 Feedlot Performance of Cattle Fed Pelleted Diets Containing Whole Aspen Tree Material (93 days) 48% Alkali 12% 24% 36% 48% Treated Performance Alfalfa Aspen Aspen Aspen Aspen Aspen Initial weight (kg) 321 32~) 318 318 319 319 Final weight (kg) 375 405 425 431 422 422 Average daily gain (kg/day) 0.6 0.9 1.2 1.2 1.1 1.1 Kg feed/kg gain 19.5 13.8 12.1 12.0 11.6 12.0 SOURCE: Singh and Kamstra (1981 a). performance compared favorably with the control group, which was fed hay. Aspen Sawdust Chemical Composition Aspen sawdust has a composition similar to that of whole aspen tree. Protein content is less than 2 percent, and acid detergent fiber is usually in excess of 60 percent. Ash content is usually less than 1 percent, unless there is soil contamination, in which case ash content may range up to 10 percent. Klason lignin content is usually between 16 and 22 percent. Nutrient Utilization The digestibility of aspen sawdust cut from bark- free logs has been determined. Mellenberger et al. (1971) incorporated sawdust into either high-roughage or high-concentrate diets at levels of 0, 20, or 40 percent, air-dry basis. As expected, the overall digestibility of the high-roughage diets was lower than that of the high-concentrate diets. Measures of apparent digestible dry matter decreased linearly with both types of diets as the percentage of aspen sawdust increased from O to 40 percent of the diet. Dry-matter digestibility of the sawdust portion of the diet was 41 percent when it was incorporated into a high-roughage diet and 28 percent in a high-concentrate diet. Whole Aspen Tree Chemical Composition As with other wood residues, crude protein con- tent of whole-aspen-tree material is very low (see the section, "Forest

Forest Residues 79 Foliages. Total ash and mineral content is normally low, but can be high as a result of soil contamination (see Appendix Table 11. Nutrient Utilization Little information is available on the digestibility of whole-aspen-tree material. Digestibility studies of diets containing 0, 20, 40, or 60 percent of ensiled whole-aspen silage suggested that the diges- tible dry-matter content of aspen silage in an 80 percent grain diet was only 5 percent but approached 37 percent when incorporated into a 40 percent grain diet (Robertson et al., 1971~. This latter digestibility figure is about 80 percent of that expected with wheat straw. Singh and Kamstra ( 1981 a) reported dry-matter digestibilities of 51, 52, 54, 60, and 63 percent for total mixed diets containing 0, 12, 24, 36, or 48 percent ground whole aspen tree. Animal Performance A series of three experiments involving rather large numbers of cattle fed whole-tree material have been reported from South Dakota. In the first of these experiments (Singh and Kamstra, 1981a), 60 Hereford steers, each weighing approximately 320 kg, were allotted to 12 pens of 5 animals each. Mature aspen trees, including bark and leaves, were harvested in summer in the Black Hills region of South Dakota. The trees were chipped and dried to about 10 percent moisture. The dried chips were hammermilled prior to incorporation into complete pelleted diets. The six treatment diets were (1) 93 percent alfalfa (control), (2) 12 percent aspen, (3) 24 percent aspen, (4) 36 percent aspen, (5) 48 percent aspen, and (G) 48 percent aspen treated with- 4 percent sodium hydroxide. Soybean meal was incorporated into the pellets at a level equal to two- thirds of the aspen present. A mixture of molasses and mineral and vitamin mix constituted 7 percent of the pellet, with the balance of the pellet supplied by alfalfa. A summary of animal performance in this trial is shown in Table 10. The cattle fed aspen performed well, and actually were more efficient in converting feed to body gain than were the cattle fed alfalfa pellets. It should be pointed out that substantial amounts of soybean meal were present in the aspen treatments, and that much of the animal response may be due to the highly digestible soybean meal. The authors concluded, however, that the aspen had a net energy for main- tenance (NEm) of 1.53 Mcal/kg and net energy gain (NEg) of 0.48 Meal/ kg. These values are slightly higher than what one would expect from medium-quality hay. In the second study (Singh and Kamstra, 1981b) 60 Hereford steers, each weighing 327 kg, were divided into six treatment groups and fed to a slaughter weight of 500 kg. The six treatment groups received the following diets: (1) a high-roughage control composed of 93 percent al

80 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS falfa; (2) an all-concentrate diet; (3) a high-concentrate diet with 15 percent alfalfa as roughage; (4) a high-concentrate diet with 15 percent aspen as roughage; (5) a diet of 48 percent aspen, 13 percent alfalfa, and 32 percent soybean meal; and (6) a diet of 48 percent aspen, 13 percent alfalfa, 16 percent soybean meal, and 16 percent chicken manure. All diets were fed in meal form. The results of this trial are shown in Table 11. Animals receiving the high-grain diets (2, 3, and 4) gained the most and had the highest feed efficiency. It would appear from this study that aspen was contributing relatively little to the performance of feedlot cattle fed high-grain diets. Animals tended to eat more of the aspen-diluted diets, but feed efficiency dropped sufficiently with aspen addition to suggest that aspen was con- tributing little or nothing to the net energy available for growth. The third experiment (Singh, 1978; Singh et al. 1978) consisted of a field trial involving feeding aspen to wintering beef cows. Three treatment groups of 67 cows each were fed for 25 weeks. The three treatments were (1) mixed grass hay as a control; (2) aspen-alfalfa (60:40) pellets; and (3) 88.5 percent aspen silage, 10 percent corn, 1 percent limestone, and 0.5 percent urea. The cows were placed in the experiment on November 16. Animals on treatments 1 and 2 were given 0.72 kg protein supplement starting February TABLE 11 Feedlot Performance of Cattle Fed Complete Mixed Rations Containing Alfalfa and/or Whole Aspen Tree Material (126 days) Treatment 48% Aspen + 13% 48% Alfalfa Aspen + 16% + 13 % Soybean Concen- Concen- Alfalfa Meal All bate bate + 32% + 16~o Concen- + 15% + 15% Soybean Chicken Performance Alfalfa bate Alfalfa Aspen Meal Manure Initial weight (kg) 327 326 327 327 327 326 Final weight (kg) 459 533 Average daily gain (kg/day) 1.04 Kg feed/kg gain 12.4 529 1.65 1.61 5.5 6.0  534 467 432 1.64 1.11 0.84 6.8 9.8 12.4 SOURCE: Singh (1978)

Forest Residues 81 1 and March 26, respectively. To maintain animals after calving, an additional supplement of 2.7 kg corn/head/day was given during May. The silage group needed additional supplementation. After 6 weeks, 2 percent dried molasses was added to the silage diet for 4 weeks. A top dressing of soybean meal at the rate of 1 .2 kg/head/day was offered starting December 15 to increase the crude protein content of the silage diet to 7 percent. Soybean meal supplementation was increased to 1.7 kg per head per day from January 27 until termination of the experiment. Corn sup- plementation of 1.5 kg during the month of February and 2.0 kg thereafter was offered to the silage group. Feeding of the treatment diets and their supplements continued until cows and calves were turned out to open range. Total feed consumption data were not obtained in this experiment, but dry-matter consumption of the three basal diets was similar. The perfor- mance of the cows, divided into young and old groups, is shown in Table 12. All cows lost some weight, with the silage group losing slightly more than the control and pellet group. Normal and healthy calves were born. There were no significant differences in birth or weaning weights among the group fed hay, pellet, or silage rations. Cow and calf losses were 3.5 percent and 0.5 percent, respectively, as compared to 1 percent and 10 percent observed during previous years at the ranch where this field trial was conducted. Although the number of open cows was slightly higher with the control group, this difference was not statistically significant. This series of trials involving large numbers of animals suggests that whole-aspen-tree material is best suited for inclusion in high-roughage, maintenance-type diets. Net energy content of the aspen appeared con- siderably higher when fed to steers receiving high forage diets than when fed to steers receiving high-grain diets. This observation agrees with the finding of Robertson et al. (1971) that digestibility of whole-aspen silage was very low when incorporated into high-grain diets. The digestibility of most roughages would be depressed when incorporated into high-grain diets. Forest Foliage Serious and sustained studies on the use of coniferous foliage in feeds for domestic animals and poultry began in the 1930s. Increased interest in more complete utilization of forest tree components has resulted in con- siderable forest foliage research, especially in the Soviet Union. Forest foliage is used primarily as a source of carotene, trace elements, and vitamins in poultry, swine, and cattle diets in the Soviet Union. When used, it typically constitutes 3 to 8 percent of the total diet (Keays, 19761. Muka is the Russian term for flour or meal, and it has become the term

82 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS TABLE 12 Performance of Young and Old Cows Fed Wintering Diets Containing Whole Aspen Tree Material Control Pellet Silage Performance Younga Oldb Younga Oldh YoungU Oldb Initial weight (kg) Final weight (kg) Change in weight (kg) Initial condition scores Final condition scores Number of bred cows Number of open cows Birth weight of calves (kg) Weaning weight of calves (kg) 439.8 411.1 493.0444.0490.8 448.9403.5449.8 28.7- 44.1- 40.5- 41.0 5.1 4.8 28 5.4 5.1 23 26 9 5 4  35.8 168.3 5.2 4.3 34.3 34.3 163.9 171.7 435.6 389.5 46.1 5.5 4.9 28 6 5.0 4.2 24 34.6 33.6 179.5 164.9 490.9 433.3 57.6 5.4 4.5 25 4 33.8 168.7 a2 to 5 years.  b6 to 13 years. 'Scored on a scale of 1 to 10 with higher values representing cows in better condition. SOURCE: Singh (1978). generally used to describe animal feed derived from tree foliage. Muka can be derived from softwood needles or hardwood leaves. The foliage is usually dried and ground and may be pelleted or fed in loose form. The Soviet Union produced approximately 100,000 metric tons of muka in 1975, and plans were for doubling that by 1980 (Keays, 19761. Foliage is about 10 percent by weight of the unbarked, full bole for mature softwoods and 25 percent for young softwoods. Corresponding values for hardwoods are 5 and IS percent. Keays (1976) has estimated that the foliage in the world's forests could produce more than lOO million tons of muka. A conservative estimate of foliage practically available would yield somewhat less, perhaps 10 million tons. Machine harvest of trees results in imperfect separation of foliage from bark and wood. Con- sequently commercial foliage is usually defined as needles, leaves, twigs, shoots, and branches up to 0.6 cm in diameter. In general, the biomass of commercial foliage available from a tree is approximately double that of the leaves alone (Keays, 19761. There is growing interest in North America in the short-rotation intensive culture of trees to meet growing demand for pulp and paper products. Populus spp. has been given the most emphasis in the north central region

Forest Residues 83 because in this location it has repeatedly outproduced other species. Pop- ulus tristis grown under a short-rotation, intensive-culture system can produce up to 42 metric tons aboveground biomass/ha in 5 years (Isebrands et al., 19794. The commercial equipment presently available for whole- tree processing will convert about 80 percent of the tree into wood chips suitable for pulp manufacture, and 20 percent into leaf, bark, and twig material. This latter fraction has been examined as a ruminant feedstuff. Preliminary observations on the digestibility of this material by goats suggest a digestibility value of about 40 percent (L. D. Salter, University of Wisconsin, Madison, 1979, personal communication). Chemical Composition The chemical composition of muka varies with the species of tree used, but the values given in Table 13 may be considered characteristic of commercial muka from Pinus sylvestris. Muka contains only about 5 percent protein, and it is quite fibrous. It appears to be a good source of carotene and riboflavin. Nutrient Utilization The digestibility of conifer muka by sheep has been estimated by Apgar et al. (19771. Sheep were fed a timothy hay diet containing 25 percent of a mixture of spruce and fir muka. The sheep TABLE 13 Nutrient Composition of Muka Made From Pinus sylvestris Component Composition, Air-Dry Basis Dry matter Organic matter Protein Ether extract Cellulose Nitrogen-free extract Ash Calcium Phosphorus Carotene Riboflavin Vitamin C Vitamin E Manganese '7 ~lnc Copper Cobalt (%)  90.4 84.6 4.9 4.5 32.1 43.0 5.8 0.8 0.5 (mg/kg) 100-250 20 20-2000 20-40 120-350 25-35 10-20 0.5-1 SOURCE: Keays and Barton (1975). Courtesy of estate of John Keays.

84 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS readily consumed the muka-timothy hay mixture, but the dry-matter di- gestibility of the mixture was 58.9 percent compared to 64.1 percent for timothy hay. This suggested that the dry matter digestibility of the muka itself was 43.4 percent or a little less than what would be expected of straw. Animal Performance Keays and Barton (1975) cite quite a number of Soviet studies in which muka was fed, but it is difficult to evaluate the animal response from the information presented. The general impression is that muka addition to the diet improved animal performance. This conclusion is not supported, however, by studies conducted in North America with growing chicks. Gerry and Young (1977) replaced ground corn with 5 percent fir (Abies spp.) muka, spruce (Picea spp.) muka, or a combination of the two in a starter and finisher broiler diet. At 7.5 weeks of age the broilers given the muka diets had significantly lower body weights. Feed consumption was not different from the control, but the feed:gain ratio was significantly poorer when spruce muka was added alone or in combination. Hunt and Barton (1978), using broiler chicks up to 4 weeks of age, also found that spruce muka incorporated at levels of 2.5 to 10 percent in basal growing diets resulted in lower growth rates. The North American studies with muka in broiler diets have not indi- cated any benefit. The high-fiber content of muka does lower the energy value of diets for poultry. In addition, muka made from some tree species, such as spruce, contains phenolic compounds and tannins, both of which may contribute to depression in feed intake and growth (Hunt and Barton, 1978). Processing Methods Braconnot (1819) first discovered that cellulose could be converted to fermentable sugars by means of concentrated mineral acid. Research on wood hydrolysis did not progress until Simonsen (1898) suggested using dilute sulfuric acid. Ten to fifteen years later, three commercial-scale wood hydrolysis plants were constructed in the United States. Two plants used dilute sulfuric acid hydrolysis, with the major product being ethyl alcohol, and the other plant used anhydrous sulfur dioxide gas and pro- duced feed (Kressman, 19221. Research to utilize wood in animal feeds began at the Forest Products Laboratory in 1920 when eastern white pine (Pinus strobus) and Douglas fir (Pseudotsuga menziesii) sawdust was hydrolyzed and fed to animals at the University of Wisconsin and the U.S. Department of Agriculture,

Forest Residues 85 Beltsville, Md. The work was started as a result of high feed-grain prices during 1918 and 1919. Wood was hydrolyzed and the washings and hy- drolyzate were neutralized, concentrated, mixed with the unhydrolyzed residue, and dried (Sherrard and Blanco, 19211. This type of material was used in several feeding experiments with sheep and dairy cows (Archibald, 1926; Morrison et al., 1922; Woodward et al., 19241. Results indicated that certain animals could eat diets con- taining up to one-third hydrolyzed sawdust mixture. Animals such as dairy cows requiring considerable energy intake could eat up to 15 percent of the hydrolyzed mixture without noticeable effects on milk production. It was determined that the eastern white pine mixture was 46 percent di- gestible and that the Douglas fir mixture was 33 percent digestible. It was concluded that feeding hydrolyzed wood was practical only when natural feed grains were in short supply. Research on wood hydrolysis was conducted in the 1940s to produce concentrated sugar solutions suitable for stock and poultry feed. Over 200 tons of molasses were produced in pilot plants and sent to universities, agricultural experiment stations, and other agencies for feeding tests with milk cows, beef cattle, calves, lambs, pigs, and poultry (Lloyd and Harris, 1955; U.S. Department of Agriculture, 19601. In general, the tests indi- cated that wood-sugar molasses is a highly digestible carbohydrate feed and was comparable to blackstrap molasses. In addition, torula yeast was grown on neutralized dilute wood hydrolyzate, and the yeast was evaluated as an animal feed ingredient (Harris et al., 19511. Results from feeding tests with wood molasses led to commercial pro- duction during the early 1960s of a concentrated hemicellulose extract called Masonex, a by-product from hardboard production (Turner, 1964~. Various chemical, biochemical, and physical treatments have been sug- gested to increase the digestibility of wood by animals. These treatments include: · Hydrolysis with various acids to solubilize the cellulose · Alkali and ammonia treatment to saponify the ester bonds and pro- mote swelling beyond the waterswollen dimensions to increase micro- biological penetration into cell wall structure · Breaking of lignin-cellulose chemical bonds or actual delignification with various chemicals to yield a digestible cellulose · Comminution to very small particle size to alter the crystalline struc- ture of cellulose · High energy electron irradiation to break chemical bonds of the lignin and cellulose

86 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS Hydrolysis Early research and commercial methods increased cellulose digestibility by hydrolysis. This resulted in the production of a molasses product or a product containing a mixture of the hydrolyzed cellulose and the residue after hydrolysis. Operation of commercial production plants was short- lived, however, because of high costs. At the present time in the United States, three plants produce a wood molasses (hemicellulose extract) from wood as a by-product of the production of hardboard by the wet process. Two of these plants use a 1- to 2-minute high-pressure steam cook after which the steam pressure is suddenly released (Turner, 19641. The other plant uses a steam-pressurized refiner to defibrate the wood chips. In each plant the solubilized wood materials, which are mainly hemicellulose sugars and organic acids, are neutralized and either concentrated or spray dried for use in animal diets. The combined production of this molasses and dried product is about 68,000 tons/year on a dry-matter basis. Since this is a by-product, production depends on the markets for hardboard. Typical analysis of this molasses is 0.5 percent protein, fat, and fiber; 6 percent ash; 0.8 percent phosphorus; 3 percent calcium; 34 percent total sugar; and 35 percent moisture. In addition to the 3 plants presently producing a molasses product, 13 additional plants produce hardboard or medium-density fiberboard by a wet process and do not concentrate solubilized wood sugars and acids. These plants could supply an estimated 180,000 tons, dry-matter basis, of additional molasses. One pulpmill, using a prehydrolysis prior to pulp- ing, is also capable of producing an estimated 27,000 tons molasses/year, dry-matter basis. Steaming is also an approach to increasing the digestibility of wood. While steaming generally involves no added acid, the cleavage of acetyl groups early in the process provides an acidic medium conducive to hy- drolytic action. Steaming has been applied to both straw and wood. Recent applications to wood are described in studies by Bender and colleagues (Bender et al., 1970; Heaney and Bender, 19701. These showed that steam treatment of aspen (Populus spp.) chips for about 2 hours at 160° to 1 70°C gave a product that is readily accepted by sheep at up to 60 percent of the diet and provided normal weight gains and carcass yields. This research led to the process developed by Stake Technology Ltd., Toronto, to make a steamed product from wood and agricultural residues (Bender, 1979; Stake Technology, Ltd., 1979~. Steaming time is a few minutes, after which the pressure is rapidly reduced. Two plants in the United States are using this process to produce feed from aspen (Timber Resources, Inc., Bangor, Maine, and Enfor Feeds, Aitkin, Minnesota).

Forest Residues 87 Another process has been developed by Iotech Corp., Ltd., Kanata, Ontario, to steam and explosively release the pressure to produce a wood product that has increased digestibility (Marchessault and St. Pierre, 19784. A hydrolysis process was developed by Jelks (1976) to produce animal feedstuffs. This process is based on additions of a mineral acid in the presence of a catalyst and oxygen. Several plants have been constructed in the United States to produce animal feedstuffs from wood and agri- cultural residue. The Madison process using dilute sulfuric acid could also be used if additional molasses is needed (Lloyd and Harris, 1955; U.S. Department of Agriculture, 19601. Over 40 plants based on a modified Madison process are operating in the Soviet Union to make wood sugars for fermentation to yeast for human food or animal fodder. Yeast is a desirable end product from the wood sugars because of the protein value and ability of the yeast to metabolize both 5- and 6-carbon sugars. Some of the wood sugars are fermented to ethanol. Alkali and Ammonia Although alkali swelling of wood has not been used commercially to prepare feed, the relative effects of alkali on wheat straw and on poplar (Populus spp.) wood were studied by Wilson and Pigden (19641. Both materials were soaked in increasing amounts of sodium hydroxide up to 15 percent of the material weight, and the digestibilities of the products were obtained for straw and wood (see Figure 24; however, the digestion coefficient of the straw was always about 30 percent greater than that of the wood, the maximum digestibilities being about 80 percent for straw and 50 percent for wood. Thus, while the overall reaction mechanism of the alkali appears to be the same, the advantage of straw resides in the greater initial availability of the carbohydrates. A further demonstration of the effect of alkali on the digestion of wood was provided by Pew and Weyna (19621. By alternately swelling 80-mesh spruce (Picea spp.) wood in cold 2 N sodium hydroxide and following this by digestion with Tri- choderma viride enzyme, 80 percent of the total carbohydrates were con- verted to sugars. Without the alkali treatment, the maximum attainable digestion was only a few percent. Data are presented relative to the in vitro digestibility and lignin content of some typical hardwood species treated with 1 percent sodium hydroxide at a 20:1 solution-to-wood ratio (see Table 14~. All wood samples were ground through a 40-mesh screen in a Wiley mill. Treatment effectiveness was monitored by a Tilley and Terry in vitro rumen technique (Mellen- berger et al., 1970) augmented by animal feeding trials with the more . . promising proc ucts.

88 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS 90 1 1 1 1 1 1 1 1 1 1 1 1 1 1 - ~ if WHEAT ST/?AW S ~ ~ / ~ -~ Cot - 20 ,` / /0 o _ ~ Y l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 O 5 /O /5 THEA OMEN t ~ (G. Na 0~//00 G. MA TE~/A L ~ FIGURE 2 Effect of NaOH pretreatment on the in vitro digestion of straw and poplar wood. SOURCE: Wilson and Pigden (1964). No uniformity exists in the response of the various species to sodium hydroxide treatment. For example, while the digestibility of basswood increased from 5 to 56 percent, the digestibility of American elm (Ulmus americana) increased from 9 to only 14 percent. Softwoods, as a group, exhibited digestibilities in the 1 to 5 percent range and were essentially unresponsive to alkali treatment. This difference in response appears to be related to the lignin content of the wood (see Figure 31. A similar

Forest Residues 89 TABLE 14 Effect of NaOH Treatment on the In Vitro Digestibility of Hardwoodsa Digestibility . . ~ Llgnln Content Untreated Treated Species (%) (I) (%) Quaking aspen (Populus tremuloides) 20 33 55 Bigtooth aspen (Populus grandidentata) 20 31 49 Black ash (Fraxinus nigra) 20 17 36 American basswood (Tilia americana) 20 5 55 Paper birch (Betula papyrifera) 21 8 38 Yellow birch (Betula alleghaniensis) 21 6 19 American elm (Ulmus americana) 23 8 14 Silver maple (Acer saccharinum) 18 20 41 Sugar maple (Acer saccharum) 23 6 28 Red oak (Quercus rubra) 24 3 14 White oak (Quercus alba) 23 4 20 a5 g wood treated 1 hour with 100 ml 1 percent NaOH, washed and dried. SOURCE: Baker et al. ( 1975). relationship appears to apply to other lignocellulosic materials, as evi- denced by the common observation that, as forages mature, their lignin content increases and the digestibility of their cellulose and hemicellulose components by ruminants decreases. To better define conditions for optimum processing, Feist et al. (1970) and Millett et al. (1970) investigated the influence of alkali concentration on the extent of in vitro digestion of aspen and red oak (Quercus rubra). The results (see Figure 4) show that from 5 to 6 g NaOH/100 g wood are necessary for maximum effect. This level of alkali is about equivalent to that required to react with acetyl and carboxyl contents of the woods. This led Tarkow and Feist (1969) to postulate that the main consequence of alkali treatment is the saponification of intermolecular ester bonds, thus promoting the swelling of wood beyond waterswollen dimensions and favoring increased enzymatic and microbiological penetration into the cell wall fine structure. Wood substance lost by water extraction at this level of treatment is about 5 percent for aspen and can be ascribed largely to the removal of saponified acetyl groups. Verification of the in vitro results was obtained in a feeding trial with  goats (Mellenberger et al., 19711. Hammerrnilled aspen sawdust was treated with 0.5 percent NaOH at a 10:1 liquid-to-solid ratio for 2 hours at room temperature. After draining, washing once with water, and air drying, the

-90 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS 60 i To - 30 /0 o o o ~ ~ 1 1 1 /9 /9 20 2/ 22 L/GN/~/ CON TE//T, % FIGURE 3 Relationship between lignin content and in vitro digestion for NaOH pretreated hardwoods. SOURCE: Feist et al. (1970). product was incorporated into pelleted diets at levels of 0 to 60 percent. A similar set of diets was prepared from untreated sawdust. Overall diet  digestibility, calculated from feed and feces analysis, is plotted as a func- tion of aspen content (see Figure 51. Extrapolation of the curves to 100 percent aspen yielded dry-matter digestibilities of about 41 percent for untreated aspen and 52 percent for the alkali-treated aspen. - Alkali treat- ments can thus increase ruminant utilization of aspen wood by approxi- mately 25 percent, making it equivalent to a medium-quality hay as a source of dietary energy. Another long-standing approach to upgrading the feeding value of lig- nocellulosic materials involves treatment with aqueous or gaseous am- monia. It has long been recognized that liquid ammonia exerts a strong swelling action on wood and cellulose and can effect a phase change in the crystal structure from cellulose I and cellulose III (Segel et al., 1954; Tarkow and Feist, 1969~. In an attempt to exploit these effects, a number of woods were exposed to anhydrous ammonia in both liquid and gaseous

Forest Residues 91 60 50 i 40 30 /0 ,: ol l l 0 2 / :1 o .: / o ~  /I/OR rfIEf;,N /?ED OA/{ O~AK/NG ASPEN I I _ ; 1 1 4 6 ~ /0 /2 /4 /6 /8 20 GfFAMS //~0~/ PER /00 GRAMS H/OOD FIGURE 4 Relationship between level of NaOH pretreatment and in vitro digestion for quaking aspen and northern red oak (aspen: 0 = 2 hours in 0.5 percent NaOH; ~ = 1 hour in 1 percent NaOH; red oak: 0 = 4 hours in 1 percent NaOH). SOURCE: Feist et al. (1970). form and the results assayed by changes of in vitro digestibility (Millett et al., 19701. Aspen appeared to be unique in the extent of its response to ammoniation, attaining a digestibility coefficient of about 50 percent as contrasted with Sitka spruce (Picea sitchensis (Bong.) (Carr.~) and red oak (Quercus rubra), whose ammoniated-product digestibilities were only 2 and 10 percent, respectively. In an extension of this work, 363 kg ammonia-treated aspen were  prepared by subjecting air-dried aspen sawdust to a 2-hour pressure treat- ment with anhydrous NH3 gas at 482 kPa and 25°C. After blowdown and removal of excess NH3, the product was blended into pelleted diets at levels of 0 to 50 percent and fed to goats, with ready acceptance over a 6-week feeding period. From extrapolation of the curve for ration diges- tibility in relation to wood content, an estimated digestibility of approx- imately 46 percent was obtained for the ammonia-treated product. This is about 6 percent lower than the in vivo digestibility of the NaOH-treated aspen but still sufficient to rank NH3-treated aspen in the same range as hay as a potential source of energy. As an added benefit, the NH3 treatment would provide some nitrogen. As with sodium hydroxide, Tarkow and Feist (1969) described the pertinent action of ammonia treatment to the hydrolysis, ammonolysis in this case, of glucuronic acid-ester crosslinks,

92 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS 70 - By w CL - z 60 o - o 50 J - m - ~ 40 1 C] 30 ~NaOH- TREATED O_ . UNTRE ATED' _. - _ - _. _ _, __ ~ _ ~1 1 1 1 1 0 20 40 60 80 100 ASPEN CONTENT OF RATION (PERCENT )  FIGURE 5 In vitro dry-matter digestion of rations containing untreated and NaOH-treated aspen. SOURCE: Mellenberger et al. ( 1971). thereby providing more ready access to structural carbohydrates by rumen bacteria and their associated enzymes. Similarly, the wide response-range exhibited by the various wood species toward NH3 treatment can be as- cribed, in part at least, to the extent of signification. Delignification Because lignin is the major roadblock to widespread utilization of the carbohydrate content of the abundant lignocellulosic residues, delignifi- cation would appear to provide a straightforward solution to increasing the digestibility of lignocellulosics. That it can be is indicated by the more than 1.5 million tons of sulfate and sulfite wood pulps from pine, spruce, and fir consumed by cows and horses during World War II in the Scan- dinavian countries (Edin, 1940; Edin et al., 1941; Hvidsten, 1940; Nord- feldt, 1951; Saarinen et al., 19581. This was an emergency situation,

Forest Residues 93 however, and the use of high-quality wood pulps declined rapidly with the renewed availability of the more conventional feedstuffs. Since the presence of lignin appeared to result in low digestibility, Baker (1973) determined what degree of delignification was required to obtain carbohydrate utilization by the rumen microflora. By selecting pulping variables, a series of kraft pulps was prepared having a broad range of residual lignin contents whose digestibilities were then determined in vitro. Four wood species were included in the investigation: two hard- woods, paper birch (Betula papyrifera) and red oak; and two softwoods, red pine (Pinus resinosa) and Douglas fir. As shown in Figure 6, an appreciable difference exists in the deligni- fication-digestibility response between hardwoods and softwoods. With too 80 - ~ 60 m C) o: 40 20 I' / , / ' o PAPER BIRCH o ~, BRED OAK / ~ _ --  o ~ _~' /~/ / ~ ~' / /1 ~ - ~ / ,' /o / to' to At' ' RED PINE/ ~ // Ill l/l /// / l '' DOUGLAS FIR O. 20 40 60 EXTENT OF DELIGNIFICATION ( % ) FIGURE 6 Relationship between in vitro digestibility and extent of delignification for kraft pulps made from four wood species. SOURCE: Baker (1973).

94 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS the two hardwoods, digestibility increases rapidly with delignification and approaches a plateau of about 90 percent as delignification approaches completion. With the two softwoods there is a distinct lag phase, especially pronounced with Douglas fir, during which extensive delignification is accompanied by only minor increases in digestibility. Following this lag phase, digestibility rises rapidly and almost linearly with delignification up to the digestibility maximum. As interpolated from these four curves, the extent of delignification necessary to obtain a product having an in vitro digestibility of 60 percent, that of a good-quality hay, is shown in Table 15 along with data on the lignin content of the original woods. As with alkali treatment (see Figure 3), digestibility response strongly correlated with lignin content, the re- sponse being measured in terms of the degree of lignin removal needed to achieve a specified level of product utilization. Additional support for this lignin dependence was obtained by Saarinen et al. (1959) in an in- vestigation of the in vivo digestibility of a series of birch and spruce pulps prepared by 10 different pulping techniques. Recalculation of their data to fit the format of Baker (1973) provided the results shown in Figure 7, which also includes Baker's curves for red pine and paper birch for com- parison. In spite of the wide variation in delignification techniques em- ployed by the two investigations, their results are quite comparable. This leads to the further conclusion that it is primarily the degree of deligni- fication that governs pulp digestibility, not the method of pulping. Kamstra et al. (1980) reported that ponderosa pine wood had a 60 percent in vitro digestibility after removal of 78 percent of the lignin with peroxyacetic acid. A similar relationship was encountered with respect to the growth of the fungus Aspergillusfumigatus on a variety of commercial pulps prepared TABLE 15 Degree of Delignification Required to Attain 60 Percent In Vitro Digestibility Required Lignin in Delignificationa Original Wood Wood (%) (%) White birch 25 20 Red oak 35 23 Red pine 65 27 Douglas fir 73 32 aBased on original wood. SOURCE: Baker (1973). Courtesy of Journal of Animal Science.

Forest Residues 95 too 80 A_ ~ 60 - m co Is 40 C:) oL . I ~I I I T _ AX- / PAPER BlR j x x x / X/ X O / _ / / - ' X/x / ~/ / RED PINE X / / ~/ x BIRCH - / x / o SPRUCE 20 1 / ~--1 1 1 1 1 ~ 1 1 0 20 40 60 80 100 EXTENT OF DELIGNIFICATION ( %) FIGURE 7 Relationship between digestibility and extent of delignification for wood pulps (data points from Saarinen et al., 1959; curves from Figure 6). SOURCE: Baker (1973).  under different conditions (Baker et al., 19731. As determined by the protein content of the fungal mass, reasonable growth on hardwood could be obtained at lignin contents of 14 percent or less, whereas fungal growth on softwoods was restricted to pulps having less than 3 percent residual . . gnln. Because complete delignification is not required to prepare a nutrition- ally acceptable feedstuff, attention has turned to chemical pretreatments that can substantially disrupt the lignin-carbohydrate complex, but may demand less costly procedures than conventional pulping processes. Treat- ment with sodium chlorite has been one approach. In acetic acid solution, sodium chlorite exerts a marked specificity for lignin solubilization and has long been a quantitative technique for the preparation of "holocel

96 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS lulose," the total carbohydrate portion of a lignocellulose (Green, 19631. The high digestibility of a birch holocellulose was described by Saarinen et al. (19591. All pulping procedures investigated thus far depend upon the selective removal of lignin for their beneficial effects. There is yet another possi- bility for enhancing the availability of wood-residue carbohydrates, that of disrupting the lignin-carbohydrate association in situ, without the se- lective removal of either constituent. Under proper conditions, it was found that gaseous sulfur dioxide can effect a disruption; the treatment appears applicable to both hardwoods and softwoods. For this work, wood residues in the form of sawdust were reacted for 2 hours (hardwoods) or 3 hours (softwoods) at 120°C with an initial SO2 pressure at room temperature of 207 kPa and a water-to-wood ratio of 3:1 (no free liquid) (Baker et al., 19751. After blowdown and a brief evac- uation, the treated woods were neutralized to about pH 7 with sodium hydroxide and then air dried. Table 16 presents analytical data and values for 48-hour enzyme digestion for both the original woods and the treated products. A sample of alfalfa was included for comparison. The enzyme digestion was by the method of Moore et al. (19721. As expected, cellulase digestion of the original woods was minimal, from a high of 9 percent for aspen to essentially 0 percent for the two softwoods. Even with alfalfa, only about half of the available carbohydrate was converted to sugars. Yields of the sulfur dioxide-treated products were 106 to 112 percent, based on starting material, as a result of the sulfonation and neutralization add-on reactions. Although all of the lignin was retained in the products, Klason lignin analysis of the five treated hardwoods showed lignin values of only 5 to 9 percent. This suggested TABLE 16 Composition and Cellulase Digestion of Various Woods Before and After SO2 Treatment Lignin (%) Carbohydrate (%) Digestibility (%) Wood Before After Before After Before After Quaking aspen 20 7 70 71 9 63 Yellow birch 23 9 66 67 4 65 Sweetgum 20 5 66 64 2 67 Red oak 26 8 62 60 1 60 Douglas fir 30 24 65 63 0 46 Ponderosa pine 31 19 59 58 0 50 Alfalfa 17 51 25 SOURCE: Baker et al. ( 1975).

Forest Residues 97 that the original lignin had been extensively depolymerized during sulfur dioxide treatment and converted to soluble lignins, a fact subsequently confirmed by extraction with boiling water. Depolymerization was less extensive with the two softwoods, and the higher Klason lignin values are reflected by a decreased digestibility. Enzymatic hydrolysis of the hard- wood carbohydrates was essentially quantitative, indicating a complete disruption of the strong lignin-carbohydrate association in the original woods. The 60 to 65 percent digestibility of the treated hardwoods ranks these materials equivalent to moderate- or high-quality forage. The two softwood products would be equivalent to a low-quality forage, but might be upgraded through a better choice of processing conditions. For information on palatability, possible toxic side effects, and in vivo nutritional value, a 136 kg batch of sulfur dioxide-treated material was prepared from red oak sawdust and fed to goats at levels of O to 50 percent of a pelleted roughage-type ration over an 8-week period. Average in vivo digestibilities for dry matter and carbohydrate as a function of wood content of the rations are plotted in Figure 8. Extrapolation of the curves to 100 percent sulfur dioxide-treated wood yielded values of about 52 percent for dry-matter digestion and 60 percent for carbohydrate digestion. From the shallow slope of the curves, it appears that a vapor-phase treat- ment with sulfur dioxide effectively converts red oak sawdust into a rum- inant feedstuff having the dietary energy equivalence of a medium-quality forage. Neutralization of the treated product with ammonia rather than sodium hydroxide would augment its crude protein content. llJ ~ 80 Q - z o 60 'I 100 a o 40 f C_ __ CARBOHYDRATE  _ of DRY MATTERS ~ t . -A 20 _ _ ~ ~1 1 1 1 1 1 1 1 1 C) O 0 10 20 30 40 50 60 70 80 90 100 A CARBOHYDRATE _ DRY MATTER_ SULFUR DIOXIDE-TREATED WOOD IN RATION ( PERCENT ) FIGURE 8 In vivo dry-matter digestion of rations containing SO2-treated red oak. SOURCE: Baker et al. (1975).

98 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS Biochemical Nine white-rot fungi were examined for their ability to remove lignin faster than they removed polysaccharides from aspen and from birch (Kirk and Moore, 19721. During decay most of the fungi decreased the lignin content of the aspen and the birch; that is, they removed a larger percentage of the lignin than of the polysaccharides. Lignin removal was always accompanied by removal of polysaccharides, but lignin removal did not correlate with removal of any particular component of the polysaccharides. During decay, lignin was usually more selectively removed in the first few percentage points of weight loss than were the polysaccharides. The decayed woods with less lignin were more digestible by rumen fluid than were the control samples. A summary of results of the study appears in Table 17. There is a very high inverse correlation between lignin content of the decayed wood and in vitro rumen digestibility. The coefficient of correlation is 0.95 for aspen and 0.97 for birch. The results show that some white-rot fungi are effective in removing lignin faster than polysaccharides from wood. However, delignifying wood by these fungi under the test conditions is relatively slow. Optimizing factors such as aeration, moisture, temperature, and source and amount of nutrient nitrogen could enhance the rate of decay. In addition, it may be possible to find conditions that would improve the selectivity of lignin removal. It may also be possible to find better fungus species or better isolates than were used, or to obtain desirable mutants. Grinding Millett et al. (1970) found that, while vibratory ball milling is indeed an effective pretreatment, the milling response is quite species-selective. As an example, Figure 9 shows the in vitro rumen digestibilities of five hardwoods as a function of the time of milling. With 140 minutes of milling, all woods more or less attained a digestibility plateau. However, the plateau of digestibility was widely different for the different species, ranging from about 80 percent for aspen and sweetgum (Liquidamber styraciflua) to only 20 percent for red alder (AImus rubraJ. Softwoods were even less responsive than red alder; five different softwoods showed a maximum digestion of only 18 percent after 120 minutes of vibratory milling. Thus, this selective species response severely limits the broad application of the milling technique. Time and energy costs impose further restrictions.

Forest Residues 99 TABLE 17 Lignin and Carbohydrate Content and Digestibility of Sound and Decayed Aspen and Birch Wood Approx- Range imate Average of Total Rumen Decay Weight Weight Carbo- Fluid Time Loss Losses Lignin hydrates Digestibility Fungus (days) (%) (%) (%) (%) (%) Aspen None (control) 0 0 15.9 71.3 46 Fomes ulmarius 77 13 10-17 11.9 73.6 64 71 51 46-56 18.7 61.5 46 Polyporus berkeleyi 88 21 16-23 8.5 70.5 77 Polyporus frondosus 64 20 19-22 10.5 71.8 71 101 50 45-53 12.8 65.0 66 Polyporus giganteus 28 15 14-18 14.9 68.8 57 Polyporus versicolor 46 40 15.4 70.6 52 Birch None (control) 0 0 - 21.2 66.8 20 Fomes ulmarius 64 15 14- 16 14.7 69.8 57 Ganoderma applanatum 34 17 15-20 21.1 63.4 29 64 32 30-33 20.0 63.0 41 Polyporus berkeleyi 44 8 6-10 16.2 66.4 54 69 22 19-26 16.0 64.1 60 Polyporus frondosus 44 14 12-16 14.7 66.8 63 Polyporus resinosus 40 11 9-14 15.7 65.7 57 111 22 18-24 15.4 64.6 57 SOURCE: Kirk and Moore (1972). Courtesy of Wood and Fiber Science. Irradiation The technique of irradiating wood or straw by gamma rays or by high-  velocity electrons substantially improves digestibility by rumen organisms (Lawson et al., 1951; Millett et al., 1970; Pritchard et al., 19621. However a strong species specificity appears again (see Table 181. The digestion of aspen carbohydrates is essentially quantitative after an electron dosage of 108 red, while spruce is only 14 percent digestible at this dosage.

UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS /00 so 60 - 40 20 ~1 1 1 1 1 1 1 1 1 1 1 1 _~f S WEE TG(JM A RED OaK f1/C/YORY NED A L DEN ~- _ O' ~1 1 1 1 1 1 1 1 1 1 1 1 0 20 40 60 80 /00 /20 /4 0 TIME OF MILL /NG (M/N. ) FIGURE 9 Relationship between in vitro digestibility and time of vibratory ball milling. SOURCE: Baker et al. (1975). Utilization Systems Experimental Previous research has demonstrated that certain fractions of forest biomass can be made useful in animal diets without additional treatments. The research has also shown that many species and fractions of forest biomass are not useful and that physical and chemical treatments are needed to enhance digestibility. Advancement of the experimental systems requires TABLE 18 Effect of Electron Irradiation on the In Vitro Digestion of Aspen and Spruce  Digestibility (%) Electron Dosage (reds) Aspen Spruce 0 55 3 lob 52 3 107 59 5 5 X 107 70 8 lo8 78 14 SOURCE: Baker et al. (1975).

Forest Residues 101 a close and long-range working arrangement between animal and wood sc~ennsts. Industrial .. The use of products from wood for animal diets is currently in practice. These systems have developed slowly, perhaps because of the lack of working arrangements between animal scientists and the wood industry. Again, close research and working arrangements are needed to promote these systems. Closer cooperation should be sought between the feed industry and the wood industry to increase research on products from the forest biomass that can be useful in animal diets. Potential The forest biomass is the world's largest storehouse of carbohydrates. In the past it has been used in animal diets on a small scale in emergencies. There will be competition for the forest biomass, both for traditional forest products and, increasingly, for fuel. There will be very little in the way of unused forest products industry residues. By choosing proper growing, harvesting, and processing methods, forest biomass can be wisely used for traditional forest products, fuel, and livestock feed. PULPMILL AND PAPERMILL RESIDUES Quantity The quantity, location, and composition of pulpmill and papermill primary sludges have been reported by Joyce et al. (19791. About one-third of the U.S. production of pulp and paper was included in the survey. The data are reported on geographic regions of the United States based on similarity of tree species, pulping processes, and end products within the region (see Figure 10~. Table 19 shows the quantity of primary sludge by pulping process. These data indicate a wide range in rates of primary sludge production. The amount of sludge produced is related to the quantity of fines that can be tolerated in the final product and, in the case of de- inking, the quantity of fines that are in the raw material or developed in refining the raw material. Table 20 shows the quantity of primary sludge produced per unit of production by the reporting mills. These data indicate an average of 84 kg/ton of production and a range of 2 to 400 kg/ton.

102 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS ~- / 6 - 1 111 ~ ~ 'art 5 :~ \J' Jo FIGURE 10 Division of the United States into six regions, based on best judgment of similar tree species, pulping processes, and end products within a region. SOURCE: Joyce et al. (1979). Courtesy of Elsevier, Amsterdam. Additional data are reported from papermills (see Table 211. Some papermills make pulp by one process and purchase market pulp produced by another process. Table 22 indicates the quantity of primary sludge production per unit of paper produced by region. This indicates an average of 52 kg/ton of paper production and a range of 2 to 185 kg/ton. The type of mill and of product produced influences the composition of the primary sludge, as shown in Table 23. These data indicate that the sludge can contain 10 to 90 percent inorganic content. The primary sludge data indicate that a substantial amount of cellulosic fiber is lost in pulp and papermaking processes. It is difficult to predict, however, how much usable fiber can be reclaimed from the primary sludges because of inherent variability of the manufacturing processes. The quantity and quality of the primary sludge are site-specific, and com- plete nutritional and chemical analyses are required to determine the use- fulness of a sludge in animal diets. For use in animal diets, the fibers should be collected prior to entry into the primary collection site. If the average primary sludge produced per ton of pulp is 84 kg and 50 percent of this is considered to be cellulosic fiber, at the 1980 pulp

Forest Residues 103 TABLE 19 Kilograms Primary Sludge per 1000 Kilograms Pulp Produced by Different Pulping Processes Mills Pulp Process Reporting Sludge Production Average Minimum Maximum Soda 1 14 NSSCa 1 20 Kraft 27 58 17 103 Sulfite 3 102 22 234 Chemi-mechanical 4 113 40 204 De-inking 3 234 24 400 aNeutral sulfite semichemical pulping. SOURCE: Joyce et al. (1979). Courtesy of Elsevier, Amsterdam. production rate of nearly 52 x 106 tons/year, 2.2 x 106 tons/year of cellulosic fibers are available in the primary sludge. North American sulfite pulp production and estimated associated sulfite spent liquor (SSL) solids appear in Table 24 (J. N. McGovern, University of Wisconsin, Madison, 1979, personal communication). Technological advancements and economic and environmental demands have resulted in important changes in sulfite mill practice over recent years that have affected the nature, production, and availability of SSL solids. These changes have primarily involved substitution of the classical sulfite- pulping reagent of calcium bisulfite (in sulfurous acid) with the more soluble and recoverable magnesium, sodium, and ammonium compounds TABLE 20 Kilograms Primary Sludge per 1000 Kilograms Pulp Production, by Region Sludge Production Mills Regiona Reporting Average Minimum Maximum 1 4 96 17 154 2 18 54 2 267 3 10 178 24 400 4 7 64 14 167 5 1 26 6 5 36 20 60 All reporting mills 45 84 2 400  aSee Figure 10 for regions. SOURCE: Joyce et al. (1979). Courtesy of Elsevier, Amsterdam.

104 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS TABLE 21 Kilograms Primary Sludge Produced per lOOO Kilograms Paper Produced by Different Types of Mills Sludge Production Mills Type of Mill Reporting Average Minimum Maximum NSSCa 2 12 8 17 Soda 1 15 Kraft 26 41 2 123 Chem i - mechanic al 3 60 5 0 8 0 Sulfite 2 65 50 81 De-inking 4 105 24 185 aNeutral sulfite semichemical pulping. SOURCE: Joyce et al. (1979). Courtesy of Elsevier, Amsterdam. and recovery of the heat and chemicals in the SSL by evaporation, com- bustion, and chemical recovery. Evaporation and recovery of lignosul- fonate by-products with prior desugaring to produce alcohol and yeast is practiced in several plants. A summary of these practices in U.S. sulfite mills is shown in Table 25 (J. N. McGovern, University of Wisconsin, Madison, 1979, personal communication). There appears to be only one sulfite mill in the United States that does not recover its spent sulfite liquors, while 14 others evaporate and burn the liquor either to recover the pulping chemical base and sulfur dioxide or, in the case of ammonia-base pulping, to incinerate the spent liquor to recover the sulfur dioxide. TABLE 22 Kilograms Primary Sludge Produced per 1000 Kilograms Paper Production, by Region Sludge Production Mills Regiona Reporting Average Minimum Maximum 1 4 30 15 50 2 17 43 2 1 14 3 15 79 8 185 4 9 41 15 100 5 1 23 6 3 44 20 62 All reporting mills 49 52 2 185  aSee Figure 10 for regions. SOURCE: Joyce et al. (1979). Courtesy of Elsevier, Amsterdam.

Forest Residues 105 TABLE 23 Inorganic Content of Primary Sludges Mills Reporting Following Inorganic Content (do) Pulp Process <10 11-20 21-40 41-60 61-80 >81 Kraft 5 4 7 9 2 0 Chemi- mechanic al 1 0 0 0 0 0 NSSCa O 1 0 0 0 1 Sulfite 2 0 0 0 0 0 Soda 0 0 1 0 0 0 De-inking 1 1 0 2 0 0 Total 9 6 8 11 2 1 aNeutral sulfite semichemical pulping. SOURCE: Joyce et al. (1979). Courtesy of Elsevier, Amsterdam. The SSL situation is quite different in Canada (see Table 26) where over 80 percent of the mills do not recover SSL solids and over half of the mills produce a high-yield newsprint pulp using a sodium acid sulfite process (J. N. McGovern, University of Wisconsin, Madison, 1979, per- sonal communication). Two mills produce SSL by-products, but about 2.7 million tons of SSL solids are discharged unused. TABLE 24 Sulfite Pulp and Spent Liquor Solids Production from North American Sulfite Mills Spent Liquor Number ProductionSolids Pulp Grade of Mills (1,000 tons)(1,000 tons) United States 28a Unbleached 359389 Bleached 1,6531,836 Dissolving and high alpha 950b952 Total 2,9923,177 Canada 37 Unbleached 2,3523,352 Bleached 8891,182 Dissolving and high alpha 378580 Total 3,6194,114 a27 as of 1979. bKraft excluded.  SOURCE: J. N. McGovern, University of Wisconsin, Madison, 1979, personal commu- nication.

106 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS TABLE 25 Spent Sulfite Liquor Handling in United States Sulfite Mills (number of mills) Desugaring, Evaporation, Evaporation By-Product No Reagent Combustion Recovery Recovery Total Calcium 1 6 7 Magnesium 10 10 Sodium 1 - 1 Ammonia 3 5 1 9 Total 14 12 1 27 SOURCE: J. N. McGovern, University of Wisconsin, Madison, 1979, personal commu . . n~cat~on. Physical Characteristics Primary sludge or cleaner cellulosic fibers at pulpmills and papermills are collected wet from a water slurry. Primary sludges contain the material removed from the bottom of a primary collection pond. Primary sludge contains about 65 percent water and, due to its consistency, it is difficult to remove additional water mechanically. The cellulosic fibers can either be in the form of fine fibers too short for use in paper or in the form of screener or knotter rejects that are too large to be used directly. Most mills recycle the screener rejects to refiners, but the smaller mills often do not recycle them for economic reasons. The fiber fines are also difficult to dewater to less than 65 percent water content. The screener rejects can be squeezed to remove water to about 55 percent water content. TABLE 26 Canadian Sulfite Pulpmills, by Process and Spent Liquor Solids Handling Mills Without Recovery Mills with Recovery Pulp Estimated Pulp Capacity Solids Capacity Base Number(tons/day) (tons/day) Number (tons/day) Calcium 71,697 2,100 Magnesium 2262 290 2 790 Sodium 196,850 6,850 2 360 Ammonia 31,275 1,275 2 1,200  Total 3110,084 10,515 6 2,350 SOURCE: J. N. McGovern, University of Wisconsin, Madison, 1979, personal commu- nication.

Forest Residues 107 The fiber fines are usually broken cellulosic fibers and parenchyma cells that are less than 1 mm long and have passed through a 100-mesh screen. The screener rejects are undercooked wood that sometimes has come from a knot in the tree, hence the name knotter rejects. These rejects are from 20 to 50 mm long and contain delignified wood on the surface and nearly whole wood on the inside. The primary sludge can contain these materials and, in addition, wood and bark fines from the wood room, dirt, grit, and papermaking additives such as coating clay, calcium carbonate, and ti- tanium dioxide. The inorganic content is usually in the form of less than 200-mesh fines. Nutritive Value Fines and Screen Rejects Four fiber types constitute the bulk of residues generated by pulpmills: ( 1 ) groundwood fines, fully lignified fiber fragments created during grind- ing or milling operations in newsprint mills; (2) semichemical pulping fines, partially delignified fiber fragments produced in the manufacture of pulps for corrugating materials; (3) screen rejects, which are partially pulped but contain unbleached chip fragments and fiber bundles that are undesired by-products of all chemical pulping operations; and (4) chemical pulping fines, fiber fragments generated in all pulp-making procedures along with ray parenchyma cells and vessel elements from hardwood tissue mill operations. Chemical Composition Information on the composition and in vitro ru- men digestibility of several pulpmill residues is shown in Table 27 and in Appendix Tables 1 to 4. Fines and screen rejects are very low in protein, and high in acid detergent fiber. Their mineral content reflects the exposure to chemical processing. Care must be used in feeding some of these residues because potentially toxic levels of mineral elements could be present, depending upon how the papermill handles its waste streams. Nutrient Utilization Digestibility values in Table 27 are expressed in terms of percentage weight loss after 5 days of incubation with buffered rumen fluid at 39°C (Mellenberger et al., 19701. As expected, groundwood fines yielded digestibility values comparable to those observed for sawdust of the same species: O percent for the pine and spruce and about 35 percent for aspen. All of the screen rejects and chemical pulp fines had digesti- bilities of more than 40 percent, and digestibility of two of the pulp fines was more than 70 percent. Thus, based on in vitro dry-matter digestibility,

108 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS TABLE 27 Composition and In Vitro Rumen Digestibility of Pulpmill Residues Composition (%) Type of Residue Groundwood fines Aspen Southern pine Spruce Screen rejects Aspen sulfite Mixed hardwood, sulfite Mixed hardwood, kraft Chemical pulp fines Mixed hardwood, kraft (bleached) Aspen sulfite Lignin Carbohydrate Ash In Vitro Digestibility (%) 21 31 31 19 24 25 1 109 73 59 60 37 o o 77 2 66 65 14 54 74 9 44 95 (parenchyma cells) 20 73 2 73  Southern pine, kraft (unbleached) 28 68 4 46 SOURCE: Baker et al. ( 1975). any of the screen rejects and chemical pulp fines could serve as useful sources of dietary energy for ruminants. The mixed hardwood, bleached- kraft chemical pulp fines are essentially pure cellulose. Table 27 shows that the lignin and total carbohydrate contents of the aspen groundwood, aspen sulfite screen rejects, and aspen sulfite parenchyma cell fines are almost identical, whereas the in vitro dry-matter digestibility ranges from 37 to 73 percent. The digestibility of fines of aspen parenchyma cells, for example, is higher than would be predicted on the basis of lignin content because the parenchyma cells contain substances that analyze as Klason lignin. Southern pine kraft (unbleached) pulp also contains substances that could analyze as lignin. In general, animals have readily accepted diets containing up to 75 percent pulp residues, and digestibility of the residues has been quite acceptable. For example, Millett et al. (1973) determined the digestibility of pelleted diets containing increasing levels of three types of pulpmill residues up to 50 percent of the total diet. Estimates of residue digestibility were obtained by regressing digestibility on the amount of residue in the diet. Mixed hardwood bleached-kraft fines were estimated to be 80 percent digestible, while screen rejects were 66 percent and parenchyma fines 50 percent. In vitro digestibilities of the three residues were 95, 66, and 73 percent, respectively. The relatively low in vivo utilization of the car- bohydrates in the aspen sulfite parenchyma fines may be related to their

Forest Residues 109 small particle size (about 0.1 mm long), which might favor short retention times and inadequate bacterial action in the rumen. Additional confirmation of the feeding value of wood pulps is provided by Saarinen et al. (19591. In vivo digestibilities of wood pulps prepared by various pulping techniques ranged from 27 to 90 percent. While these were experimental pulps, the results indicate the digestibility range to be expected from the fines fractions generated by commercial pulping op- erations. Dinius and Bond (1975) fed commercially available unbleached fiber fines along with activated sludges as a protein source to steers. The fines were screened from maple, birch, and beech pulp made by an ammonia- base sulfite pulping process. The effects of feeding pulp fines on ruminal pH, ammonia, and volatile fatty acid concentrations and on voluntary intake and growth of pregnant beef heifers throughout gestation were also measured. The fines were estimated to be 92.8 percent digestible. Weight gain, calf birth weight, or incidence of calving problems for heifers fed diets containing 75 percent fines did not differ from those of heifers fed hay. Lemieux and Wilson (1979) used pulp fines from the same com- mercial source and noted that diets containing 20 to 67 percent of the wood residue were as digestible as the corn-hay control diets. The growth rate of lambs fed these diets was comparable to that of the control lambs. There were no adverse effects of wood-containing diets on blood serum urea, protein or mineral levels, or rumen and liver histopathology. Animal Performance Millett et al. (1973) offered Hereford steers diets containing 0 and 50 percent unbleached southern pine kraft pulp fines for a period of 58 days. Average daily gain was 0.77 and 0.54 kg for the 0 and 50 percent pulp diets, respectively. Daily feed intake as a percentage of body weight was 2.84 and 2.48, and feed/gain ratio was 9.4 and 11.7, respectively. Steers fed the pulp diet tended to sort the ingredients and reject the fines fraction, indicating a lower palatability for this material. Fritschel et al. ( 1976) conducted five different feeding experiments with sheep and steers in which aspen sulfite parenchyma fines were fed at levels of 72 and 83 percent of the total diet. One experiment included pregnant ewes fed over an 11-month period, including lambing and lactation. Re- sults indicated ready acceptance of the diet by the ewes. The number and growth of lambs was equivalent to those from control hay-fed ewes. Equally good results were obtained with a group of Angus beef cows, some of which were pregnant at the beginning of the experiment and calved during the trial. The value of pulp-containing diets in lamb-fattening operations was tested in two feeding experiments by Riquelme et al. (1975~. In one, a

110 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS diet of 60 percent alfalfa and 40 percent concentrate was compared with two diets containing 66 percent pulp fibers (a bleached hardwood kraft and a bleached mixed species sulfite), 5 percent alfalfa, and 29 percent concentrate. In the second, a diet of 70 percent barley and 30 percent alfalfa was compared with those diets containing approximately 70 percent of the bleached sulfite pulp and three different nitrogen sources. In all cases the pulp-fed lambs had desirable carcass characteristics equivalent to or higher than the controls. Based on growth measurements, the pulp fiber was considered to have about 85 percent of the energy value of barley. Processing Many pulp and paper residues are highly digestible, and it is not expected that they will require processing. In the case of utilization of the fines from a mechanical pulpmill where the fines are merely groundwood, processing requirements are the same as for wood and sawdust of the same species. If the materials are too contaminated with inorganics or other papermaking additives, it is doubtful if further processing could be justified on the basis of economic and animal health considerations. Utilization Systems Presently, fines from a sulfite pulpmill are being incorporated into live- stock diets. The fines contain parenchyma cells, short fibers, and fiber bundles obtained from a pulpmill making fiber for tissue from birch, beech, and maple species. The process is an ammonia-base sulfite pulping pro- cess. The experimental program started with the testing of fines from an identical pulpmill operated by the same company that made pulp from aspen. The testing consisted of chemical analysis of the fines for lignin, carbohydrate content, ash, mineral elements, fats, and in vitro rumen digestibility. In vivo digestibility was determined with goats, and addi- tional feeding trials with sheep were made to determine palatability and the effect of palatability on general health indicators (Millett et al., 19731. The fines used are identified in Millett's report as aspen sulfite parenchyma cell fines (unbleached). Additional feeding trials were made with beef animals (Fritschel et al., 1976~. Nearly 50,000 kg were fed during a 3- year cooperative study between the Forest Products Laboratory, University of Wisconsin-Madison, and Procter and Gamble Company. Additional feeding trials were conducted by the Agricultural Research Service (Dinius and Bond, 19751. As a result of this and other research (Lemieux and

Forest Residues 111 Wilson, 1979), pulp fines containing about 65 percent water are presently being fed to dairy cows and beef animals in Pennsylvania. This experimental program required close cooperation between re- searchers of various disciplines, pulpmill management personnel, and corporate research staff. At the end, it involved close cooperation between pulpmill management and county agricultural agents to explain to the farmers how the fines are produced and how they should be used. The use of additional spent sulfite pulp fines in animal diets requires study to determine how to make the material available without causing serious changes in the material and energy balances in the pulpmills and how best to utilize it near its source. Potential The results of these studies involving pulpmill residues showed that the cellulosic fiber fines from two pulpmills, each producing about 31 metric tons (dry basis) per day, could be used in animal diets. The extension of these results to other pulpmill residues could have a lasting and continued impact on agricultural land use, food production, and wise use of our renewable resources. Sludges Chemical Composition In addition to the residues described above, pulpmills and papermills may produce primary clarifier or lagoon sludges. Table 28 shows the com- position and the in vitro dry-matter digestibility of various combined pulpmill and papermill sludges (Millett et al., 19734. Most of the sludge- containing by-products are very high in ash, and moderate to high in lignin. Both of these characteristics diminish the feeding value of the sludge. The Klason lignin values in Table 28 also include acid-insoluble paper additives (ash) as lignin. Errors in the lignin analysis are evident in the data listed in Table 21 for the combined pulpmill and papermill residues that have high ash content. Nutrient Utilization Because the groundwood mill sludges are mostly groundwood, digesti- bility is expected to be low, although the total carbohydrate content is high. One of the semichemical pulpmill sludges was high in digestibility

1 12 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS TABLE 28 Composition and In Vitro Rumen Digestibility of Combined Pulpmill and Papermill Sludges Composition (dry-matter basis) Total In Vitro Klasona Carbohydrate Dry-Matter Type of Residue Lignin (%) (%) Ash (%) Digestibility (%) Groundwood mill Mixed species plus some mixed chemical pulps 50 41 38 24 Southern pine plus some hardwood kraft 24 60 15 19 Semichemical pulpmill Aspen 20 71 2 57 Aspen plus mixed hardwoods 55 29 13 6 Chemical pulpmill De-inked wastepaper, tissue 23 71 22 72 Softwood sulfite, glassine 13 74 14 66 Reprocessed milk carton stock 28 67 25 65 Mixed chemical pulps, tissue 17 76 13 60 Mixed hardwood bleached kraft, printing 17 75 11 59 Aspen sulfite, tissue 19 77 2 50 Aspen and spruce sulfite, tissue 45 46 45 35 Secondary waste treatment sludge 38 5 45 5 aIncludes ash not soluble in H2SO4 SOURCE: Millett et al. (1973). Courtesy of Journal of Animal Science.  and total carbohydrate and low in ash, but the other was low in digestibility and total carbohydrate. This indicates the amount of variation that can be observed between products from mills that use the same pulping process. Dinius and Bond (1975) fed a mixture of 65 percent pulp fines and 35 percent sludge. The activated sludge was the residue from processing total pulpmill wastes through a lagoon system and constituted 30 percent of

Forest Residues 113 the diet. At 38 percent protein, it provided the sole nitrogen source. The remaining 5 percent of the diet consisted of inorganic salts. Organic-matter digestibility of this diet was 78.9 percent, but animal acceptance of the diet was poor. Dry-matter consumption was only 1.2 percent of body weight. Spent Sulfite Liquor The chemical composition of spent sulfite liquor (SSL) varies greatly and depends upon wood species, pulping reagent, reagent composition, and pulping conditions. For a normal yield of calcium bisulfite-sulfurous acid reagent pulp, the spent sulfite liquor can contain 12 to 16 percent total solids, 30 percent of which consists of monosaccharides with some oli- gosaccharides. The sugars in the presently unutilized spent sulfite liquor can be concentrated and mixed with molasses. Klopfenstein et al. (1973) reported that calcium and ammonia spent sulfite liquor can be effectively used as a liquid feed ingredient in beef cattle diets, but that it contains only about 50 percent as much energy as cane molasses per unit of dry matter because of lignin residues. WOOD RESIDUES AS ROUGHAGE SUBSTITUTES IN RUMINANT DIETS Ruminants generally require a certain level of roughage in their diet. Roughage provides tactile stimulation of the rumen wall and promotes rumination, which in turn increases salivation and supply of buffer for maintenance of rumen pH. Roughage in the dairy diet can help maintain normal milk fat test, and in high-grain feedlot cattle diets can lower the incidence of rumen parakeratosis and liver abscess. When traditional sources of forage or roughage are expensive or in short supply, alternative rough- age sources could be helpful. Wood residues have been investigated for this purpose. Aspen sawdust has been shown effective as a partial forage substitute in a high-grain dairy diet (Salter et al., 19701. Cows fed 2.3 kg hay and about 17 kg of a pelleted diet containing one-third aspen sawdust main- tained normal milk fat. Cows receiving a similar diet but without sawdust produced milk with 50 percent as much fat. Diets with one-third sawdust are not practical, however, for high-producing dairy cows. A subsequent study (Satter et al., 1973) was done in which lower levels of sawdust were fed with high-grain diets, with or without sodium bicarbonate and sodium bentonite. It was concluded that aspen sawdust can be a partial roughage substitute in lactating dairy cow diets and can be helpful in maintaining near-normal milk fat content in high concentrate diets. The

114 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS sawdust cannot serve as the only source of roughage for lactating cows because of the irregular feed intake that results if no other forage is fed. Oak sawdust is essentially indigestible, and has been used in several growth or feedlot trials as a roughage substitute (Anthony and Cun- ningham, 1968; Dinius et al., 1970; El-Sabban et al., 1971~. Inclusion of 5 to 15 percent oak sawdust in the diet has generally supported animal performance equal to the other experimental groups being tested. Unfor- tunately, the design of these experiments or the numbers of animals used prevent firm conclusions regarding the value of oak sawdust as a direct replacement for conventional roughage. The study of El-Sabban et al. ( 1971 ) does suggest that the incidence of liver abscesses did decrease with increase in dietary sawdust up to 15 percent of the diet. Rumens of steers fed diets containing sawdust were parakeratotic, but did show improve- ment, particularly when coarse sawdust particles were fed. It has been generally concluded that oak sawdust is an effective roughage substitute when used as 5 to 15 percent of the total diet. Similar conclusions were reached when pine sawdust was used as a roughage substitute in beef finishing diets (Slyter and Kamstra, 1974~. ANIMAL HEALTH Generally, whole-tree or tree residues are not considered dangerous to the health of livestock. It is essential, however, that diets containing wood residues be properly balanced for all of the essential nutrients. Wood residues must be considered primarily as energy sources. They contain only small amounts, or are nearly devoid, of many essential nutrients. Thus, animals consuming diets containing large amounts of wood residues will encounter ill health if the diets are improperly balanced. Pine needles have been shown to be toxic and can cause abortion in cattle. Pregnant mice fed an extract of ponderosa pine (Pinus ponderosa) needles had reduced litter size, and thus can serve as indicators of potential toxicity problems (Cogswell and Kamstra, 19801. Wood residues that have been exposed to chemical treatments, such as in the pulping process, should be carefully examined for residual chem- icals. In the pulp and papermaking industry, the process or method of handling waste streams can unwittingly result in contamination of the potential feedstuff. Sludges in particular need careful scrutiny. A number of sludges have digestibility values comparable to hay. Their suitability for animal feed, however, will depend on the amount of ash and the chemical nature of the individual ash constituents. For example, moderate levels of clay-type filler could be tolerated, but the presence of more than trace amounts of certain heavy metals would rule out use as a feedstuff. n

Forest Residues 115 Thus, each pulp and papermaking residue should be chemically charac- terized before it can be recommended as a feedstuff. REGULATORY ASPECTS The use of wood and wood-derived products in animal feeds requires the consideration of and compliance with state and federal regulations. The Association of American Feed (control Officials (1980) reported that ground whole-tree aspen (Populus spp.) and/or its parts can be fed to animals provided that the diets are supplemented with protein, vitamins, and minerals. RESEARCH NEEDS The relatively low digestibility of most wood residues has stimulated research into ways of treating or processing wood residues to increase digestibility. A number of approaches have proven effective in increasing digestibility, but they are costly and not economically feasible. The need for a cost-effective process is clear. While the need is very great for such a process, the likelihood of developing one may be diminished as cost of energy and chemical inputs increase. A major portion of the total wood residue is the residue that is left in the forest at the time of harvest. This suggests that improved techniques for whole-tree harvest be developed so the heretofore underutilized por- tions of the tree may be available for collection, processing, and transport to a site for utilization. In this connection, more research is needed on the nutritive value of the foliage portion of trees used for pulpwood. Commercial equipment exists for chipping whole trees, and sorting out the wood chips from leaf, bark, and twig material at the point of harvest. As this method of harvesting grows, so will the potential supply of foliage material. Much more information is needed on the feeding value of the foliage, and how feeding value is influenced by time of harvest, storage of the foliage, and the genetic background of the tree (clone from which the foliage was derived). Most of the wood residues that have potential as feedstuffs are relatively high in moisture. This causes problems in storage because of deterioration. Research is needed to identify low cost means of preserving high moisture wood residues. Preservation through ensiling and the use of mold inhibitors are approaches that might be fruitful. A significant portion of the research that has been done with wood residues has involved sheep or goats. Unfortunately they are not as well suited as cattle are to digest low-quality fibrous feeds. Sheep and goats

1 16 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS are selective browsers, and normally select the more easily digested plant material from their grazing environment. Cattle, because of their larger size and capacity for retaining slowly digested materials in the rumen for longer periods of time, can more easily accommodate the poorer-quality feedstuffs (Van Soest, 19824. Future research with the relatively low qual- ity wood residues should, wherever possible, utilize cattle rather than sheep or goats. SUMMARY The carbohydrates of whole-wood residues are, with few exceptions, re- sistant to attack by cellulolytic organisms in the rumen. This resistance apparently stems from the close physical and chemical association between cellulose and lignin, augmented by the crystalline nature of cellulose itself. Whether or not the carbohydrates contained in wood lignocellulosic res- idues can be utilized by rumen microbes will depend largely on how extensively the lignin-carbohydrate complex can be altered or opened up. Of the woods tested, all of the coniferous species are essentially un- digested by rumen microorganisms. Deciduous species, with a few ex- ceptions, are only slightly digested. Aspen is the most digestible species tested, giving both an in vitro and in vivo digestibility of about 35 percent. Aspen bark is about 50 percent digestible. All of the chemical and physical pretreatments discussed are effective to some extent, but exhibit a strong species preference that severely limits their applicability. Hardwoods are generally more responsive to pretreat- ment than softwoods, but even hardwoods exhibit a broad range of re- sponsiveness. Aspen is particularly susceptible to treatment. Several of the treatment methods, technically speaking, can be readily adapted to a commercial process. The cost of treatment has presented the biggest bar- rier. Conventional feedstuffs need to be relatively high priced before treated wood residues can compete in the marketplace. Prices of conven- tional feedstuffs have on occasion reached levels that would make treated wood residues attractive. The lack of a steady market has discouraged development of commercial wood processing enterprises. Wood has been shown to be effective as a roughage replacement. De- pending upon the other dietary ingredients, concentrations of 5 to 15 percent screened sawdust in diets for beef cattle appear practical. For lactating dairy cows, aspen sawdust could be used as a roughage extender or as a partial roughage substitute in high-grain diets. Some long hay appears to be necessary in the diet to stabilize feed intake. Finally, it must be realized that wood residues are generally low in protein and other essential nutrients that livestock require. This necessitates

Forest Residues 117 more extensive supplementation. Treated wood residues are primarily an energy source, and some may compare as an energy source to an average- or low-quality hay. For this reason, treated residues are best suited for ruminants having relatively low nutrient requirements, such as overwin- tering beef cows and ewes and "dry" dairy cows, and for larger size dairy and beef replacement heifers. Foliage does, however, contain sig- nificant protein and other essential nutrients that could make foliage quite useful as a feedstuff. If the foliage can be economically harvested and stored, it may be a potential animal feedstuff. Some of the many pulp and papermaking residues that are already partially delignified but that have little fiber value for paper manufacture have excellent potential as ruminant feedstuffs. Care must be used in their selection as feedstuffs, however, because some residues may contain toxic materials. LITERATURE CITED Anthony, W. B., and J. T. Cunningham, Jr. 1968. Hardwood sawdust in all concentrate rations for cattle. J. Anim. Sci. 27: 1159 (Abstr. ). Apgar, W. P., H. C. Dickey, and H. E. Young. 1977. Estimated digestibility of conifer muka fed to sheep. Res. Life Sci., University of Maine at Orono 24:1. Archibald, J. G. 1926. The composition, digestibility and feeding value of hydrolyzed sawdust. J. Dairy Sci. 9:257. Association of American Feed Control Officials. 1980. Official publication of Association of American Feed Control Officials, Box 3160, College Station, Tex. Baker, A. J. 1973. Effect of lignin on the in vitro digestibility of wood pulp. J. Anim. Sci. 36:768. Baker, A. J., M. A. Millett, and L. D. Satter. 1975. Wood and wood-based residues in animal feeds. P. 75 in Cellulose Technology Research. A. F. Turbak, ed. ACS Sym- posium Series 10. Washington, D.C.: American Chemical Society. Baker, A. J., A. A. Mohaupt, and D. F. Spino. 1973. Evaluating wood pulp as foodstuff for ruminants and substrate for Aspergillus fumigates. J. Anim. Sci. 37: 179. Bender, F., D. P. Heaney, and A. Bowden. 1970. Potential of steamed wood as a feed for ruminants. For. Prod. J. 20:36. Bender, R. 1979. Method of treating lignocellulose materials to produce ruminant feed. U.S. Pat. No. 4,136,207. Jan. 23. Braconnot, H. 1819. Hydrolysis of cellulose into sugar. Gilbert's Annalen der Physik. 63:348. Cogswell, C., and L. D. Kamstra. 1980. Toxic extracts in ponderosa pine needles. J. Range Manage. 33:46. Dinius, D. A., and J. Bond. 1975. Digestibility, ruminal parameters and growth by cattle fed a waste wood pulp. J. Anim. Sci. 41:629. Dinius, D. A., A. D. Peterson, T. A. Long, and B. R. Baumgardt. 1970. Intake and digestibility by sheep of rations containing various roughage substitutes. J. Anim. Sci. 30:309.  Edin, H. 1940. Experiments in the use of chemical pulp as fodder. Sven. Papperstidn. 43:82.

1 18 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS Edin, H., T. Helleday, and S. Nordfeldt. 1941. Feed cellulose for milk cows and~horses, including other feeding experiments during the shortage and crisis in feedstuffs. Lant- brukshoegsk. Husdjoursfoersoeksanst. Medd. No. 6:1. Ek, A. R., and D. H. Dawson. 1976. Actual and projected growth and yields of Populus "Tristis #1" under intensive culture. Can. J. For. Res. 6:132. El-Sabban, F. F., T. A. Long, and B. R. Baumgardt. 1971. Utilization of oak sawdust as a roughage substitute in beef cattle finishing rations. J. Anim. Sci. 32:749. Enzmann, J. W., R. D. Goodrich, and J. C. Meiske. 1969. Chemical composition and nutritive value of poplar bark. J. Anim. Sci. 29:653. Feist, W. C., A. J. Baker, and H. Tarkow. 1970. Alkali requirements for improving digestibility of hardwoods by rumen micro-organisms. J. Anim. Sci. 30:832. Fritschel, P. R., L. D. Satter, A. J. Baker, J. N. McGovern, R. J. Vatthauer, and M. A. Millett. 1976. Aspen bark and pulp residue for ruminant feedstuffs. J. Anim. Sci. 42:1513. Gerry, R. W., and H. E. Young. 1977. A preliminary study of the value of conifer muka in a broiler ration. Res. Life Sci., University of Maine at Orono 24:1. Gharib, F. H., R. D. Goodrich, J. C. Meiske, and A. M. E1 Serafy. 1975. Effects of grinding and sodium hydroxide treatment on poplar bark. J. Anim. Sci. 40:727. Green, J. W. 1963. Methods in carbohydrate chemistry. P. 12 in Vol. III, Cellulose. New York: Academic Press. Harris, E. E., G. J. Hajny, and M. C. Johnson. 1951. Protein evaluations of yeast grown on wood hydrolyzate. Ind. Eng. Chem. 43:1593. Heaney, D. P., and F. Bender. 1970. The feeding value of steamed aspen for sheep. For. Prod. J. 20(9):98. Hunt, J. R., and G. M. Barton. 1978. Nutritive value of spruce muka (foliage) for the growing chick. Anim. Feed Sci. Technol. 3:63. Hvidsten, H. 1940. Fodder cellulose, its use and nutritive value. Nord. Jordbrugsforsk. 22:180. Biol. Abstr. 22(5):1123(1948). Hvidsten, H., and T. Homb. 1951. Survey of cellulose and Beckmann-treated straw as feed. Pure Appl. Chem. 3:113. Inman, R. E. 1977. P. 62 in Silvicultural Biomass Farms. Vol. I Summary. McLean, Va.: MITRE Corp. Isebrands, J. G., J. A. Sturos, and J. B. Crist. 1979. Integrated utilization of biomass. A case study of short-rotation intensively cultured Populus raw material. Tappi 62(7):67. Jelks, J. W. 1976. Process for oxidizing and hydrolyzing plant organic matter particles to increase the digestibility thereof by ruminants. U.S. Pat. No. 3,939,286. Feb. 17. Joyce, T. W., A. A. Webb, and H. S. Dugal. 1979. Quantity and composition of pulp and papermill primary sludges. Resour. Recovery Conserv. 4:99. Kamstra, L. D., D. Ronning, H. G. Walker, G. O. Kohler, and O. Wayman. 1980. Delignification of fibrous wastes by peroxyacetic acid treatments. J. Anim. Sci. 50: 153. Keays, J. L. 1976. Foliage. 1. Practical utilization of foliage. Appl. Polymer Symp. 28:445. Keays, J. L., and G. M. Barton. 1975. Recent Advances in Foliage Utilization. Information Report VP-X-137. Vancouver: Canadian Forestry Service, Forest Products Laboratory. Kirk, T. K., and W. E. Moore. 1972. Removing lignin from wood with white-rot fungi  and digestibility of resulting wood. Wood and Wood Fiber 4(2):72. Klopfenstein, T., W. Koers, and S. Farlin. 1973. Sulfite liquor in beef cattle rations. Pp. 25-26 in 1973 Nebraska Beef Cattle Report. ED 73-218. Omaha: University of Nebraska. Kressman, F. W. 1922. The manufacture of ethyl alcohol from wood waste. USDA Bull. No. 983.

Forest Residues 1 19 Lawton, E. J., W. D. Bellamy, R. E. Hungate, M. P. Bryant, and E. Hall. 1951. Some effects of high velocity electrons on wood. Science 113:380. Lemieux, P. G., and L. L. Wilson. 1979. Nutritive evaluation of a waste wood pulp in diets for finishing lambs. J. Anim. Sci. 49:342. Lloyd, R. A., and J. F. Harris. 1955. Wood hydrolysis for sugar production. Report No. 2029. Madison, Wis.: USDA Forest Products Laboratory. Marchessault, R. H., and J. St. Pierre. 1978. A new understanding of the carbohydrate system. Chemrawn Conf., Toronto, Ontario. Mellenberger, R. W., L. D. Satter, M. A. Millett, and A. J. Baker. 1970. An in vitro technique for estimating digestibility of treated and untreated wood. J. Anim. Sci. 30: 1005. Mellenberger, R. W., L. D. Satter, M. A. Millett, and A. J. Baker. 1971. Digestion of aspen, alkali-treated aspen and aspen bark by goats. J. Anim. Sci. 32:756. Millett, M. A., A. J. Baker, W. C. Feist, R. W. Mellenberger, and L. D. Satter. 1970. Modifying wood to increase its in vitro digestibility. J. Anim. Sci. 31:781. Millett, M. A., A. J. Baker, L. D. Satter, J. N. McGovern, and D. A. Dinius. 1973. Pulp and papermaking residues as foodstuffs for ruminants. J. Anim. Sci. 37:599. Monteith, D. B. 1979. Whole tree weight tables for New York. Appl. For. Res. Inst. Res. Rep. No. 40. Syracuse: State University of New York. Moore, W. E., M. J. Effland, and M. A. Millett. 1972. Hydrolysis of wood and cellulose with cellulytic enzymes. J. Agric. Food Chem. 20:1173. Morrison, F. B., G. C. Humphrey, and R. S. Hulce. 1922. Unpublished Report. Madison, Wis.: USDA Forest Products Laboratory. Nehring, K., and J. Schutte. 1950. Composition and nutritive value of foliage and twigs. I. The change in composition of leaves and twigs from different seasons. Arch. Tier- ernahr. 1: 151. Nehring, K., and W. Schramm. 1951a. Composition and nutritive value of leaves and branches. II. The digestibility of summer leaves and twigs. Arch. Tierernahr. 2:264. Nehring, K., and W. Schramm. l951b. Composition and nutritive value of leaves and branches. III. Nutritive value of dead leaves and winter branches. Arch. Tierernahr. 6:342. Nordfeldt, S. 1951. Problems in animal nutrition. The use of wood pulp for feeding farm animals; the value of silage made with added acids. Proc. 11th Inst. Cong., Pure and Appl. Chem. 3:391. Pew, J. C., and P. Weyna. 1962. Fine grinding, enzyme digestion and the lignin-cellulose bond in wood. Tappi 45:247. Pritchard, G. I., W. J. Pigden, and D. J. Minson. 1962. Effect of gamma radiation on the utilization of wheat straw by rumen micro-organisms. Can. J. Anim. Sci. 42:215. Ribe, J. H. 1973. Puckerbush weight tables. Life Sci. and Agric. Exp. Stn. Misc. Rep. 152. Orono: University of Maine. Riquelme, E., I. A. Dyer, L. E. Bariba, and B. Y. Couch. 1975. Wood cellulose as an energy source in lamb fattening rations. J. Anim. Sci. 40:977. Robertson, J. A., S. E. Beacom, and R. Shiels. 1971. Feeding value of poplar silage in rations for yearling steers. Can. J. Anim. Sci. 51:243. Saarinen, P., W. Jenson, and J. Alhojarri. 1958. Investigation on cellulose fodder. I. Birch wood. Pap. Puu 40:495. Saarinen, P., W. Jenson, and J. Alhojarri. 1959. Digestibility of high yield chemical pulp  and its evaluation. Acta Agral. Fennica 94:41. Satter, L. D., A. J. Baker, and M. A. Millett. 1970. Aspen sawdust as a partial roughage substitute in a high-concentrate dairy ration. J. Dairy Sci. 53:1455. Satter, L. D., R. L. Lang, A. J. Baker, and M. A. Millett. 1973. Value of aspen sawdust as a roughage replacement in high-concentrate dairy rations. J. Dairy Sci. 56:1291. -

120 UNDERUTILIZED RESOURCES AS ANIMAL FEEDSTUFFS Scott, R. W., M. A. Millett, and G. J. Hajny. 1969. Wood wastes for animal feeding. For. Prod. J. 19(4): 14. Segel, L., L. Loeb, and J. J. Creely. 1954. An X-ray study of the decomposition product of the ethylamine-cellulose complex. J. Poly. Sci. 13:193. Sherrard, E. C., and G. W. Blanco. 1921. Preparation and analysis of a cattle food con- sisting of hydrolyzed sawdust. J. Ind. Eng. Chem. 13:61. Simonsen, E. 1898. Vorlaufige resultate der Fabrikmassigen versuche mit darstellung von Spiritus aus Sagespahner. Zeit. Angew. Chem. Part 1:962. Part II:1007. Singh, M. 1978. Utilization of whole aspen tree material as a ruminant feed component. Ph.D. Thesis. South Dakota State University. Singh, M., and L. D. Kamstra. 1981a. Utilization of whole aspen tree materials as a roughage component in growing cattle diets. J. Anim. Sci. 53:551. Singh, M., and L. D. Kamstra. 1981b. Utilization of aspen trees as a ruminant feed component. Proc. S.D. Acad. Sci. 60:54. Singh, M., L. D. Kamstra, J. A. Minyard, D. E. Moore, and R. Healy. 1978. Winter feeding demonstration with pregnant stock cows. Proc. S.D. Acad. Sci. 57:46. Slyter, A. L., and L. D. Kamstra. 1974. Utilization of pine sawdust as a roughage substitute in beef finishing rations. J. Anim. Sci. 38:693. Society of American Foresters. 1979. Forest Biomass as an Energy Source. Washington, D.C. Springer, E. L., L. L. Zoch, Jr., W. C. Feist, G. J. Hajny, and R. W. Hemingway. 1978. Storage characteristics of southern pine whole tree chips. In Complete Utilization of Southern Pine, Charles W. McMillin, ed. Madison, Wis.: Forest Products Research Society. Stake Technology, Ltd. 1979. Pro-Cell Nutritional Report No. 1. Ottawa, Ontario. Stranks, D. W. 1959. Fermenting wood substrates with a rumen cellulolytic bacterium. For. Prod. J. 9:228. Tarkow, H., and W. C. Feist. 1969. A mechanism for improving digestibility of ligno- cellulosic materials. In Cellulases and Their Applications. Advances in Chem. Ser. 95. Washington, D.C.: American Chemical Society. Turner, H. D. 1964. Feed molasses from the Masonite Process. For. Prod. J. 14(7):282. U.S. Department of Agriculture. 1960. Wood Molasses for Stock and Poultry Feed. Report No. 1731. Madison, Wis.: USDA Forest Products Laboratory. U.S. Department of Agriculture. 1976. Utilization and marketing as tools for aspen man- agement in the Rocky Mountains. Gen. Tech. Rep. RM-29. Fort Collins, Colo.: Rocky Mountain Forest and Range Experiment Station. U.S. Department of Agriculture. 1977. The Nation's Renewable Resources An Assess- ment, 1975. For. Resource Rep. 21. Washington, D.C.: Government Printing Office. 243 pp. Van Soest, P. J. 1982. Limitations of ruminants. Pp. 336-339 in Nutritional Ecology of the Ruminant. Corvallis, Oreg.: O and B Books. White, M. S. 1979. Guides for Reducing the Incidence of Fires During Outside Storage of Sawmill Residues. Blacksburg, Va.: Department of Forestry and Forest Products, Virginia Polytechnic Institute and State University. Wilson, R. K., and W. J. Pigden. 1964. Effect of a sodium hydroxide treatment on the  utilization of wheat straw and poplar wood by rumen microorganisms. Can. J. Anim. Sci. 44:122. Woodward, F. E., H. T. Converse, W. R. Hale, and J. G. McNulty. 1924. Pp. 9-12 in Values of Feeds for Dairy Cows. USDA Bull. No. 1272. Washington, D.C. Zerbe, J. I. 1977. Wood in the energy crisis. For. Farmer 37(2):12-15.