The digestive maturation of the young rabbit is reviewed. It indicates that their nutritional needs and those of lactating females are antagonistic in many aspects. Energy requirements of lactating females are very high, whilst a low starch, high fibre diet around weaning improves the health of the young after weaning. To solve this problem, several feeding and management strategies are presented and discussed. If weaning occurs between 28 and 35 days of age, feeding the young with a specific diet, different from the female’s, seems an effective solution. Otherwise, it is necessary to find a compromise between the needs of the litter and of the doe. Early weaning (< 26 days) could be also a promising way to provide adequate feeding for the young as soon as they begin to eat solid food.

Sheep of a line (S) selected on an index to increase lean weight and decrease fatness at an age, and a control line (C), were given a high quality food at different levels including ad libitum. Live performance was measured from about 21 to 114 kg live weight. The carcasses of each line were analysed for lean, fat and bone at three widely varying weights in both males and females. Level of feeding did not affect the extent to which S was superior to C in either the level of fatness in the carcass (0·86 as much) or the ratio of lean to fat (1·28 as much). The lean to bone ratio was slightly greater in S (1·028 of the value of C; P 0·05) and was higher on the lowest level of feeding compared with the two higher levels used (P 0·05 in one experiment on females and P 0·001 in another on males). On ad libitum feeding the S line grew 1·19 times as fast and was 1·17 times as efficient compared with C. These advantages to S decreased as level of feeding decreased to become virtually zero at the lowest level of feeding used, which allowed C to grow at only 0·53 of the rate seen on ad libitum feeding. On ad libitum feeding growth was well described by a Gompertz growth function of the form W = (Z/B) exp(-exp (G0 –B t)). The maximum growth rate is (Z/e). Line S had a value of Z that was 1·10 that of C averaged across the two sexes. A Spillman function W = W0 + (A-W0) (1-exp (-k F)) was used to describe weight, W, in terms of cumulative intake, F. It worked well for ad libitum feeding and for the two restricted regimes used. The value of the combined parameter (A k) varied across treatments in the same way as efficiency did.

The aims of this study were to assess genetic variances, trends and mode of inheritance for hip and elbow dysplasia in Finnish dog populations. The influence of time-dependent fixed effects in the model when estimating the genetic trends was also studied. Official hip and elbow dysplasia screening records of 42 421 dogs from seven breeds were analysed with REML. To investigate the mode of inheritance of hip and elbow dysplasia, trait distributions, genetic variances and regressions of offspring phenotypes on parental predicted breeding values were studied separately in males and in females. Genetic trends for hip dysplasia between the years 1983 and 1998 were favourable only in the Rottweiler. In elbow dysplasia, the trends were favourable after the year 1992 in all the four breeds studied but the overall changes were small. The reason for this seemed to be negligible selection pressure against these traits. Time-dependent fixed effects in the model had an influence on the estimated genetic trends, resulting either in a more negative or more positive genetic trend compared with the model from which the time-dependent effects were removed. Mitochondrial or sex-linked inheritance did not seem likely in the expression of hip and elbow dysplasia in the populations studied. Regression coefficients of offspring phenotypes on estimated parental breeding values were approximately equal to their expected value in a situation with equal parental contribution. Furthermore, the phenotypic frequency distributions of hip and elbow dysplasia grades were similar among males and females in each breed studied. No indication of major genes was found in the offspring frequency distributions within individual sires. According to these Finnish data, mode of inheritance for both hip and elbow dysplasia is polygenic (quantitative) with equal expression of the genes from both parents, although the estimates of heritability for hip dysplasia in the Rough Collie and for elbow dysplasia in the German Shepherd and the Golden Retriever were somewhat different in males compared with females.

The objective of this study was to examine the effect of breed, parity and body fatness on the lipolytic response of dairy cows. The lipolytic response was estimated as the plasma non-esterified fatty acid (NEFA) response to an adrenergic challenge. Four challenges per lactation were carried out on 124 Danish Holsteins, 101 Danish Red and 82 Jerseys through consecutive lactations. Within breed, there were two genetic lines. Cows were equally distributed across two feeding treatments, a normal and a low energy total mixed ration. Diet composition was constant throughout lactation. The lipolytic response was significantly affected by breed, parity and stage of lactation. Lipolytic response was greater in early lactation than other stages in lactation (P 0·001). The larger breeds had a higher lipolytic response than Jerseys (P 0·05), and showed an increase in lipolytic response between first and second parity (P 0·05), but not between second and third lactation. Lipolytic response in Jerseys was not significantly affected by parity. No line or feeding treatment effects were observed on the lipolytic response. Ultrasound measurement of the area of subcutaneous backfat was used to estimate effect of body fatness on the lipolytic response. There was an increase in lipolytic response with increasing body fatness (P 0·05) in mid lactation and the dry period. The slope of this was not affected by breed or parity. It was concluded that the lipolytic response of dairy cows, and by implication the responsiveness of the lipid reserves, varies according to breed, parity and physiological state.

A total of 77 heifers (3⁄4 Charolais crossbred) completed a study which examined the effects of parity number on growth, changes in size of the carcass, carcass composition and meat quality. The experiment had a factorial design based on the factors: parity and level of finish. The parities were: 0 (maiden), 1, 3 and 5. At each parity animals were slaughtered at medium (EU fat class 3) and high (EU fat class 4H) levels of finish. All data were analysed by analysis of variance using individual animal observations, and linear, quadratic and asymptotic trends were explored. Where appropriate, regression equations were derived using individual animal values to describe the relationships between key parameters and the parity status of the finished animal. Animals calved at turn-out to grass in spring (April/May) and the cows were housed at weaning in mid October (mean lactation length of 163 days) and all pregnant animals were given a diet of grass silage without concentrate supplementation during the winter period. The medium level of finish animals were slaughtered 24 days after commencement of the breeding programme for parity 0 or at the end of lactation for the remaining parities, except for parity 1 animals which had a short finishing period. For the high level of finish treatment, parity 0 heifers had a longer finishing period, while all bred animals had a post-weaning finishing period. The high finish animals were finished on a diet of grass silage and concentrates. Birth weights of calves increased until the fourth parity, while weaning weights increased linearly until the fifth parity. Mean daily milk yield increased until the third lactation. Live weight, carcass weight and weights of saleable beef, separable lean and separable fat all followed asymptotic patterns of rapid initial increase and then tended to plateau with increase in parity. The asymptote values for live weight and carcass weight were 723·0 and 383·2 kg respectively. Statistically significant asymptotic relationships with parity number were obtained for live weight and various tissue weights. Nominal mature weight was assumed to be 0·99 of the asymptote and nominal mature weights for live weight, carcass weight, saleable beef, separable lean and separable fat were attained at parity (age, years) 4/5 (6·1), 3/4 (5·3), 4/5 (6·2), 3/4 (4·7) and 1/2 (2·7) respectively. The proportion of high-priced joints declined with increase in parity (linear trend P 0·01) thus indicating a reduction in relative growth of the main muscle groups of the hind limb. Shear value measurements on cooked muscle from the maiden and first parity animals indicated very tender meat, but tenderness declined with increase in parity (asymptotic trend P 0·001). Taking animals to the higher level of finish effected increases in carcass weight and separable fat in the carcass of 55·1 kg (P 0·001) and 72 g/kg (P 0·001) respectively, while separable lean declined by 53 g/kg (P 0·001). Cooked muscle from the high finish animals was less tender than from the medium finish animals (P 0·05). It was concluded that slaughtering cows at the third parity would enable a self replacing closed herd to be operated, while obtaining 0·97 of the potential maximum output of saleable beef from the cow, but there may be some compromise in the tenderness of the meat for some consumers by the third parity.

A basal diet was formulated to meet the nutrient requirements of growing rabbits. Another diet was formulated by substituting 152 g/kg of the basal diet with sunflower hulls (SH diet). One hundred and sixty-eight weaned 30-day-old rabbits were given these diets and finishing performance was recorded. Eighty animals were used to study the effect of SH inclusion on caecal fermentation traits at two ages (5 and 35 days after weaning) and disaccharidase activity in the small intestine at 35 days after weaning. Inclusion of SH in the diet reduced growth rate by proportionately 0·056 in the first 2 weeks after weaning (P 0 ×001), but had no effect from 14 to 65 days after weaning. Accordingly, daily gain was lower by a factor of 0·035 over the whole finishing period (P 0×01). There was no effect of treatment on food intake during the 14 days after weaning, but SH inclusion tended to increase it from this time onwards ( +0·026; P = 0 ×06) and over the whole finishing period ( + 0·018; P = 0 ×09). This effect was parallel to a 0·09 proportional decrease in the weight of caecal contents (P 0 ×01) observed in animals of 2 kg live weight. Food efficiency was lower by a factor of 0·05 (P 0×001) in all the periods considered when SH was included in the diet. Mortality rate (6%) was not affected by treatment nor was caecal pH or caecal concentrations of volatile fatty acids and ammonia nitrogen either at 5 days (5×75, 72×7 mmol/l and 16×6 mmol/l, respectively) or at 35 days after weaning (5×70, 74×3 mmol/l and 9 ×7 5 mmol/l, respectively). Inclusion of SH increased sucrase specific activity at the ileum by a factor of 0·47 (P 0×01) but had no effect on maltase specific activity at the jejunum or ileum or on sucrase specific activity at the jejunum. In conclusion, SH included at moderate levels (150 g/kg) in the diet reduced accumulation of digesta in the caecum, which increased voluntary food intake but impaired growth rate and food efficiency. Inclusion of SH did not affect caecal fermentation or mortality.

The aim of this study was to assess the effect of weaning weight and post-weaning diet on the performance of weaned pigs. In experiment 1, 30 litters with more than 10 pigs born alive per litter were selected. At 11 days of age, pigs of average weight for the litter were removed from 15 litters so that eight pigs remained per sow. These litters were given access to creep food (16·5 MJ digestible energy (DE) per kg and 18·7 g/kg lysine). The remaining litters were left complete and were not given creep food. Pigs were weaned at 28 days of age and pairs of pigs (a male and a female littermate of similar weight) were formed from each litter (no. = 54 pairs). Pairs were blocked on the basis of litter origin and weight and assigned at random to one of the following treatments: (1) 10 kg starter diet (16·1 MJ DE per kg and 17·4 g/kg lysine) followed by link diet (15·3 MJ DE per kg and 15·0 g/kg lysine) to 27 days (high dietary regimen; HDR); or (2) 4 kg starter diet, 10 kg link diet and weaner diet to 27 days (low dietary regimen; LDR). Thereafter pigs were given common diets to slaughter at about 95·6 kg live weight. On experiment 2, four pigs (two light and two heavy) were taken from each of 32 litters (no. = 128 pigs; age = 22 days), blocked on sex, litter origin, and weaning weight and within weight category individually assigned at random to two dietary treatments: (1) high density diet (16·1 MJ DE per kg and 17·4 g/kg lysine); and (2) low density diet (15·1 MJ DE per kg and 15·7 g/kg lysine). The duration of the experiment was 26 days. In experiment 1, reducing litter size and creep feeding increased weaning weight by 0·6 kg (P 0·05). This weight advantage at weaning was lost by day 14 post weaning (P > 0·05). From day 0 to day 27 post weaning daily gain was 472 and 427 g/day (s.e.12·8; P 0·05) and food conversion efficiency was 1·26 and 1·36 g/g (s.e. 0·026; P 0·05) for HDR and LDR, respectively. In experiment 2, weaning weight was 7·1 and 5·8 kg (s.e. 0·08; P 0·01) and pig weight at day 26 post weaning was 17·5 and 15·4 kg (s.e.0·23; P 0·01) for heavy and light weight categories, respectively. In the period from day 0 to 26, food intake was 440 and 396 g/day (s.e. 8·0; P 0·01) and daily gain was 389 and 355 g/day (s.e. 8·0; P < i 0·01) for heavy and light weight categories, respectively. Weaning weight was found to be a good determinant of weight at day 26 when terms for litter origin were included (R2 = 0·67; P 0·001 for the high density diet and R2 = 0·77; P 0·001 for the low density diet). It is concluded that weaning weight can be increased by pre-weaning management but that this weight advantage is lost in the early post-weaning period. Where weaning weight was naturally higher the weight advantage was still evident at day 26 post weaning.

In the domestic environment, horses are often presented with foods to which they are not evolutionarily adapted, such as low fibre pellets. Horses may not have the ability to learn the consequences of consuming unnatural foodstuffs and adapt their selection accordingly. This study aimed to investigate the horse’s feeding preferences when presented with concentrate pellets differing in nutrient content. Using a choice test, the relative preferences of 12 horses for mint and garlic, in iso-caloric diets, was first assessed over 29 meals. A mint preference, calculated as the proportion of mint in the total food intake, was shown by 11 horses. The horses were then divided into two groups, approximately balanced on the basis of mint preference. Group A was exposed to a choice of a mint-flavoured lower energy food or a garlic-flavoured higher energy food, while group B was exposed to mint-flavoured higher energy food and garlic-flavoured lower energy food for 29 meals. Next the flavours were presented in iso-caloric foods, initially for 10 meals, then a further 40, before the flavour-energy pairings were reversed for 30 meals. A final iso-caloric test was carried out for 30 meals. Both groups showed a preference for mint in the initial iso-caloric choice test but no such preference was shown in later iso-caloric tests. Both groups showed a higher preference for mint when paired with higher energy (proportion of mint intake to total intake was 0·75 (s.e.0·02) and 0·73 (s.e.0·02) for A and B respectively). Group B also showed a preference for garlic when paired with higher energy (proportion of mint intake: 0·32, s.e. 0·02) whilst group A showed a significant decrease in preference for mint when paired with lower energy (by 0·21 (s.e. 0·03), T = 6·88, P 0·01). The results suggest that horses can select a higher energy diet over a lower energy one and that horses can form associations between foods and their nutritional composition, even if they do not resemble those found in their natural environment.

Seventy-four New Zealand ✕ Californian females were used to determine the influence of dietary energy source on the performance and perirenal fat thickness (PFT) change measured by ultrasound of primiparous rabbit does. Three highly concentrated diets were designed with different dietary energy sources: animal fat (diet F), vegetable oil (diet O) and cereal starch (diet S), and does were fertilized by artificial insemination after parturition (AP group) or around weaning (AW group), litters being weaned at 28 days of age. There were no significant differences in live weight and food intake of females throughout gestation, but does given diet S showed a higher increase in their PFT until 28th day than those given diets F and O (P < 0·01), although all females had a similar PFT at parturition. Rabbit does given diets S and O showed a significantly higher dietary energy intake than those given diet F (P < 0·01) during the first weeks of lactation, and similar thereafter until the second parturition. Milk yield and litter weight gain were significantly higher with fat-enriched diets (P < 0·05), and milk composition of rabbit does given diet F showed a higher total solids content (P = 0·01), fat (P < 0·001) and energy (P < 0·01) than milk of does receiving diets O or S. Although pups on fat diets showed a higher weaning weight (P < 0·05), dietary treatment during lactation did not have any effect on the performance of pups during the subsequent growing period. PFT always decreased during lactation for does given fat-enriched diets, while it increased for does given diet S and fertilized after parturition. Live weight of AP does was significantly higher at 21st day of lactation (P < 0·01) and at weaning (P < 0·001), and they showed a significantly lower food intake during the final week of lactation. The period from weaning to parturition was too short for AP does, which showed a decrease of their PFT, while AW does showed an increase of their PFT and a greater number of pups alive at birth (P < 0·05). Due to the lower milk yield of rabbit does fertilized AP, their pups showed a significantly higher solid food intake (P < 0·001), resulting in a similar value for the live weight of pups in both groups at weaning. The earlier promotion of solid food intake significantly affected the food intake of litters during the subsequent growing period, presenting higher values for pups coming from AP does. In conclusion, the addition of dietary fat, especially from animal origin, seems to improve the utilization of energy for milk production, while the use of a high level of starch could decrease the negative balance of primiparous rabbit does during lactation.

The plasma concentrations of cortisol and progesterone during reproductive maturation in the heifer calf were examined. Six beef heifer calves were handled and bled every 2 weeks (control), 30 were left unhandled (naive). At 13, 21, 30, 39 and 47 weeks of age, a different group of naive heifers and the control heifers were bled every day for 5 days (puberty seen at 57·4 (s.e. 1·3) weeks). Thirty-nine dairy heifer calves were allocated to seven groups to receive adrenocorticotropic hormone (ACTH) at either 4, 7, 9, 16, 26, 36 or 46 weeks of age (puberty expected around 43 weeks of age). Plasma concentrations of cortisol increased at 21, 30 and 39 weeks of age in frequently handled and naive beef heifers; the increase was greater at these ages in the naive beef heifers (age and treatment P < 0·01). No age trend was apparent for plasma concentrations of cortisol and progesterone prior to ACTH injections (t = 0) in frequently handled dairy heifers. Adrenal progesterone secretion did not change with age in beef heifers, but naive beef heifers had greater plasma concentrations than frequently handled beef heifers. ACTH induced cortisol release in dairy heifers as early as 4 weeks of age (P < 0·05), but the response was greater in dairy heifers 16 weeks old and older (P < 0·05). A progesterone response to ACTH in dairy heifers was not seen until animals were 9 weeks old. It is concluded that as heifers mature reproductively, there is a parallel increase in the sensitivity of the adrenal gland to ACTH and handling stress.

Efficacy of oestrous synchronization and post-treatment fertility was studied in a herd of 149 dairy goats (Swiss breed no. = 104, indigenous Greek breed no. = 45) kept under an extensive farming system. All does were synchronized during the transitional period (during July prior to the breeding season) using fluorogestone acetate (FGA) impregnated intra-vaginal sponges, equine chorionic gonadotropin (eCG) and different prostaglandin F2a (PGF2α) analogues. Four doses of prostaglandin F2a (group 1: 4·5 mg luprostiol; group 2: 2·25 mg luprostiol; group 3: 0·2 mg cloprostenol; group 4: 0·1 mg cloprostenol) were administered i. m. 48 h prior to the end of progestagen treatment and 400 IU eCG was administered at sponge removal. All does were inseminated with fresh semen (300 ✕ 106 spermatozoa per insemination) 42 to 44 h after sponge removal. There were no significant differences (P > 0·05) in the number of goats showing ovarian activity (blood progesterone concentration ³ 1 ng/ml) prior to oestrous synchronization between PGF2α treatment groups. No significant differences (P > 0·05) were observed for the kidding rates between the treatment groups. Our results indicate that goats with ovarian activity prior to sponge application who received high doses of PGF2α showed lower conception rates than those detected in anoestrus (P 0·05). No differences (P 0·05) were observed in conception rates between groups that received lower doses of PGF2α. July is considered to be a transitional period suitable for oestrous synchronization for both Swiss and indigenous dairy goats. Natural mating during the two subsequent cycles after artificial insemination improved kidding rate, but without any significant difference either between treated groups or between Swiss and indigenous breeds.

The feeding of housed lambs on conserved forages and pelleted rations is accompanied by a high risk of chronic copper (Cu) poisoning (CCP) which might be reduced by selecting sires for low liver Cu status. Livers were therefore retrieved from Suffolk, Texel and Charollais lambs, slaughtered during the course of a performance trial, to ascertain sire and possibly breed effects on the rate of Cu accretion in the liver. In total, 160 livers were obtained, 100 from Suffolk, 40 from Texel and 20 from Charollais lambs, the progeny of 14, eight and eight sires, respectively. Lambs came from three separately managed flocks but were brought together at 8 weeks of age, weaned onto a common complete diet containing 6·1 mg Cu per kg dry matter (DM) and offered ad libitum. One-fifth of each breed group was slaughtered at 14, 18 or 22 weeks and the remaining 40% at 26 weeks of age. Mean (s.e. ) liver Cu concentrations at those ages were 3220 (450), 4639 (464), 6426 (468), and 6513 (370) µmol/kg DM for Suffolk, and 5843 (811), 6579 (857), 8017 (811) and 10406 (589) µmol/kg DM for Texel, respectively. The pattern of liver Cu accretion differed, the Suffolk starting at a low value yet reaching a plateau at about 22 weeks of age (significant quadratic regression coefficient), the Texel, continually increasing from a high initial value at an average rate of 53·7 (s.e.10·6) µmol/kg DM per day. There was a significant effect of sire on liver Cu in the Suffolk (P 0·05) with a heritability of 0·85 (s.e. 0·44); in the Suffolk and Texel combined, the heritability was 0·60 (s.e. 0·33). The data available on the Charollais were too limited to test for sire effects but at 26 weeks of age, where most information was available, the mean liver Cu concentration was 7285 (s.e.826) µmol/kg DM. At a given age, food intake, liver weight and live weight were each lowest in the Texel but when expressed as a proportion of live weight (LW), both food intake (43 g/kg LW) and liver weight (5·15 g DM per kg LW) were similar among breeds (P 0·05). Thus, differences in liver Cu accretion are unlikely to reflect differences in Cu intake per unit liver weight. There was a tendency for liver size per kg LW to decrease as liver Cu rose in the Texel but not in the Suffolk. Continued hepatic Cu accretion in the Texel may reflect a breed-specific inability to cope with Cu overload. Increases in liver Cu to marginally toxic levels in some Suffolk, some Charollais and most Texel lambs, and to a level commonly associated with toxicity in one Texel lamb, on a ration of moderate Cu concentration highlights the difficulty of controlling risk of CCP by manipulating dietary composition. The current EC limit for Cu in ovine diets, 17 mg Cu per kg DM, is clearly too high for the breeds and dietary conditions used in this study. However a safe limit would be hard to achieve and hence the need to exploit sire variation in propensity to accumulate liver Cu to reduce disease risk.

The influence of naturally acquired nematode infections on food intake and utilization was investigated in grazing lambs. It was first demonstrated in a pilot study with penned lambs that the application of chrome-oxide in a ruminal bolus could be used for comparison of food intake in infected and uninfected lambs on grass. In a field trial, repeated in two consecutive seasons, 48 lambs grazed on clover grass in six groups. Three stocking rate groups were grazed on contaminated pastures (I-low, I-medium and I-high) and three groups on clean pastures (U-low, U-medium and U-high). Faecal dry-matter (DM) output of individual lambs was estimated in August each year by the chrome-method. DM intake and food conversion efficiency were calculated based on hand-picked herbage samples and in vitro assessment of digestibility. Substantial worm burdens, increasing from year 1 to year 2, were observed and some lambs had clinical symptoms in the 2nd year. In year 2, lambs on contaminated pastures had significantly lower DM intake (proportionately 009 lower), higher faecal protein losses and lower food conversion efficiency compared with the uninfected lambs on clean pastures. The effect of infection on food intakes was eliminated if corrected for body weight. The study showed that if exposure to nematode infections is large enough, increased protein losses in faeces and anorexia may contribute to reduced performance in lambs in late season.

The first objective of this experiment was to investigate the effect of using different bacterial isolates on the estimation of microbial protein production in semi-continuous fermenters (RUSITEC) given four diets, and to test whether a ‘mixed’ bacterial pellet was representative of the whole bacterial population in the fermenters. A second objective was to compare two different microbial markers (nitrogen-fifteen (15N) and purine bases (PB)). Diets consisted of neutral-detergent fibre from grass hay (10 g/day) and sugar-beet pulp (2 g/day) and 280 mg/day of one of four N forms (isolated soya-bean protein, soya-bean peptides, amino acids blended to profile soya-bean protein and NH4Cl). Two 14-day incubation runs were carried out and in each run each of the four different diets were given to two vessels. On days 12 and 13, total digesta (effluent plus nylon bags residues) was collected for analyses of non-ammonia N, 15N enrichment and PB concentration, and for isolation of total mixed bacterial pellets (TB). On the last day of each incubation run, the system was stopped for isolation of liquid- (LAB) and solid-associated (SAB) bacteria. Microbial N flow was estimated from the 15N enrichment and PB concentration in both total digesta and in the three different bacterial pellets (TB, LAB, and SAB). For all diets, LAB presented a greater (P < 005) 15N enrichment and PB: N ratio than SAB, with TB having an intermediate value. For both markers, the use of LAB produced the lowest (P < 005) estimates of microbial N flow and the use of SAB produced the greatest (P < 005) estimates. The use of TB produced intermediate values with all diets, suggesting that TB consisted of SAB and LAB. For all bacterial pellets, PB produced greater (P 005) values of microbial N flow than 15N. However, there was a positive relationship (r = 0·883; P 0001; no. = 15) between the values of microbial N flow determined with the two markers when TB were used as reference.

Ruminal bacterial counts and in vivo digestibility were determined on four Mediterranean buffalo bulls and four Friesian bulls, all fistulated at the rumen, and given at maintenance level (50 g/kg M0·75 per day of dry matter) four different diets with the same crude protein content (N ✕ 6·25 = 140 g/kg dry matter) and with forage: concentrate ratios as follows: diet D12·5 = 0·875: 0·125; diet D25·0 = 0·75: 0·25; diet D37·5 = 0·625: 0·375; diet D50·0 = 0·5: 0·5. All the animals received the diets during four consecutive periods in a Latin-square design. Buffaloes had higher total microbial counts (10·78 v. 10·08 log10 cells per g dry rumen content, P < 0·01) as compared with cattle; differences in total ruminal bacterial counts among the diets were only observed within the buffalo species (diet D12·5 v. diets D25·0, D37·5, D50·0: 10·04 v. 10·92, 10·98, 11·17 log10 cells per g dry rumen content, P 0·01) and when comparing the two species for each diet, significantly higher values for bacterial counts in buffaloes were found for diets D25·0: 10·92 v. 10·28 (P 0·05), D37·5: 10·98 v. 10·08 (P 0·01) and D50·0: 11·17 v. 9·76 (P 0·01) log10 cells per g dry rumen content. Cattle showed significantly higher digestibility values for: organic matter (0·696 v. 0·676, P 0·05), neutral-detergent fibre (NDF; 0·548 v. 0·511, P 0·05) and cellulose (0·621 v. 0 · 509, P 0·01), while the crude protein digestibility (CPD) values were similar (0·667 and 0·671). Comparing the two species for each diet, cattle showed significantly higher digestibility values for organic matter in diet D50·0 only (0·714 v. 0·688, P 0·01), for NDF in diet D12·5 only (0·578 v. 0·531, P 0·05) and for cellulose in all diets (0·660 v. 0·546, 0·630 v. 0·525, 0·605 v. 0·505, 0·588 v. 0·460, P 0·01); in contrast buffaloes showed higher values of the CPD for diet D12·5 (0·662 v. 0·632, P 0·05).