The nutritional value of peas (Pisum spp.) in human diets has been known for many years. In contrast, there is a lack of information regarding the nutritional value of peas for livestock, particularly ruminants. Peas contain a moderate amount of high quality protein and a high level of starch. This makes peas a unique dual purpose feed, rich in both energy and protein. Pea protein is rapidly degraded in the rumen and the starch is slowly degraded. Therefore, the value of peas differs depending on diet formulation, age of the animal and processing (Marquardt and Bell 1988). The purpose of this section is to review the factors influencing the nutritional value of peas in ruminant diets.

Pea seeds contain high quality protein, and have an average crude protein content between 20-25% DM (Lalles 1993, Monti 1983). Reichert and MacKenzie (1982) reported a considerable range (14-28.5% DM in Pisum sativum L. cv. Trapper) in protein in feed peas and reported that starch accounted for most of the difference in protein content while the remainder of variation was due to lipid, NDF, soluble sugars and ash. The composition of five pea varieties reported by Christensen et al. (1998; Table 2), illustrated considerable varietal variation in crude protein and starch values. Peas are characterized by a low ADF and NDF nitrogen content compared to other feedstuffs. Pea protein is highly soluble at over 70% of CP (Christensen et al. 1998; Mustafa et al. 1998; Kossen et al. 1994; Walhain et al. 1992). Pea protein is characterized by high rumen degradability and low bypass protein values (22% NRC 1989; Mustafa et al. 1998).

The amino acid composition of peas is characterized by a high lysine and arginine content and a low methionine, cystine and tryptophan content (Leterme et al. 1990, Lalles 1993). In preruminant calves, in which the limiting amino acids are the sulfur amino acids, lysine, threonine and isoleucine (van Weerden and Huisman 1985 as cited by Lalles 1993) it was concluded with the addition of methionine calf requirements would be met.

Feed peas have a relatively high and variable starch content (27-50% DM) making them a rich energy source for animals (Table 1) (Christensen et al. 1998). Starch is the main constituent in dehulled peas (Daveby et al. 1993). Synthesis of amylose and amylopectin, the two major components of starch, increases rapidly after the first third of seed dry matter is produced (Haeder 1989 as cited by Daveby et al. 1993,). The starch in wrinkled peas contains greater amounts of amylose (60-90%) in comparison to that of smooth peas (30-45%) (Otto et al. 1997). Flatus-producing sugars are common in grain legumes but are least troublesome in peas (Bond and Smith 1988). The oligosaccarides (% dry matter) present in dry peas include raffinose (0.6%), stachyose (1.9%) and verbascose (2.2%) (Saini 1989).

The neutral detergent fiber (NDF) content of peas, which approximates the sum of hemicellulose, cellulose, and lignin present in the cell wall, was negatively correlated with the content of pea protein (Reichert and MacKenzie 1982). Because the majority of fiber exists in the form of hemicellulose and cellulose with very little lignin, the fiber content is considered highly digestible (Racz 1997 Feed Industry Guide). A modification of the van Soest method has been developed for analyzing NDF in peas and other ingredients (Chai and Uden 1998).

^zChristensenet al. 1998. SCP=soluble crude protein, NPN=non-protien nitrogen, NDICP=neutral detergent insoluble crude protein, ADICP=acid detergent insoluble crude protein, ADF=acid detergent fiber, NDF=neutral detergent fiber, ADL=acid detergent lignin, NSC=non-structual carboydrates

In ruminal fluid, untreated pea was characterized by a slow starch degradation rate and a rapid protein solubilization (Focant et al. 1990). The protein in peas is almost completely digested by ruminant animals. Degradation of the total nitrogen of SBM and field pea in the rumen of 2 cows from 0 to 48 hours was estimated at 70% for SBM and 94.7% for field peas (Aufrere et al. 1994). Pea protein is highly soluble with only 22% of CP considered bypass protein or ruminally undegraded protein (NRC 1989). Soluble protein estimates range from 40% (Aguileera et al. 1992 as cited by Corbett 1997, Feed Industry Guide) to 48.6% CP (Walhain et al. 1992). The non-soluble degradable protein fraction is estimated at 51.2% CP (Walhain et al. 1992) to 38% (Aguileera et al. 1992 as cited by Corbett 1997, Feed Industry Guide). Walhain et al. (1992) reported a totally potentially degradable protein fraction of 99.8% CP and a ruminal effective degradability of 88.3%. The ruminal protein breakdown of a pea based diet fed to dairy cows was 77%. In comparison, maize silage, meadow hay, sunflower meal, brewer's residues, maize, barley and wheat bran had ruminal protein breakdowns of 70, 62, 72, 53, 35, 79 and 66% (Grigorova et al 1990). Other researchers found corresponding values for the in situ ruminal degradability (5% outflow rate) of peas (62.4%), maize meal (37.1%) and wheat bran (44.6%).

The energy content of peas fed to ruminants is high (Table 1), and is comparable to the energy value of wheat. The starch content of peas ranges from 41-50% DM, with 35-50% of that starch being soluble. The non-soluble rumen degradable fraction is characterized by its slow degradation rate (Walhain et al. 1992; Robinson and McQueen 1989 as cited by Corbett 1997, Feed Industry Guide). Cerneau and Michalet (1991) investigated pea starch degradability in Holstein cows. Starch degradability of barley, oats, wheat bran, peas and maize was 98, 97, 96, 90 and 58 %, respectively. Maize and pea starches were degraded at the same rate as non-starchy components. Barley, oats and wheat starch were degraded more rapidly than the other DM components. Cerneau and Michalet (1991) concluded that part of the variation could be contributed to differences in particle size. Aleksic et al. (1999) found in situ (5% rumen outflow rate) pea, maize and oat starch digestiblities of 75, 20 and 61%, respectively.

Intestinal digestibility of rumen undegraded pea protein was measured using two cows and was reported to be 92.5% and DM digestibility was 76.0% (Frydrych 1992). Intestinal digestion of protein was similar to maize (94.6%) or rye (94.4%) (Frydrych 1992).

Much research is being done in using alternative sources of plant protein to replace costly milk protein in the diets of pre-ruminant calves (Madrigal-Ambriz et al. 1992; Lalles 1993; Molchanov et al. 1983). Soybean protein is the most common alternative for veal calves, but pea protein also has potential for this use (Kolar and Wagner 1991 as cited by Lalles 1993). Peas contain approximately one-half the protein as soybean, but the protein quality is high (Lalles 1993). The first limiting amino acids for pre-ruminant calves are the sulfur amino acids followed by lysine, threonine and isoleucine (van Weerden and Huiman 1985 as cited by Lalles 1993). Therefore pea protein should be adequate as long as methionine is added (Lalles 1993). The limiting factor in incorporating pea flour into milk replacers is the high concentration of starch (Lalles 1993). This limitation is reduced with the use of pea concentrates and isolates. A pea concentrate, from alkaline extraction and precipitation of pea flour, has a lower starch concentration (2.99 vs. 39.96 g/100 g dry basis), total carbohydrates (20.91 vs. 53.87), crude fiber (0.77 vs. 11.92), and ether extract (2.58 vs. 3.55) and a higher level of crude protein (68.19 vs. 27.39), ash (7.55 vs. 3.27) and DM (92.73 vs. 91.80) than pea flour (Madrigal-Ambriz et al. 1992). Pea isolates are characterized by a further reduction in starch and oligosaccharides.

Results using pea proteins are variable and seem to reflect the age of the animal, processes used before incorporation, the level included in the milk replacer and the level and type of anti-nutritional factors present (Table 3). Milk substitutes containing peas are less readily accepted than skim milk replacers by young calves (Mbugi et al. 1989; Drean et al. 1995). Replacing skim milk powder with legume proteins reduces calf growth in proportion to their incorporation rate (Troccon and Toullec 1989 as cited by Lalles 1993; Kvasha and Gritsai 1990). In 5- to10-week-old calves, when 34% of the protein was replaced with raw pea flour, gains decreased from 0.66 kg/day in week 1 to 0.11 kg/day by week 4. Calves fed the all-milk diet gained 0.97 kg/day (Bush et al. 1992; Bush et al 1991). In another experiment with 50 day-old calves, replacing 16.5% milk protein with dehulled raw pea flour significantly reduced weight gain compared to the control (milk protein) (1145 g/day vs. 1366 g/day). However, when flaked pea flour replaced raw pea flour at 16.5% of the milk protein, there was no difference in gain compared to the all milk control (Drean et al. 1995). The low tolerance to the raw pea flour appeared to be due to anti-nutritional factors since flaked pea flour, in which most anti-nutritional factors had been destroyed, had no negative effect on the growth of calves. In 0- to 45-day-old calves, 30% of the milk protein was replaced with pea protein concentrate (80% CP). Gains did not differ from the skim milk control, but at 60% they were significantly reduced (Mbugi et al. 1989). When milk proteins were replaced with either raw or flaked pea protein, pancreas weight decreased significantly (16-18%), and amylase-specific activity increased significantly (43%) (Drean et al. 1995) indicating a change in digestive processes.

Pea protein concentrate and gelatinized pea flour have been reported to have a low digestibility in preruminant calves (Bell et al. 1974 as cited by Bhatty and Patel 1983; Prado et al. 1989; Kirilov et al. 1988). Pea isolate has a greater digestibility, but is not equal to a skim milk replacer (Bhatty and Christison 1980 as cited by Bhatty and Patel 1983). Lower protein digestibility in comparison to milk replacer has been contributed to higher levels of starch and oligosaccarides, to the presence of resistant proteins or peptides (Bhatty and Patel 1983; Spencer et al. 1988) and to anti-nutritional factors that are difficult to destroy completely with common technology (Lalles 1993). Extruded or heated peas have a higher fiber digestibility and feeding value than raw peas in starter diets (Kolobov 1985; Natsyuk 1985; Kirilov et al. 1988). Inclusion of pea proteins cause changes in digestion including increased digesta flow rates (Seegraber and Morrill 1986 as cited by Bush et al. 1992; Prado et al. 1989), increased volumes of digesta and decreased nutrient digestibility (Bush et al. 1992).

a Do Prado et al.1989 as cited by Lalles 1993;
b Bush etal. 1991 as cited by Lalles 1993;
c Mbugi et al. 1989;
d Gwenola et al. 1995;
e Bush t al. 1992.

Bell et al. (1974 as cited by Mbugi et al. 1989) reported that a milk replacer with pea protein used at 50% in the milk replacer for young calves was 25% digestible by calves < 2 weeks old and 65-70% digestible by 3 weeks of age. The apparent digestibility of DM, crude protein, and energy were lower (P < 0.05) for a milk replacer containing 60% protein from pea concentrate than the all-milk diet (Mbugi et al. 1989). The total volume and DM content of digesta were increased (P < 0.01) at week 1 and further by week 4 due to the substitution of pea protein (Experiment 1) (Bush et al. 1992). Ileal digestibility of DM, organic matter, nitrogen, fat, minerals and nitrogen-free extract decreased (P < 0.01) with the substitution of pea protein. The lower digestibility of the nitrogen free extract in the pea diet resulted from its high content of raw starch (18% of DM) and from cellulose and oligosaccarides (1 and 3% DM, respectively). A large amount of starch escaping through the small intestine may have increased bacterial growth in the hindgut. The proliferation of bacteria in the hindgut tended to reduce nitrogen digestibility (83 vs. 72%) and increase pea nitrogen-free extract (53 vs. 69%) (Bush et al. 1992). The decreasing apparent digestibility of the pea protein appeared to result from the increased endogenous and bacterial protein loss. An overall increase in ileal flow of dietary, endogenous and bacterial protein was observed with the supplementation of pea protein (Lalles 1993). This was consistent with decreased digestion of legumin (Bush et al. 1991 as cited by Lalles 1993) and a fourfold increase in the flow of active trypsin at the ileum (Lalles 1992 as cited by Lalles 1993). The pea diet was lower in ileal digestibility for individual and total amino acids as well as amino acid nitrogen (Bush et al. 1992).

Results using pea protein in calf milk replacer report that nutrient digestibility, nitrogen balance and blood urea concentrations in calves were lower with diets containing pea meal compared to those containing SBM or rapeseed meal until the age of 80 d. Performance and digestibility numbers for calves older than 80 d fed SBM, pea or canola protein are expected to be similar (Namiotkiewicz et al. 1987).

Soya and pea proteins are able to induce antibody formation when given to calves (Bush et al. 1992; Hessing et al. 1993; Prado et al. 1989; Nunes et al. 1988; Nunos et al. 1988). Preruminant calves at weaning are prone to developing gut immune-mediated hypersensitivity reactions. Unlike piglets that become tolerant after a few weeks, calves remain hypersensitive for months (Lalles and Peltre 1996). The antibody response consists mainly of immunoglobulin (Ig) G isotypes, but IgE and IgA may be involved. At the first introduction of pea protein (21.6 legumin mg/g of DM) to calves the concentration of immunoreactive legumin found in the ileal digesta was much higher (1.04 and 1.27 vs. 0.03 mg/g of DM) in comparison to calves receiving the all milk diet (Bush et al. 1992). Antibodies were formed against both legumin and vicilin with a titer of 1 (start), 5.5 (7 d) and 9-9.5 at 28 d. After 28 d calves were placed on an all-milk diet and antibody titers dropped immediately. Following the second 4-week introduction to pea protein, antibody titers increased, but not until d 9 and by d 35 titers were 9-9.5, similar to the previous experiment. Between 0.5-3.0% of the legumin fed was still immunologically active in the ileal digesta and the feces (Bush et al. 1992; Prado et al. 1990). These results indicate that the effects of incorporating peas into calf milk replacer depend on the level of processing, the age of the calf, the inclusion level and the level of anti-nutritional factors.

Recently weaned calves have mature rumen function within 2-3 weeks after dry feeds are introduced (Lalles and Poncet 1990a; Lalles and Poncet 1990b). When crushed peas replaced soybean meal in the diet of 7-20 week old calves that had recently been weaned, both diets had similar effects on intake, liveweight and gain (P > 0.05; Table 4). It was concluded that the protein degradability of peas and soybean meal in the rumen had little effect on the performance of recently weaned calves (Lalles and Poncet 1990b). In fact, ruminal N degradation (73%) and OM digestibility (64%) of the pea diet was unaffected by age, but increased (47-74% for N degradation and 50.7-71.2% OM digestibility) for the SBM diet. This indicates a more rapid ruminal adaptation to the pea diet. A mature N metabolism for both diets was evidenced by similar volatile fatty acid concentrations in the rumen and blood, 65% of microbial N in the duodenal N flow and efficient microbial N synthesis (24 g microbial of N synthesized per kg truly fermented OM) (Lalles and Poncet 1990b). The amino acid compositions of feed, duodenal and ileal digesta were similar between SBM and pea meal fed calves. However, ileal methionine levels were higher (P < 0.01) in the pea meal diet than in the SBM diet. Plasma free amino acids, asparagine, valine and leucine were higher (P < 0.05) and glutamine was lower (P < 0.01) in SBM fed calves compared to pea fed calves (Lalles et al. 1990c). Other studies report no difference in nutrient digestibility, nitrogen retention or blood values when replacing peas for SBM in young weaned cattle (Szyszkowska et al. 1987).

Feeding a diet containing yellow peas or SBM to young cattle (21-120 d) did not significantly affect nutrient digestibility, nitrogen retention or carcass dressing percentage. It was concluded that the level of peas should not exceed 35% or 0.8 kg/d (Namiotkiewicz et al. 1989). In contrast, Namiotkiewicz et al. (1989) reported a slight reduction in fat and crude protein digestibility and N retention when extruded ground yellow peas replaced SBM in the prestarter/starter concentrates that were fed to 24-120 d old calves. Fiber digestibility was increased with the supplementation of ground yellow peas. This slight reduction in CP digestibility and N retention did not affect animal performance (Namiotkiewicz et al. 1989). Unlike the preruminant calf that depends greatly on the type of protein and the quantity and quality of amino acids for growth, the weaned ruminant calf relies more on the introduction of dry feed and the development of a functional rumen (Lalles et al. 1990c). Peas can act as the sole protein source for young ruminants that have a functioning rumen with little or no effect on performance

Adapted from Lalles and Poncet 1990b.

Throughout the literature there was no indication that feed intake decreased as a result of pea supplementation. Feed intake of late-lactation (200 d in milk) dairy cows was not affected by substitution of peas for 33, 67 or 100% of SBM in the control diet (Khorasani et al 2001). In fact, DM intake of oat hay and grain by four lactating Friesian cows was higher (P < 0.05) when the cows were given legume grains rather than barley grain (8.6 ± 0.32 kg/day-peas vs. 6.6 ± 0.71 kg/day-barley) (Valentine and Bartsch 1987).

A growing body of evidence exists regarding the use of feed peas in diets for both high-producing and late-lactation dairy cows. Due to the lower effective degradability of crude protein in peas compared to soybean/canola meal (12 vs. 28% Khorasani et al. 1992 as cited by Corbett et al. 1995) and a lower undegradable protein content relative to soy/canola meal (22 vs. 35 and 28%), milk production may decrease in early lactation when the demand for undegradable protein is high (NRC 1989) (Corbett et al. 1995). In some studies this finding has been confirmed and the reduction in milk production is attributed to the greater degradation of pea protein in the rumen (Khasan et al. 1989). However, other studies indicate that dairy rations balanced to meet the bypass protein requirements of dairy cattle are able to effectively utilize peas.

Dairy production studies from the University of Saskatchewan reported no differences in high producing dairy cows (41 kg/d) fed raw pea, SBM, or micronized pea rations. The rations consisted of 47% concentrate and 53% forage (DM) and the forage contained 75% barley silage and 25% second cut alfalfa hay. Barley was the only grain in the concentrate (48-64%). The concentrate also consisted of 3% molasses, 1.3% canola oil, and mineral and vitamin premixes. Soybean meal was used to adjust the protein content of the ration to exceed the NRC (1989) requirements by 10%. Milk yield was maintained at 41 kg over three months with no difference between treatments in milk fat % or milk fat yield or protein % or yield (Jackman MSc. thesis as cited by Christensen and Mustafa 1998). Peas can be used in combination with other protein sources or as the sole protein source in well balanced diets for high producing dairy cattle.

Results from another feeding study with peas in a high producing Holstein dairy herd (31.3 kg/d) are located in Table 5 (Corbett et al. 1995). Milk production for the entire experimental period was higher (P < 0.05) for cows fed the soy/canola meal supplement (32.1 kg/d) than for cows fed the pea supplement (30.5 kg/d), but fat corrected milk was equivalent for both groups. In early lactation, fat corrected milk was lower for cows fed the soy/canola supplement (29.7 kg/d) compared to the cows fed the pea (31.3 kg/d) concentrate diet (Corbett et al. 1995). In mid-lactation, milk production was higher for cows fed the soy/canola supplement compared to cows fed peas. This may be due to decreased milk production (17% decrease) of the primiparous cows fed peas compared to those fed the soy/canola supplement. It was suggested that the primiparous cows' requirements for growth could be the cause for the depressed milk production in mid-lactation (Corbett et al. 1995).

Another study using 80% first time lactating cows (n=21) fed a control concentrate based on maize silage and an experimental concentrate based on peas. Milk yield (23.5 kg), milk fat (948 g) and milk protein yield (692 g/d) was similar for both groups. However, the most productive cows had a decrease in 4%-fat corrected milk (FCM) yield (P < 0.07), fat yield (P < 0.02) and fat content (P < 0.02) when they received the pea diet. They concluded that the use of peas in dairy rations is possible up to production levels of 25-30 kg of milk/day, but higher production rates will require a higher quality protein (Hoden et al. 1992).

In conclusion, peas can be used as a source of energy and protein in the diets of high-producing dairy cows, provided that the ration is balanced to meet bypass protein requirements (Jans 1993; Corbett et al. 1995; Hoden et al. 1992).

Studies have reported that peas can be substituted for soy/canola meal as a protein source for late-lactation cows (Khorasani et al 1992 as cited by Corbett et al. 1995; Corbett et al. 1995) and in commercial dairy herds with moderate milk production (23 kg/d) (Ward et al. 1989 as cited by Corbett et al. 1995; Valentine and Bartsch 1990; Jutz and Leitgeb 1989). A study using Holstein Friesian cows (n=8 for 60 d) with a high daily milk yield (29 kg) utilized peas, sunflower oilmeal and brewer's residues as the main protein sources. Mean daily milk yield was 27.7, 27.9 and 29.5 kg/d, milk protein 3.13, 3.14 and 3.17 % and milk protein yield was 866.0, 878.5 and 938.3 g/d for peas, sunflower oilmeal and brewer's yeast, respectively (Grigorova et al. 1990). Use of peas in grass silage/corn diets (202 g kg^-1 diet) for mid-lactation cows supported performance equivalent to that of a SBM-supplemented control (Petit et al. 1997).

When the requirement for bypass protein is low (for late lactating cows or a herd with moderate milk production) peas can be the sole source of protein for dairy cows. In southern Australia feeding peas instead of cereal grains to cows supplemented with a cereal hay based diet in early lactation resulted in an increase in milk yield and improved rumen pH and ammonia levels (NH₃)(Bartsch and Valentine 1986; Valentine and Bartsch 1987). However, peas are best used in conjunction with roughages such as corn or barley silage that can provide the high level of energy required for the maximal incorporation of ammonia-nitrogen into microbial protein (Wilkins and Jones 2000).

Studies using organic peas as a protein source for mid-lactation Finnish-Ayrshire dairy cows showed performance equivalent or superior to that of the non-organic barley-oat control (25.1 vs. 24.6 kg milk d^-1). Diets were not formulated to be isoenergetic or isonitrogenous, therefore the additional energy and protein supplied by the peas probably contributed to the superior performance of this ration (Khalili et al 1999).

Milk fat percentage was higher (P < 0.05) in cows fed a pea concentrate in all stages of lactation compared to cows fed a soy/canola concentrate. This was attributed to the low degradation rate of non-structured carbohydrates in peas. Previous reports suggest this prevents a depression in rumen pH (Valentine and Bartch 1987 as cited by Corbett et al. 1995; Valentine and Bartsch 1987). This maintains a more stable rumen, resulting in increased cellulolysis and a higher acetate:propionate ratio, leading to an increase in milk fat (Valentine and Bartch 1987 as cited by Corbett et al. 1995). Robinson and McQueen (1989 as cited by Corbett et al. 1995) reported a low pea starch degradation rate of 3.9-5.3% hr^-1 compared to barley starch (21.3-34.2% hr^-1) when peas were fed in a high concentrate:low forage diet. However, the University of Saskatchewan trials reported similar milk fat percentage and yield in dairy cattle fed SBM, micronized or raw peas (Christensen et al. 1998).

Milk protein percentage and yield were not affected by diet at any stage of lactation when peas were substituted for soy/canola in the concentrate and the diets were balanced for undegradable protein. This may be a reflection of a similar amino acid profile and supply to the small intestine between the two protein sources (Corbett et al. 1995).

Peas may provide better dairy production for cows fed cereal hay based diets by promoting a more stable rumen environment. Supplementing hay diets with high levels of barley grain in dairy cow diets causes major changes in rumen fermentation. This leads to digestive disorders, reductions in hay intake and losses in milk production due to the rapid fermentation of starch to volatile fatty acids and lactic acid. The result is a low rumen pH (below 5.8) and a severe inhibition of cellulolysis. Rumen bacteria normally associated with fiber digestion are almost eliminated when this occurs (Valentine and Bartsch 1987; Bartsch and Valentine 1986). Replacement of barley with peas (and other legumes) as 70% of the total ration administered twice daily to cows resulted in a rumen pH that was significantly higher three to six hours after the feed was administered. The rumen pH did not fall below 6.0 in contrast to barley (including 2% urea) fed cows. When barley was supplemented the rumen pH was below 6.0 for approximately 7 hr of the 12 hr feeding period (Bartsch and Valentine 1986).

Rumen ammonia-nitrogen concentration in the rumen of the cows offered hammermilled barley grain with 2% urea was below 5mg/100ml for 7 hours of the 12 hr feeding interval. Replacing barley with legume grains (including peas) resulted in higher ammonia-nitrogen concentrations at 0, 3, 4, 5, 6 and 8 hr in comparison to barley. Rumen ammonia-nitrogen concentrations below 5 mg 100/ml are sub-optimal for maximum bacterial protein synthesis by the less competitive cellulolytic bacteria (Pisulewske et al. 1981 as cited by Valentine and Bartsch 1987).

Studies at the University of Alberta indicated that as peas replaced up to 100% of SBM in rations for late-lactation cows, rumen pH decreased (P<0.01) and rumen acetate, butyrate, isovalerate and valerate increased (P<0.01; Khorasani et al 2001). Conversely, there were no differences in the concentration and the proportion of volatile fatty acids in Friesian cows supplemented with hammermilled lupin, pea, faba bean or barley grain. The volatile fatty acid (VFA) concentration of cows supplemented with barley and peas were not different (90.8 and 102.4 mM), respectively. The VFA proportion for cows fed barley were 59.5% acetic, 26.5% propionic and 14.0% butyric acid. The VFA proportion for cows supplemented with peas were the same for acetic (58.3%) and propionic (26.5%), but was higher (P < 0.05) for butyric (20.3%) acid. Production response and rumen characteristics suggest that grain legumes are less likely to cause problems than cereal grains when introduced suddenly into a diet (Bartsch and Valentine 1986; Valentine and Bartsch 1987).

a-b means in the same row with different letters are different (P<0.05).
^zAdapted from Corbett et al. 1995.

A two-year study out of North Dakota indicated equivalent (Anderson 1998a) or improved (P<0.01; Anderson 1999) growth rate in beef calves when wheat midds creep feed was substituted with up to 100% pea grain. Intake of pea-based creep feed can be limited to 3 lbs/day by step-wise inclusion of 80 and 160 g kg^-1 salt in the ration (Landblom et al 2001b). Peas supported excellent growth rates in creep fed calves, and the major determinant for their use was the net return per pound of gain.

Studies at North Dakota State University indicated that peas are an ideal feed for the receiving rations of newly weaned calves entering feedlots. Stress related to weaning, transportation, mixing and disease often depresses the intake of cattle entering feedlots; therefore, the CP content of receiving diets should be increased in order to ensure the reduced intake by calves meets the daily CP requirement (Fluharty and Loerch 1995 as cited in Anderson and Schoonmaker 2004). Feed peas, which were highly palatable and had high protein levels, stimulated significantly greater dry matter intake and average daily gain than either the barley (42 days; 16.53 vs. 14.56 lb/day intake, P<0.05; 3.53 vs. 3.32 lb/day, P<0.01; Anderson and Stoltenow 2003) or corn/canola components (20 days; 11.8 vs. 10.2 lb/day intake, P<0.05; 3.18 vs. 2.48 lb/day; Anderson and Schoonmaker 2004) that the peas replaced.

Growing heifers (Anderson 1998), backgrounding calves (Poland and Landbloim 1998) and finishing steers (Anderson 1999) consuming diets in which feed peas replaced the barley or barley/canola meal concentrate component had equivalent performance and carcass characteristics. Cost was the major factor limiting the use of feed peas in diets for growing and finishing calves. Conversely, a small-scale trial indicated positive effects on carcass quality, ADG and mortality (P<0.05) from replacing whole corn with up to 200 g kg^-1 peas in the diets of finishing steers (Anderson 1999).

Young bulls (120 d) were fed a diet in which 34.8% of skim milk was replaced with pea meal. Feeding lasted 637 d and there was no difference in pre-slaughter weight or dressing percentage (Natsyuk and Prikhod 1990). Black Pied bulls were fed either 28% peas or a 28% rapeseed with hay, silage and a grain mixture. Dry matter, protein, fat, fiber and N-free extract digestibilities were similar, and average daily gain and feed efficiency were unaffected (Soloshenko et al. 1990). Replacement of soybean meal with peas in a diet fed to German Simmental bulls (149 d old and weighed 192 kg) resulted in similar average daily gains, feed intakes and carcass characteristics (Schwarz and Kirchgessner 1989). Thirty Lowland Black and White bulls were fed diets supplemented with 22% pea meal or 15% rapeseed meal until they reached body weights over 400 kg. Average daily gain was 709 and 693 g/d, 6.43 and 7.01 barley units/kg liveweight gain, 720.1 and 736 g digestible protein/kg liveweight gain, 69.3 and 68.6% meat yield for pea meal and rapeseed meal diets, respectively. There were no differences in bull performance when rapeseed was substituted for peas (Stenzel et al. 1994). Inclusion of up to 400 g kg^-1 peas in the diets of Black and White Lowland bulls and heifers (28-120 d) further substantiated the equivalency of feed peas as a concentrate source for growing calves (Korniewicz 1998). Forty-nine Simmental bull calves (29 d old) were given free access to conventional maize-soyabean starter and finisher diets. Peas replaced SBM at 0, 50, 75 and 100%. Average liveweight did not differ at 125 d, but at 365 d weights were 472, 466, 417 and 442 kg, respectively. Carcass characteristics were similar for all bulls fed all diets (Pichler 1990). Peas can be used as the sole protein source for beef cattle.

Lambs (40-45 d to 180 d) initially weighing 12 kg were fed diets containing 81.4, 63.0 and 34.2 barley, 15.6, 9.5 and 0% SBM and 0, 24.5 and 62.8% peas (DM). Digestibility of the DM increased with increasing pea content of the diet-79.5, 80.5 and 84.9% and nitrogen retention was 5.99, 3.36 and 4.43 g/day with 0, 24.5 and 62.8% peas in the diet (Purroy et al. 1992). Peas and SBM were equally capable of supplying protein to growing lambs.

Peas would have a higher nutritional value if ruminal starch utilization was increased and a lower rate of protein degradation could be obtained (Walhain et al. 1992). An optimum heat treatment of peas would minimize soluble CP and maximize NDICP without a substantial increase in ADICP (Van Soest 1989 as cited by Mustafa et al. 1998). The search to find a process that will increase bypass protein and ruminal starch degradation has included moist heat treatment, extrusion, expansion, steam flaking and micronizing.

Many protein supplements consumed by dairy cattle are pelleted. Pelleting of rolled peas (3.2 mm. roll gap; 90°C, 35 s. conditioning, 5X45mm die, 10 s. residence time, 80°C pellet temperature) significantly reduced undegraded intake protein (29 vs. 65% of CP) and undegraded intake starch (38 vs. 78% of CP) values without affecting intestinal digestibility of these components (P<0.001). This was due to significant increases in the washable fraction and in the rate of degradation (k_d) for both starch and protein components (Goelema et al 1999). Conversely, work by Ljokjel et al. 2003) indicated that pelleting (ground - 3 mm. screen; 60°C preconditioning, 8 mm die, 81°C pellet temperature) numerically reduced the effective rumen degradation of protein (74 vs. 66%) and increased the effective rumen degradation of starch (56 vs. 65%).

Mustafa et al. (1998) ground field peas and treated them with moist heat at 127°C for 10, 20 and 30 minutes. Heating (0 vs. 30 min.) dramatically reduced CP solubility (788 vs. 190 g kg^-1 CP), and increased NDICP (28 vs. 137 g kg^-1 CP) without significantly increasing ADFICP (5 vs. 7 g kg^-1 CP). Moist heat significantly increased the rumen undegraded protein (10 minutes; 286 vs. 542 g kg^-1 CP) and the slowly degradable fraction (30 minutes; fraction B3; 24 vs. to 127 g kg^-1 CP) of intake protein. In conclusion, heating peas for 10 to 30 min will change the site of CP degradation from the rumen to the small intestine without reducing the total CP available for digestion postruminally (Mustafa et al. 1998).

Pressure toasting of whole or broken pea seeds (132°C, 3 min.) increased the undegraded intake protein (UIP) content from that of untreated whole peas (133 and 149 vs. 75 g kg^-1 P<0.0001), without affecting total tract digestibility of UIP. Ruminally undegraded intake starch (UIS) was increased (50.1 and 53.2 vs. 38.9%; Goelema et al 1998), resulting in increased absorption of both protein and starch from the gut (Goelema et al 1999).

Micronizing peas (115°C) enhanced the bypass value of the protein, as indicated by a large increase in the B2 fraction (73 vs. 23% of CP; medium rate of availability) and a large reduction in the B1 fraction of protein (high rate of availability; 68 vs. 10% of CP). Micronization at 125 or 135°C damaged the pea protein, as evidenced by substantial increases in ADIN values. Micronizing had little effect on pea carbohydrate fractions (Christensen et al. 1998).

Extrusion increases the nutritive value of legume seeds by decreasing anti-nutritional factors, gelatinizing starch and lowering the rate of protein degradation in the rumen (Walhain et al. 1992; Focant et al. 1990; Table 6).

Extrusion of crude peas at 140°C did not affect rumen dry matter effective degradability (83.6 vs. 79.5%), but dramatically reduced protein effective degradability (88.3 vs. 65.5% at an outflow rate of 0.06/h) in steers and bulls (Walhain et al. 1992). Conversely, Petit et al. (1997) indicated that extrusion at 140°C increased the rumen degradability of starch, but had no effect on ruminally degradable protein. An in vitro study reported a severe reduction in rumen fluid N digestion (68.8 vs. 28.6%) after 6 h incubation with rumen fluid, which resulted in faster DM and DM-N degradabilities in the rumen (Focant et al. 1990). Supplementation of extruded peas resulted in a lower rumen pH, a higher volatile fatty acid production, and a lower ammonia concentration compared to ground peas (Focant et al. 1990). Friesian heifers (n=6) fed a 40% ground or extruded pea diet reported a reduction in nitrogen solubility due to extrusion. This increased non-ammonia-N flow (+36%, P < 0.001) resulting from a higher rate of bacterial N flow (+53%, P < 0.01) and N flow (+19%, numerically higher, but not significantly) (Focant et al. 1990). Because of the high intestinal protein digestibility (24.9% of crude protein disappearing in the intestine, but the whole tract digestibility=99.8%) it was concluded that extrusion had not over-protected proteins or rendered them unavailable (Walhain et al. 1992). In one study, extrusion did not significantly increase the essential amino acid profile in the undegradable ruminal fraction (Walhain et al. 1992). In other studies, extrusion increased flow of all amino acids into the duodenum (Focant et al. 1990; Chapoutot and Sauvant 1997). It was concluded that extrusion temperatures above 140°C were unnecessary (Walhain et al. 1992).

^ZAdapted from Walhain et al. 1992.

Means in the same row with different subscripts differ. a=rapidly soluble fraction, b=slowly degradable fraction at time t, c=fractional rate constant of disappearence of the b fraction, a+b=potentially degradable fraction, ED=ruminal effective degradability

Bulls fed a 34% extruded pea/rapeseed oilmeal mixture had a 20.7% reduction in rumen degradation versus the control without affecting post-rumen protein digestibility (Schmidt et al. 1993). A study using extruded blends of whole peas and full-fat rapeseed (60:40 and 80:20 blends of peas:rapeseed) was conducted to determine the nutritive value of these feedstuffs. Extrusion decreased effective degradation in the rumen by 82, 96 and 93% for crude protein, dry matter and NDF, respectively. A large compensation in nitrogen digestion occurred in the intestine and increased amino acid delivery to the intestine (Chapoutot and Sauvant 1997).

Extrusion gelatinized pea starch and increased the degradation of starch in the rumen (Walhain et al. 1992; Focant et al. 1990; Chapoutot and Sauvant 1997). In an in vitro trial, starch digestion (30 min incubation with pancreatin; maltose yield (mg/g DM)) increased from 44.8 to 284.0 mg/g DM (Focant et al. 1990).

Expansion of ground peas at either of two temperatures (annular gap expander, 112 and 130°C) reduced the effective rumen degradation of protein (3mm die; 59, 63 and 74% of CP, respectively) and increased the effective rumen degradation of starch (77, 77 and 56% of CP, respectively; Ljokjel et al 2003). Conversely, expansion of coarsely crushed peas (3.2 mm gap rolled; 90C, 35°s. preconditioning; 114°C, 8s. residence) reduced the values for UIP (54 vs. 65% of CP) and UIS (60 vs. 78% of starch, respectively; Goelema et al 1999).

Steam flaking at atmospheric pressure is not an effective heat treatment. In one study it did not gelatinize the starch of peas, reduced ruminal digestion of N from 69 to 62% and posed no benefit to any in vivo parameters (Focant et al. 1990).

Reduction in particle size through fine grinding increased the rate of protein and starch ruminal disappearance in peas. Disappearance values for the nutrients in coarse particles (mean diameter = 2025µm) exceeded 76% after 8 hours, compared to nutrient disappearance values in excess of 94% with fine grinding (112 µm). The increase in nutrient degradation may be due to the increased area/mass ratio favouring nutrient solubilization and microbial access to substrates. Information regarding outflow rate of different particle sizes is required to assess the effect of particle size on rumen undegraded protein and starch in peas, and on protein digestibility index values (Bayourthe et al 2000).

Formaldehyde addition or condensation reactions with pea protein should be reversible under the action of diluted acid solutions and should not affect amino acid composition or post abomasal protein digestion (Antoniewicz et al. 1992). The addition of formaldehyde (20 g/kg) to peas increased rumen crude protein residues from 22.4% to 28.3% (Antoniewicz et al. 1992). Intestinal digestion estimated by in vitro or mobile bag technique almost completely compensated for the decline in the ruminal digestion (Antoniewicz et al. 1992). In another study, when peas were treated with formaldehyde the undegradable protein increased from 21.6 to 33.3 % CP. The soluble protein decreased from 50% to 30% CP at time 0 h, 75.1 to 60.9% CP at 6 h, 81.5 to 74.7% CP at 12 h and 98.4 to 94.3 at 24 h (Voigt et al. 1990).

The white-flowered, spring-sown varieties of peas in Western Canada contain the lectins, tannins and trypsin inhibitors; however, these antinutritional factors are present at nearly undetectable levels or in non-toxic forms. The rumen environment inactivates some antinutritional factors but, ironically, the protein-complexing properties of tannins may increase the rumen-bypass value for dietary proteins (Hill and Tammings 1998). The wide variation in anti-nutritional factors indicates that it may be possible to modify their levels by breeding (Bond and Smith 1988). However, most anti-nutritional factors have evolved to protect the plant against pests or diseases and reducing these may increase the plant susceptibility.

Whole-crop or inter-cropped pea silage is harvested to produce silage with a higher protein content than found in whole-crop cereal silage; however, alfalfa silage is already used for this purpose in western Canada. A study at the University of Saskatchewan indicated that pea silage contained more starch than alfalfa silage (12.9 vs. 0.6% DM), more crude protein than barley silage (17.0 vs. 10.1% DM), and energy values similar to both silages (1.50, 1.52 and 1.49 Mcal kg^-1 DM NE_L; 66.2, 66.7, and 65.8 % TDN, respectively). In situ analysis indicated pea silage DM, CP and NDF were more degradable than in barley silage; however, NDF degradability in alfalfa silage exceeded that in peas. Pea silage supported equivalent milk production in early-lactation cows when compared to barley silage. Milk protein levels for pea silage were lower than observed for alfalfa silage, but the authors noted this effect was not consistent with results reported by other authors (Mustafa et al 2000).

A mixture of field peas at the flowering stage (Pisum sativum L.) (67%) and triticale (33% on a DM basis) was harvested as silage and fed to midlactation Holstein cows for a 64-d lactation trial. Diets were formulated to contain 65% forage:35% concentrate (DM basis). The control ration contained 33.6% alfalfa silage, 33.0% corn silage, 16.6% ground corn, 6.2% soybean meal, 9.3% dried distillers grains, minerals and vitamins. The pea/triticale silage contained approximately three times the amount of acetic acid compared to typical alfalfa silage (Messman et al. 1992). Dry matter intake (22.6 kg/d) and milk yield (25.2 kg/d) for the pea/triticale silage did not differ from the control. Cows fed pea/triticale silage produced higher levels of fat corrected milk (27.3 vs. 22.1 kg/d; P < 0.03) because the milk fat percentage was substantially higher (4.59 vs. 3.35%; P < 0.01) (Messman et al. 1992). Energy use by both sets of cows was equivalent, but the partitioning of energy between body stores (13 vs. 6% for control and pea/triticale treatment, respectively) and milk production (53 vs. 63% for the control and pea/triticale treatments, respectively) was different. Cows fed the control diet gained more (P < 0.01) body weight than cows fed the pea/triticale silage. Cows consuming the pea/triticale diet may have been able to divert more energy into milk fat and less into body fat because the end products of digestion were more lipogenic than glycogenic (Messman et al. 1992).

Pea silage alone or in combination with barley and oats was examined for its nutritive value for dairy heifers. Holstein heifers offered pea silage had greater DM consumption (2.6% of body weight) and higher DM digestibility (64.8%) compared to other silages. Other studies have reported higher voluntary intake by dairy cows for whole-crop silages made from peas compared to barley silage. These findings were confirmed by weight gains and milk yields, which were highest when peas were included in the silage (Skovborg et al. 1987). In contrast a barley/pea silage led to a slightly lower DM intake, lower milk protein, lower fat content and reduced weight gain compared to the control. The negative performance was attributed to a high rumen solubility of pea protein resulting in low intestinal protein absorption (Jans 1993). No statistical differences in performance were seen in Ohio dairy cows consuming corn-alfalfa silage versus sorghum (71%) - soybean (18%) - pea (11%) silage, confirming the ability of peas to supply quality forage for dairy cattle and increase the protein content of forage mixtures into which they are sown (Weiss 1996).

A pea oat silage, fed to Sychevka bulls, ensiled at 70-80% moisture contained 13.6% protein, 4.4% and 30.5% fiber, 41.6 N-free extract (DM) and 20g/kg of digestible protein. When pea oat silage was substituted for maize silage the average daily gain, feed/gain ration and feed intake were similar (Epifanov et al. 1986). Pea-barley silage and pea haylage were compared to maize silage when fed to 14 months old Red Steppe bulls weighing 300 kg. Average daily gain, digestibility of DM, protein, fat, fiber, N-free extract and dressing percentage were similar between roughages (Devyatkin and Kovalev 1987).

Forage peas were of the highest quality and yield when ensiled at 12 versus 10 or 14 weeks of maturity in studies performed in the United Kingdom. Round bale silage made from peas benefited from the application of commercial silage inoculant, as evidenced by lower pH, ammonia-N, acetic and butyric acid concentrations and higher lactic acid concentrations, although these parameters were not assessed statistically. However, results were based on a study of limited scale, and intake and performance of lambs consuming the silage showed no significant differences (Fraser et al 2001). Silage quality was also maximized in the earliest sample harvested (14 weeks; pod swell stage) by other British researchers (Salawu et al 2002). Pea-oat silages of wax-ripeness (grain silage) contained significantly higher DM, CP and energy than silage harvested at milky-ripeness (Manciok 1998).

Whole plants of pea cv. Consort were harvested when DM content was 28.1%. At this time over 50% of the silage was seeds. Silage composition was 159 g crude protein, 200 g crude fiber and 566 g nitrogen free extract/kg (DM). The ensilability of whole-plant peas was only moderate because of high acetic acid and ethanol levels. The organic matter digestibility was 75% compared to 89% for the seeds. Energy values were 6.3 MJ net energy for lactation and 6.4 MJ net energy for growth/kg DM for the silage and 8.0 and 8.8 MJ, respectively, for seeds. The energy value and absorbable protein content of whole plant pea silage is comparable to that of medium quality maize silage (Daccord and Arrigo 1994).

Pea-wheat (3:1) bi-crop forage harvested at 30.1% DM ensiled equally well whether untreated or treated with inoculants (Lactobacillus spp.), formic acid or quebracho tannins. The silage achieved an average pH of 4.0 (Salawu et al 2001).

Forages generally contribute 40 to 50% of the DM in dairy rations (Jaster et al. 1985; Messman et al. 1992) and crude protein is usually the main criteria used to judge forage quality (Hafley et al. 1987). Stage of maturity at harvest is the most important factor influencing the composition and nutritive value of forage (McDonald et al. 1981 as cited by Jaster et al. 1985). The cell wall and crude protein concentrations of cool season forages including barley/pea, pea and oat/pea silages were reversed when harvest was delayed by 14 days (Jaster et al. 1985). Oat/pea forage DM increased by 21.1% and crude protein decreased from 20.0% to 13.0%. Winter pea (Pisum arevense, cv. Austrian) crude protein decreased 31.1% between March 29 and May 5 and an additional 28.4% decrease occurred between May and June due to advancing maturity (Hafley et al. 1987). The barley/pea forage acid detergent fiber content increased by 9.4% and acid detergent lignin doubled. As a plant grows, the need for structural tissue increases. Consequently, acid detergent fiber, neutral detergent fiber and acid detergent lignin will increase with a 14-d harvest delay (Jaster et al. 1985). Pea haylage was found to have a higher feed value than oat haylage (Yarmots 1986).

Pea meal developed from the whole plant had a crude fiber content of 23.9% and a neutral detergent fiber value of 51.5% (DM). The digestibility of crude fiber was 26% and neutral detergent fiber was 44% in a diet fed to feeder cattle weighing 250 kg. Hemicellulose was significantly more digestible than either cellulose or lignin (Lipiec et al. 1988).

An alternative approach to increasing the use of structural carbohydrates is supplementing with microbial additives that digest high fiber residue feeds such as pea/bean hay (Mpofu and Ndlovu 1994). Rumen fungi have cellulolytic and hemicellulolytic activities and white rot fungi have the ability to decompose lignin or lignocellulose with a minimum amount of hemicellulose degradation (Mpofu and Ndlovu 1994). Zadrazil (1976 as cited by Mpofu and Ndlovu 1994) discovered that yeasts act as secondary lignin and lignocellulosic degraders and use the simple metabolic products that are produced. An in vitro study reported that supplementing pea/bean hay with Saccharomyces cerevisiae (yeast) and Armillaria heimii (white rot fungi) increased (P < 0.05) neutral detergent fiber degradability of pea/bean hay by 50% (Mpofu and Ndlovu 1994).

Pea straw is a good feed alternative in beef diets and has a higher feed value than wheat straw. A recent study of pea straw collected from south-central Alberta measured CP (7.2%), Ca (1.61%), P (0.08%), ADF (51.1%) and NDF (64%) (DM basis). Environmental conditions affected the CP, ADF and NDF values of pea straw, and feed analysis was strongly recommended for any producers utilizing this feed source (Olson et al 2001).

Pea straw was reported to contain 69.5% NDF, 1.4% NDF Ash, 54.1% ADF, 15.4% hemicellulose, 42.4% cellulose, 11.6% lignin and 5.0% ash (Lee and Pearce 1984). Pea straw contained the highest proportion of lignin compared to barley straw, oat straw, ryegrass and lucerne hay after grinding and had the largest particle size (Lee and Pearce 1984). Conversely, degradability studies using rams indicated pea straw had the highest DM (63.2%) and NDF (57.1%; soluble plus slowly degraded fractions) digestibility when compared to chickpeas, lentils, horse beans or vetches (Bruno-Soares et al 2000). Pea straw was reported to contain 6.4% protein, 44-46.0% TDN, 0.60% Ca and 0.19% P (Vern Racz, Prairie Feed Resource Centre). Pea straw contains more protein and TDN than wheat straw. As long as there is no mold present the digestibility and acceptance of pea straw is high (Vern Racz, Prairie Feed Resource Centre).

Peas intercropped with winter wheat in the southern United States increased the rumen degradable nitrogen level to that required for peak rumen microbial efficiency in grazing cattle. Seeding rates (23 kg ha^-1) were less than used for commercial pea production (Redmon et al 1998).

Although scientific research has been limited, pea screenings are reported to contain 20.0% protein, 75.5% TDN, 0.11% Ca and 0.42% P. Screenings are also highly digestible and palatable and can compete with protein supplements in ruminant rations (Vern Racz, Prairie Feed Resource Center).

Peas provide an array of diverse feeds that can be used in ruminant diets. Peas are palatable, highly digestible and can provide the sole protein source for most ruminants. For highly productive animals the diet needs to be formulated to provide adequate bypass protein. Extrusion increases the nutritional value of peas by increasing ruminal bypass protein and gelatinizing the starch. Much research is warranted to establish the protein and starch degradability of raw and extruded peas, and to relate these findings to production parameters.

Mustafa, A. F. 1998. unpublished results.

Christensen, D.A., Mustafa, A.F. and McKinnon, J.J. 1998. Carbohydrate and protein characteristics of peas and canola meal for ruminants. Pgs. 14-27. In: Proceedings of the 19th Western Nutrition Conference. Saskatoon, Saskatchewan.

Racz, V. 1988. Industry report.