Reindeer calves(Rangifer tarandus tarandus) were used in parallel feeding trials with two different qualities of round-bale timothy silage in September 1992 (summer) and April 1993 (winter) in northern Norway, to determine the effect of season on their food intake and ability to utilize silage of different fibre contents. The silages were prepared from first-cut (FC) and regrowth (RG) timothy. The FC silage contained (on a dry matter (DM) basis) 12·3% crude protein, 30·4% cellulose and 6·2% water soluble carbohydrates (WSC), whereas the RG silage contained 14·3% crude protein, 18·7% cellulose and 30·0% WSC. For each trial, six male calves were taken directly from a natural range and divided into two groups of three animals. One group was offered FC silage and one RG silage ad libitum, for 46 days. Mean daily DM food intake (g/kg liveweight) was not significantly different in summer compared to winter(P > 0·05), but significantly higher in the animals fed RG silage compared to the animals fed FC silage (P < 0·0001). The apparent digestibility coefficient of DM was high for both silages both in summer and winter, ranging from 74·4 to 86·6%. However, the digestibility in winter was significantly higher than in summer(P < 0·002) and the digestibility of RG silage was higher than that of the FC silage (P < 0·009). The ruminal VFA concentration was significantly higher in winter than in summer (P < 0·03) and in the animals fed RG silage (range 55·3–104·9 mM) compared to the animals fed FC silage (range 35·2–61·7 mM) (P < 0·002). In spite of the high digestibility of both the FC and RG silages, the nitrogen balance was significantly higher in the animals in winter compared to summer(P < 0·0001) and in the animals fed RG silage compared to the animals fed FC silage(P < 0·0001). Thus, utilization of grass silage in reindeer appears to be influenced both by season and by plant quality.

A dynamic model of digestion and absorption of nutrients in cattle fed sugarcane-based diets is described. There are 11 rumen state variables, and four zero pools representing absorbed nutrients. The rumen state variables represent nitrogen, carbohydrate, long chain fatty acid, microbial and volatile fatty acid pools. The zero pools relate to absorbed amino acids, glucose, long chain fatty acids, and volatile fatty acids. The flux equations are described by mass-action and Michaelis-Menten forms. Wherever possible, data derived from trials with cattle fed sugarcane-based diets were used to parameterize the model. Sensitivity analyses revealed that stability and behaviour of the model was generally satisfactory. The model was most sensitive to changes in fractional passage and substratehydrolysis rates and to the microbial maintenance requirement. Within the limited information available for comparison, the simulations agreed with observations of fibre flows and ammonia and volatile fatty acid concentrations in rumen fluid. Duodenal non-ammonia nitrogen flow was consistently underpredicted and reasons for this underprediction are suggested.

Ceramic cup samplers were used to measure nitrate leaching from grass/clover pasture in Wales to which no N fertilizer had been applied and from a predominantly grass pasture receiving 100 kg fertilizer-N/ha annually. Annual leaching losses at individual sampling points, measured over a 3-year period between 1988 and 1991, ranged from the equivalent of < 0·1 to 226 kg N/ha. All data sets were positively skewed and in four out of six cases conformed to a log-normal distribution. The marked spatial heterogeneity was attributed to the uneven deposition of N in the excreta of grazing stock but variations in soil depth and hydrology may also have contributed. Particularly large losses occurred from those areas of the plots where sheep congregated. As a result of this heterogeneity, there were large standard errors associated with estimates of mean losses from the pastures as a whole. Overall losses ranged from 13 to 24 kg N/ha per year from grass/clover plots and from 10 to 13 kg/ha from fertilized grass plots. There was no consistent relationship between relative losses from the two types of pasture. The quantity of nitrate leached appeared to be independent of stocking rate, although there was a direct correspondence between the loss from grass/clover plots and the proportion of clover in the sward. Estimates of nitrate concentrations in drainage never exceeded 5·6 mg N/I for either sward.

The changes in the buffer components and pH in the surface layer of a pig and a cattle slurry were studied in the laboratory of the Department of Soil Science, Lincoln University in 1994. The slurries were spread to a depth of 7 mm in Petri dishes open to the atmosphere. Slurry pH, total inorganic carbon (TIC = CO2 + HCO- + H2CO3), total ammoniacal nitrogen (TAN = NH3 + NH4+) and volatile fatty acids (VFA = C2-C5 acids) were determined at 8–10 intervals after 1–96 h of incubation at 10, 16 and 22 °C. A great increase in pH over the first 8 h was due to the release of CO2. If the initial TIC > TAN, pH then increased steadily but slowly from 8 to 96 h. When the initial TIC < TAN, the pH declined or did not change after 20 h incubation. The initial pH elevation rate increased with temperature and initial concentration of TIC. Calculation indicated that the NH3 partial pressure (PNH3) in equilibrium with the slurry increased and pH decreased at increasing temperature if gases could not exchange between the slurry and the atmosphere. From the open slurry system PNH3 increased with temperature during the first 1–20 h. At 16 and 22 °C the PNH3 declined to low values after 20 h, whereas at 10 °C the PNH3 remained appreciable after 20 h. This explains why high accumulated NH3 losses may occur when slurry is applied to the field at low temperatures.

Dietary calcined magnesite supplements of different particle sizes and temperatures of calcination were examined at Glasgow University Veterinary School in 1981. Balance experiments with wether sheep revealed that particle sizes < 75, 75–150, 150–250 and 500–1000 μm diameter of a feed-grade calcined magnesite all increased urinary output of magnesium to a similar extent but the apparent magnesium availability coefficient for the 500–1000 μm diameter fraction (0·03) was significantly less than for fractions of smaller diameter (0·17–0·23) (P < 0·01). A 1000–2000 μm fraction of fertilizer-grade material had an apparent availability of 0·18 but had the least effect on urinary magnesium output. Calcination temperatures of 800, 900 and 1100 °C significantly increased apparent availability (c. 0·46)compared with that for temperatures of 500 and 650 °C (c. 0·12) (P < 0·01), and significantly increased urinary output of magnesium. Losses of magnesium from these supplements incubated in the rumen of cows at grass in 24 μm mesh nylon bags showed some correlations with their apparent availability and urinary magnesium output. Solubility in molar ammonium nitrate showed some good correlations with urinary magnesium output but not with magnesium availability. Supplementation of a low magnesium diet given to lactating ewes with fine particle (< 75μm) calcined magnesite resulted in significantly greater increases in plasma magnesium concentration than when coarse particle material (500–1000 μm) was given (P < 0·05), but magnesites calcined at 650 and 800 °C induced similar changes in plasma magnesium.

As part of a long-term breeding programme aimed at improving the performance of Omani goats and sheep, a 3-year performance trial was carried out between 1991 and 1993 to evaluate the reproductive and productive performance of three breeds of goat: Batinah (BT), Dhofari (DH) and Jabal Akhdar (JA), and one breed of local Omani sheep. In each year, flushing started on 1 September, mating started on 1 October and terminated on 15 November, and the kidding/lambing season started on average on 1 March. Kids and lambs were weaned at c. 3 months of age, then put on a 3-month performance test, fed on concentrates ad libitum and Rhodes grass hay. Some 3706 dam and 3011 progeny performance records were available over a 3-year period from 1991 to 1993. Ewes were lighter in body weight and lower in litter size weaned but higher in fertility than the does. Although the lambs had lighter birth weights, they were heavier at weaning and at 6 months of age. Among the goats, the DH does had lighter body weight but higher litter size weaned than both the BT and the JA goats. DH kids were somewhat lighter at birth, weaning and 6-month body weights than the other two breeds but had higher survival.

The sheep had higher output per ewe exposed (3·28 kg, 33%) and per ewe lambed (3·27 kg, 26%) and were 35% more efficient in production than the goats. Although the DH goats had a lower output per doe exposed than the BT and JA goats (0·9 kg, 9 % and 0·76 kg, 7% respectively) and per doe kidded (1·11 kg, 9% and 1·29 kg, 10%), they had higher efficiency of production than both BT (0·05 units, 6%) and JA (0·11 units, 14%).

In view of these results, attention should be paid to the sheep industry through the introduction of promotion schemes. The DH goats would be even more useful if involved in a crossbreeding scheme and the JA goats would benefit from selection for maternal care.

Pure swards of chicory (Cichorium intybus) and perennial ryegrass {Lolium perenne)were grown at Palmerston North, New Zealand. They were cut daily and fed fresh at 2 kg dry matter (DM)/day to ten hand-reared rumen fistulated castrated red deer stags kept in metabolism crates in April and October 1994. The efficiency of particle breakdown during the time allowed for rumination (<C.PART>) to below the critical size required to leave the rumen (passage through a 1 mm sieve) and jaw activities (i.e. eating and ruminating) were measured. Total eating time and the number of eating bouts were similar for deer fed each forage, but deer fed chicory had a greater chewing rate during eating (97·4 v. 81·0 chews/min), and a higher number of chews/g DM eaten (36·2 v. 31·5). Deer fed chicory had lower total ruminating time (30 v. 257 min/22·5 h), lower number of boli ruminated (38 v. 440/22·5 h), lower number of rumination bouts (5·4 v. 16·2/22·5 h) and less chews per minute ruminating (16·5 v. 44·3) than those fed perennial ryegrass. Of the ten deer used to measure (<C.PART>), only four ruminated when fed chicory compared with nine when fed perennial ryegrass.

Deer fed chicory had a higher efficiency of particle breakdown (<C.PART>; 0-64 v. 0·42), higher fractional degradation of particles > 1 mm (9·2 v. 5·1 %/h) and faster fractional disappearance of total DM from the rumen (10·2 v. 5·3 %/h). All three measurements for chicory were similar in deer that did or did not ruminate, but with perennial ryegrass, all values were considerably reduced in the deer that did not ruminate.

It was concluded that chicory can be broken down faster in the rumen, with less rumination being required than for perennial ryegrass, and that some deer (60%) could break down swallowed chicory to below the critical particle size without ruminating at all. The faster clearance of DM from the rumen explains the high voluntary feed intake (VFI) of deer grazing chicory. Future research needs to be done to partition rumen fractional disappearance rate into its components, rumen fractional degradation rate and rumen fractional outflow rate in deer fed chicory and perennial ryegrass.

A previously described model of digestion in cattle fed sugarcane diets was applied to predict nutrient supply to the host animal from dietary intake, to indicate pre-experimentally the suitability of various supplements to enhance the milk production of dairy cattle in the tropics fed sugarcane–based diets. Potential milk production was calculated according to simulated energy and lipogenic, glucogenic and aminogenic nutrient availability. The addition of small amounts of urea (10 g/kg sugarcane fresh weight) improved microbial efficiency and degradation of substrates in the rumen, but further increases could only be achieved by supplements containing rumen degradable protein. The simulated level of absorbed amino acids was low in comparison with the level of absorbed energy and glucogenic substrates. This low amino acid availability was the factor most limiting milk production on sugarcane diets and supplements which increase the amino acid availability in the gastrointestinal tract are recommended. At increasing supplementation or intake levels, and once amino acid requirements were met, an increased absorption of energy and long chain fatty acids could enhance milk production. These suggestions should be evaluated experimentally in order to improve the model and ultimately, to result in an improved system to evaluate feeds, based on the supply and utilization of individual substrates.

Genetic variation was analysed in cattle from the central area of Spain, in 1993, for the following microsatellite loci: CYP21, BOVTAU, ETH131, ILSTS002 and ILSTS005. The breeds studied were Morucha (n = 104), Sayaguesa (n = 60), Brown Swiss (n = 90) and two different populations of Avilena-Negra Iberica (n = 134 and n = 104). Morucha cattle showed the largest number of different alleles (66), whereas Sayaguesa exhibited the lowest (38). Values of number of observed alleles, gene diversity, polymorphism information content and effective number of alleles indicated that the microsatellite showing the highest variability was CYP21, followed by ETH131, ILSTS002, BOVTAU and ILSTS005.

The starchy seeds of jackfruit (Artocarpus heterophyllus Lam.) contain 124 g crude protein and 740 g total carbohydrates per kg. The possible replacement value of jackseed meal (JSM) for maize in chick diets was evaluated in Peradeniya, Sri Lanka between 1986 and 1989. In Expt 1, the inclusion of 125 and 250 g raw JSM/kg diet severely depressed growth performance and increased mortality in chicks. Organ weights, relative to body weight, were also affected in chicks fed on diets containing 250 g raw JSM/kg diet. Subsequent analyses revealed lectins and trypsin inhibitors to be responsible for the toxic effects of raw JSM. Processing of JSM by moist-heat treatment completely destroyed these antinutritional factors. This was demonstrated by the results of Expt 2, where the feeding value of processed JSM was found to be comparable to that of maize up to 250 g/kg diet. Moist-heat treatment also improved the in vitro protein digestibility (56·7 v. 84·2%) and apparent metabolizable energy (8·97 v. 13·72 MJ/kg) values of JSM. In Expt 3, the inclusion of 500 g processed JSM/kg diet resulted in reduced broiler performance and lowered the apparent dry matter digestibility and apparent energy utilization of the diets. The non-starch polysaccharide component of the JSM may have been responsible for these negative effects.

The effect of sward structure and species composition on dietary quality in cattle and sheep grazing a mesic, low quality grassland (sourveld) in South Africa was investigated over the 1990/91 and 1991/92 grazing seasons. Species composition had no measurable effect on dietary quality in either cattle or sheep but this variable could affect quality over grazing periods longer than those used in the study. In contrast, sward structure, and in particular sward greenness, had a major effect on dietary quality in both animals. Dietary in vitro digestibility (IVD) was positively correlated (P < 0·001) with sward greenness but negatively correlated (P < 0·05) with herbage mass and sward height. Dietary IVD in sheep was consistently higher (P < 0·001) then in cattle but both species reacted in a similar manner to changes in sward structure and season. Dietary quality declined over the grazing season, with IVD being highest in spring (> 60%) and lowest in winter (< 45%). Marked intra-seasonal fluctuations in quality also occurred in response to sward structure and apparent variation in soil moisture. Fire also appeared to have a significant effect on quality, with IVD being higher (P < 0·001) in the first year following the burn than in the subsequent year. It is suggested that on sourveld, dietary quality in the growing season is determined by a combination of environmental and management factors. Conversely, in the dormant season, dietary quality is largely determined by environmental factors beyond the control of the grazier.

Two grazing trials were carried out at Palmerston North, New Zealand using lactating red deer hinds in summer 1994 (Expt 1) and using weaner deer during the autumn, winter and spring of 1993 (Expt 2), to compare the feeding value of chicory (Cichorium intybus) and perennial ryegrass (Lolium perenne)/white clover (Trifolium repens) pasture for increasing the growth of deer calves. Red deer and hybrid (0·25 elk; 0·75 red deer) calves were used in both experiments. Experiment 2 concluded with slaughter at the end of spring, when the deer were c. 12 months old. In both experiments, animals were rotationally grazed on either pasture or chicory with DM allowances being 12 kg DM/hind per day (Expt 1), and 6, 6 and 7 kg DM/head per day during autumn, winter and spring respectively (Expt 2).

Perennial ryegrass comprised 62% of pasture on offer in Expt 1 and 78–90% in Expt 2, whilst chicory comprised 90–92% of forage on offer in both experiments. Relative to pasture, chicory had a higher ratio of readily fermentable: structural carbohydrate and had higher organic matter digestibility (OMD) in summer and autumn but not in spring.

Deer grazing chicory had higher voluntary feed intake (VFI), bite weight, liveweight gain (LWG), carcass dressing percentage and carcass weight and much shorter ruminating time than deer grazing pasture. Hybrid deer grew better than red deer and there were forage × genotype interactions in Expt 2, with LWG and carcass weight of hybrid deer being much greater when grazed on chicory. Carcass weight for red deer and hybrid stags was 63·2 and 73·0 kg when grazed on chicory and 56·6 and 57·0 kg when grazed on pasture. Grazing chicory advanced the date of first-cut velvet antler by 28 days and increased the weight of total harvestable (first-cut + regrowth) velvet antler. It is concluded that grazing chicory increased carcass weight, especially in hybrid stags with increased growth potential, and increased velvet antler production. This was achieved by increased VFI in all seasons and increased OMD of chicory in summer and autumn relative to deer grazing pasture. Further research is need to determine the efficiency of rumination on particle size breakdown and to measure rumen outflow rate in deer fed chicory.

The effects of species composition and sward structure on the ingestive behaviour of cattle and sheep grazing a mesic, low quality grassland in South Africa were investigated over the 1990–93 grazing seasons. Species composition had a significant (P < 0·05) effect on sheep bite size and on cattle and sheep bite rates but had no effect on dry matter intake rate (IR). Species composition could, however, affect IR over longer grazing periods than those used in the experiment.

Sward structure had a major effect on ingestive behaviour. Cattle and sheep bite rates and cattle grazing time, were negatively correlated (P < 0·05) with plant height but positively correlated (P < 0·05) with sward greenness. Bite size and hence IR in cattle and sheep were strongly correlated (P < 0·001) with plant height. Cattle IR increased from 6 to 20g/min over the range of heights encountered and appeared to reach an asymptote at a plant height of 20–25 cm. Sheep IR, expressed per unit of body mass, increased from 0·01 to an asymptote or maximum of 0·13 g/min/kg at plant heights of 10–15 cm. For sheep there was evidence of a non-asymptotic functional response at some sites with IR being maximized at certain sward heights but declining thereafter. This suggests the possible existence of a third, quality dimension to the functional response on these low-quality grasslands.

Two experiments (E1 and 2) were carried out in Northern Ireland in 1988/89 to examine the effects of a wide range of supplementary protein intakes given with grass silage-based diets on the performance and carcass composition of beef cattle, and to compare diets based on silage and dried forage. The five treatments used consisted of grass silage offered ad libitum and supplemented with 2·5 kg dry matter (DM) of barley-based concentrates containing (1) zero, (2) 0·2, (3) 0·4 and (4) 0·6 extracted soyabean meal and (5) artificially dried grass and hay supplemented with 3·3 kg concentrate DM. Experiments E3 and 4 were carried out in 1991/92 to examine the responses in ration digestibility, silage intake and animal performance to supplementary protein in relation to the requirements recommended by the United Kingdom Metabolizable Protein System. The two treatments consisted of grass silage offered ad libitum and supplemented with 3·2 kg DM of (1) barley and (2) 0·8 barley and 0·2 soyabean meal. The silages used were well preserved, containing on average 200 g DM/kg; 143 g crude protein (CP)/kg DM; 79 g ammonia-N/kg total N and 734 g digestible organic matter/kg DM. The cattle were castrated males of medium to late-maturing breed type and were initially 13 months old and 396 kg liveweight. Increasing protein intake did not affect ration digestibility, silage DM intake or animal performance in any of the four experiments, despite the fact that the effective rumen degradable protein (ERDP) intakes in the diets supplemented with barley alone were only 80–85% of ERDP requirements. However, increasing protein intake tended to increase carcass fatness in all experiments, the effect being significant (P < 0·05) for the lipid concentration in the M. longissimus dorsi muscle in E1 and 2. Offering the dried forage-based diet rather than an isonitrogenous silage-based diet did not affect carcass fatness. It is concluded that supplementing well preserved grass silages with mixtures of barley and soyabean meal, rather than with barley alone, did not affect ration digestibility, silage intake or animal performance but tended to increase carcass fatness, especially marbling fat, despite the fact that ERDP intakes for the diets with the lowest protein concentration were only 80–85% of requirements given by the United Kingdom Metabolizable Protein System.

The efficiency with which llamas and sheep digest various roughages was compared in France in 1993. Six llamas (three of which were rumen-cannulated) and six rumen-cannulated sheep were simultaneously given three different diets: (i) a grass hay low in N content (HLN), (ii) the same hay with barley, representing 25% of the total dry matter (HLN+B) and (iii) a cocksfoot grass hay, rich i n nitrogen (HRN). Daily forage dry matter intake was slightly, but not significantly, lower in llamas (55·9 g/kg W0·75 ν. 60·3 g/kg W0·75 in sheep). Intake behaviour was similar for the two species except that the number of rumination periods was lower with llamas (P < 0·05). With all diets, the digestibility of both organic matter and cell walls (estimated from neutral detergent fibre) was higher in llamas (Animal Nutrition, 9700 Angra do Heroismo, Azores, Portugal 5+6 and +6+5 units respectively, P < 0·001). The higher digestibility in llamas may be related to the slightly lower level of intake and to the lower dry matter turnover rate in the forestomachs compared to sheep (3·9%/h ν. 5·3%/h, P < 0·01). However, in sacco degradation of both hays were higher in llamas' forestomachs after 24 h of incubation (+6+5%, P < 0·01) and 48 h (+4+3%, P < 0·05) than in sheep. The high microbial efficiency of the llamas can be ascribed to the physical and chemical characteristics of their forestomach contents, which had a higher liquid turnover rate (+1+4%/h, P < 0·05), a lower volatile fatty acid concentration (P < 0·01), a higher and more constant pH (P < 0·01) than in sheep and a greater buffering capacity when pH was close to neutral. These results suggest the greater ability of llamas to control the physicochemical conditions of their forestomach contents to digest cell walls efficiently and to minimize the negative effects of concentrate supplementation.

This group, which is concerned with the applications of mathematics to agricultural science, is sponsored by the Biotechnology and Biological Sciences Research Council. It was formed in 1970, and has since met at approximately yearly intervals in London for one-day meetings. The twenty-eighth meeting of the group, chaired by Dr. D. A. Rose of the Department of Agricultural & Environmental Science, University of Newcastle, was held in the Wellcome Meeting Room at the Royal Society, 6 Carlton House Terrace, London on Friday, 29 March 1996, when the following papers were read.

An experiment was conducted at Palmerston North, New Zealand, to determine the effect of condensed tannins (CT) on the true and apparent digestion of methionine and cysteine in the small intestine (SI) of sheep fed fresh Lotus comkulatus. The lotus contained c. 30 g total CT/kg dry matter (DM) and was fed hourly to sheep in metabolism crates. Four sheep were prepared with rumen and abomasal cannulae which enabled the indigestible liquid phase marker, chromium ethylene diamine tetra-acetic acid (Cr-EDTA), to be infused into the rumen to estimate digesta flow. True digestibility of plant methionine and cysteine in the SI and their site of absorption in the SI were determined from 35S-labelled L. corniculatus homogenate continuously infused into the abomasum. After 9 h infusion of the 35S-labelled lotus homogenate, the sheep were slaughtered and digesta samples were taken at intervals along the small and large intestines. The effect of CT was determined by comparing two control sheep (CT-acting) with two sheep given a continuous intraruminal infusion of polyethylene glycol (PEG, MW 3500) to bind and inactivate the CT.

The CT reduced the true digestibility of plant methionine (0·72 v. 0·88) and cysteine (0·65 v. 0·81) in the SI relative to sheep receiving PEG. Condensed tannins also appeared to alter the site of digestion of both [35S]methionine and [35S]cysteine in the SI, and increased the flux of both amino acids in the mid and latter thirds of the SI. CT did not affect the apparent digestibility of total methionine (0·82 v. 0·84) in the SI but reduced the apparent digestibility of total cysteine from 0·77 to 0·66. In control sheep CT increased the abomasal flux (as a proportion of eaten) of total digesta methionine (0·88 v. 0·76) and total digesta cysteine (0·74 v. 0·62). The apparent absorption of total methionine (plant + microbial + endogenous) was increased by the action of CT (0·72 v. 0·63 g/g eaten) but was similar for total cysteine (0·49 v. 0·48 g/g eaten) in both groups. It was concluded that CT reduced the true digestibility of plant methionine and cysteine in the SI. However, it was calculated that the action of CT actually increased the total amounts (g/g eaten) of plant methionine and cysteine absorbed from the SI, due to its effect in increasing abomasal flux.

Two experiments were done in Mexico (1992–93) to determine the response of goats to different periods of feed restriction on liveweight, organ weights and tissue composition. In each, the weight, body condition and daily dry matter intake (DM1) of two groups of 12 female, adult, non-lactating, non-pregnant, Nubian crossbred goats were recorded for 7 weeks (stabilization period; SP), to allow the animals to reach and maintain their liveweight and body condition. At the end of the SP the animals in Expt 1 (RP1) were feed restricted for 18 weeks and for 36 weeks in Expt 2 (RP2). Animals were divided into three groups receiving either 100 (FL1), 80 (FL2) or 60% (FL3) of the levels of DM1 previously observed. At the end of each experimental period all animals were slaughtered, carcass weights were recorded and viscerae were weighed.

A sample of the soft tissues was chemically analysed. There were no significant changes in body condition in either experiment (P > 0·05). None of the feeding levels had an effect on carcass weight or the chemical composition of soft tissues (P > 0·05). Liver weights (as a proportion of slaughter weight) were significantly different among goats fed at different feeding levels in both experiments. The results of this study show the adaptation capacity of adult goats to medium and long term malnutrition, and the importance of the liver as a source for the generation of energy and for sustaining protein turnover. This could be especially important for goats grazing in arid and semiarid zones where dry periods can last for months and feed availability is therefore sometimes limited.