A simple model of vernalization, originally developed to quantify the vernalization response of fieldgrown carrots, was fitted to previously published experimental results for winter wheat cv. Norin 27. The optimum temperature for vernalization indicated by the model was c. 5·2 °C, as this induced the fastest progress to final leaf appearance, expressed as the reciprocal of number of days from sowing to final leaf. This rate decreased linearly with temperature rise or fall on either side of the optimum, extrapolating to zero at –4·8 °C (Tmin) and 26·6 °C (Tmax). When all the treatment temperatures and durations were expressed as vernalizing degree days > –4·8 °C (V °C d), there was a linear increase in post-treatment development rate with increasing vernalization up to c. 275 V °C d. Ending the effective treatment duration for vernalization at the estimated time of initiation of the final leaf primordia brought many of the data points closer to the linear trend which described the rest of the data.

Effects of using leaf number, which is linearly related to thermal time, instead of days as the unit of time to compensate for temperature differences in the original experiment were examined. Unvernalized plants had the potential to produce 18 leaves before flowering and therefore rates were expressed as the fraction of the potential total leaf number that each new leaf represented. All plants were assumed to have an initial development rate of 1/18 per leaf. This rate was assumed to increase linearly with time during the vernalizing treatment periods and then remain constant after treatment until the final leaf appeared. Leaf numbers reported from the original experiment were used with these assumptions to estimate the rate at the end of each treatment. The relationship between these rates and treatment temperatures was similar to that for rates based on post-treatment durations. There was an optimum temperature c. 5·5 °C and Tmin and Tmax of –5·1 and 18·8 °C estimated by extrapolating the decreasing linear trends to the base rate of 1/18. When plotted against V °C d calculated from these temperatures, the rates from the full data set were well represented by the model line which had been fitted to the data from just one treatment duration.

Assimilate availability may determine seed yield of caraway during three periods: (i) beforeflowering it may determine dry matter allocation to flowers and thereby the number of ovaries; (ii) during flowering it may determine seed set; and (iii) after flowering it may determine seed filling. From 1990 to 1992 in the vicinity of Wageningen, field experiments were performed in which growth, assimilate storage and partitioning, and the effects of shading during various growth stages on seed set and filling, were studied to determine the importance of these three phases. Dry matter production before flowering was correlated with dry mass of flowers, but there was no relationship between the dry mass of flowers and seed yield. Assimilate availability after flowering only slightly affected seed yield, but during flowering it had a strong effect. Shading 50% (during c. 3 weeks of flowering only) reduced yield by 30% whereas 50% shading (after flowering for > 1 month) reduced yield by only 6%. In addition, the main flowering period of biennial caraway over a number of years was estimated and the daily amount of photosynthetically active radiation (PAR) for that period calculated. There was a close linear relationship between light intensity during flowering and yield. The possibilities for yield improvement using these results are discussed.

The rate of leaf appearance of barley varies substantially with time of sowing. This variation has been related to both the length and the rate of change of photoperiod at the time of plant emergence. An outdoor pot experiment was conducted to test if rate of change of photoperiod directly affects phasic development and rate of leaf emergence of spring barley. Two photoperiod-sensitive cultivars (Bandulla and Galleon) were subjected to five photoperiod regimes: two constant photoperiods, of 14 and 15·5 h, and three different rates of change of photoperiod of c. 2, 9 and 13 min/day from seedling emergence to awn initiation.

Photoperiod treatments significantly affected the duration from seedling emergence to awn initiation in both cultivars. Rate of change of photoperiod did not affect the rate of development towards awn initiation independently of the absolute daylength it produced. Although Bandulla had a longer duration than Galleon at any photoperiod regime, the cultivars did not vary in their sensitivity to photoperiod. When this phase was divided into the leaf initiation (LI) and spikelet initiation (SI) phases, it was evident that the sensitivity to photoperiod was not constant, being in general higher during the SI than during the LI phase. However, the magnitude of the change in sensitivity was cultivar-dependent, indicating that sensitivity to photoperiod during the different phases could be under independent genetic control.

Final numbers of primordia (leaves together with maximum spikelet number) were negatively affected by increasing photoperiods, but once again, there was no evidence of any effect of the rate of change of photoperiod which was independent of the average photoperiod. Both cultivars showed similar sensitivities for final leaf number but maximum spikelet number was more sensitive to photoperiod in Galleon than in Bandulla.

Highly significant linear relationships between leaf number and thermal time were found for all combinations of cultivars and photoperiod regimes (r2 > 0·98). The rate of leaf appearance (RLA) was similar for both cultivars (c. 0·0185 leaves/°Cd) and did not alter during plant development or in response to the change in photoperiod at awn initiation. The range in RLA was greater for Galleon (0·0170–0·0205 leaves/°Cd) than for Bandulla (0·0173–0·0186 leaves/°Cd). Neither of these cultivars exhibited a significant relationship between rate of leaf emergence and photoperiod or rate of change of photoperiod. The lack of significant relationships between RLA and length or rate of change of photoperiod is in contrast with previous reports using time of sowing as a main treatment.

Data on ovulation rate and numbers of ova and transferable embryos recovered from superovulated cattle and sheep were analysed using generalized linear models, quasi-likelihood, restricted maximum likelihood (REML) and generalized linear mixed models (GLMMS). The data pertained to the operation of nucleus breeding schemes in cattle and the commercial application of embryo transfer in sheep.

Results of the analyses showed that generalized linear models involving Poisson and Binomial distributions were inappropriate because of over-dispersion, and that analyses using quasi-likelihood to model negative binomial and β-binomial distributions were more suitable. Factors identified as important in determining the results in cattle were the number of previous superovulations (a higher proportion of transferable embryos were obtained in the initial flush compared to subsequent recoveries in two out of three sets of data), the donor (significant in all analyses with repeated recoveries) and its mate (significant in some analyses). In sheep, the use of pFSH or hMG for superovulation increased embryo yields above those obtained with PMSG + GnRH. Analyses of a further data set for sheep showed the effect of breed was ambiguous.

The effects of donors and their mates were treated as random effects in analyses involving REML and GLMMS. Results showed that the repeatability of the number of transferable embryos produced per donor ranged between 0·13 and 0·23 in three sets of data and was significant in all cases. In these analyses the variance among mates was not significantly different from zero.

The results of analyses were used to develop a random generator to simulate the numbers of ova and embryos recovered from a cow following superovulation. By sampling from negative binomial distributions where the scale factor used for each cow was a normally distributed deviate, distributions were obtained which had the same mean, variance and repeatability as those observed.

The effect of sowing date and plant density are modelled in relation to three variables: Ym, the maximum yield the crop can produce; τ, the interval from sowing to Yb/Ym = 0·5 where Yb is the yield of Brussels sprout buttons and Nb, the number of buttons on a plant stem. Ym is well related to the quantity of solar energy which the crop intercepts during the main period of growth. The fraction of radiation intercepted by the crop canopy is related to leaf area index (L) and sub-models relate L to sowing date and plant density. An empirically derived parameter relates the value of τ to thermal time and photoperiod time. The time course of Nb is modelled relative to the number of buttons at the end of the growing season. To overcome influence of site, variety and season, a generalized equation relates Nb, to plant density and a ‘known’ number of buttons at a specified planting density. The yield of buttons in specific size ranges (required for marketing) is described by a normal distribution with the standard deviation (σ) representing the spread of button diameter. There were no obvious effects of sowing date and plant density on σ. The model enables the effects of sowing date and plant density to be simulated using only simple and easily understood parameters. A sample simulation is presented.

Trials were established at two sites in 1982 to compare six 2-course rotations, in which barley alternated with feed legumes (pure or mixed with barley), fallow or barley, in partial factorial combination with different N and P fertilizer regimes. This paper summarizes data from the second to the seventh cropping seasons.

Barley yielded most after fallow and more after legumes than after barley; but, on the basis of total dry matter production over both courses, barley-legume rotations outyielded barley-fallow and barley-barley rotations. In terms of net nitrogen offtake, which may be taken as a rough measure of feed value, barley-legumes surpassed the other rotations by c. 100% at both sites, whether the comparison was made on values obtained from plots with or without added fertilizer. Of the legumes, Lathyrus sativus was slightly more productive than Vicia saliva, and this carried over into the rotational yield of both dry matter and nitrogen. Rotations involving pure legumes were slightly more productive than those involving legume/barley mixtures.

Yields of all crops and rotations varied greatly from year to year depending on rainfall. Those of the barley-fallow rotation were much less variable than those of the barley-barley rotation at the wetter site (323±103 mm), with the barley-legume rotation intermediate; but variability was much greater at the drier site (250±79 mm), with much less difference between rotations. Nevertheless, fertilizer use and rotations including legumes gave a significant yield advantage under the driest conditions.

The evaluation of n-alkane concentrations in herbage and faeces is the basis of a methodology that yields precise estimates of herbage intake. Established chemical analysis procedures for the determination of n-alkane concentrations in herbage and faecal samples involve an elaborate sequence of steps utilizing n-hexane as solvent for both the liquid-liquid extraction of n-alkanes from the sample and the solid-phase separation of n-alkanes in the extract. Composites of herbage and faecal samples from studies with sheep provided the experimental material used to evaluate a series of modifications which would simplify and reduce the workload involved in the chemical analysis. The results show that a less hazardous chemical, n-heptane, can replace n-hexane at all stages of the analytical procedure. They also show that evaporation of the sample extract and redissolving it prior to the separation of the n-alkanes, using a silica gel column, is unnecessary and that the volume of solvent used can be reduced. Procedures for saponification of samples prior to extraction can also be simplified as the process can be run overnight involving a slow build-up of temperature to 90 °C. The gain in precision from processing samples in duplicate was negligible and it would be more appropriate to invest extra efforts on an increased number of experimental animals. The results snowed that the workload and cost of using the n-alkane technique to estimate herbage intake can be reduced substantially.

Intake and digestion by swamp buffaloes (Bubalus bubalis) and crossbred cattle (Bos indicus × B. taunts) of a range of diets were measured in two experiments conducted in north Queensland in 1988. In Expt 1, four animals of each species were offered rice straw ad libitum with a supplement of minerals and urea. The four dietary treatments were (i) no concentrates, (ii)cracked rice grain (900 g/day), (iii) cracked rice grain (900 g/day) plus sunflower seed meal (900 g/day) and (iv) as for (iii) but with 50% of the sunflower seed meal treated with formaldehyde solution. In Expt 2, the same animals were offered two legumes, lablab (Lablab purpureus) and verano (Stylosanthes hamata cv. verano), and two grasses, sorghum (Sorghum bicolor × S. sudanense) and pangola grass (Digitaria eriantha). In Expt 1, voluntary intake of organic matter of buffaloes was 1·22 that of cattle. Concentrate supplementation increased organic matter digestibility and total intake, but did not affect straw intake. Intake of supplements by cattle was poor in the absence of sunflower seed meal. In Expt 2, intake and digestibility was similar in both species but was higher in animals given sorghum compared with the other forages. Total time spent chewing for both species and all diets ranged between 163 and 244 min/kg intake of plant cell wall constituents. The fractional rate of fluid flow from the reticulorumen and the concentration of propionic acid in the rumen fluid of buffaloes were consistently higher than in cattle in both experiments. Cotton and rice straw, placed in polyester bags in situ in the reticulorumen, were more extensively fermented in buffaloes than in cattle in Expt 1. In contrast, there were no differences between animal species in fermentation rates of cotton and ground diets in situ in Expt 2, but a relationship was observed between the relative fermentation of cotton and the relative digesta retention times in the reticulorumen in the two animal species. Microbial biomass, estimated as microbial dry matter per kg dry matter in the reticulorumen, was less (P < 0·05) in buffaloes than in cattle in Expt 1; higher concentrations of protozoa were observed in buffaloes than in cattle in Expt 2. Digesta load in the reticulorumen of buffaloes was c. 0·88 that of cattle. In Expt 1, digesta load was inversely related to digestible organic matter intake. The results were consistent with the hypothesis that energy metabolism and digesta load in the reticulorumen interact in the regulation of roughage intake, but it appeared that the lower loads measured in both species in Expt 2 indicated the operation of an unidentified limitation to intake.

Herbage intake estimation, using the n-alkane technique, requires the determination of the ratio of ‘dosed’ to ‘natural’ n-alkane concentrations in faeces. The present study was designed to evaluate the accuracy and precision of intake estimates based on the analysis of pooled samples of total daily faecal output and those derived from rectal grab samples.

Wether lambs maintained in individual metabolism crates, offered fresh herbage ad libitumand dosed daily with 120 mg of n-dotriacontane (C32), were used. Actual intake was recorded daily and samples of total faecal output were retained either for individual analysis or to produce 6-day composites by pooling representative amounts of material from each day's faecal output. Different pooling strategies were evaluated: a constant volume of fresh faecal material from each day or a constant weight (or constant volume) of dried and ground material from each day's faecal output. In addition, daily rectal grab samples, obtained at the same time each day, were pooled using a constant daily volume of fresh faeces. Herbage intake was estimated using the ratios of faecal concentrations of n-hentriacontane (C31) or n-tritriacontane (C33) relative to that for C32.

Similar patterns of effects were observed for intake estimates derived from the C31: C32 and C33:C32 n-alkane ratios for the different faecal sample types. Comparisons based on intake estimates derived from the use of the C33: C32 ratio showed that there were no significant differences between actual intake and estimates based either on the mean of daily samples or on the different composites of total daily faecal output. Intake estimates based on the analysis of composites of total daily faecal output were consistently lower (c. 3%) than those based on the analysis of daily faecal samples. A difference of this magnitude is expected, based on the algebraic relationship between the mean of a ratio of two variables (e.g. faecal concentrations of C33 and C32) and the ratio of the corresponding means. Intake estimates derived from the use of rectal grab samples, taken 3 h after the daily dose was given, yielded a proportionate overestimation of actual intake of 0·06 (P < 0·10). No significant differences were detected in the precision of intake estimates derived either from individual daily samples, composites of total daily faecal output, or composites of rectal grab samples, although the latter yielded the lowest precision.

It is concluded that faecal samples can be pooled on a gravimetric or a volumetric basis over several days with negligible loss in precision of intake estimates. Such pooling greatly reduces the workload and costs involved in using the n-alkane technique. Sampling and analytical errors associated with t e technique were found to be very small, and the processing of faecal samples in duplicate would result in a negligible increase in the precision of intake estimates. The need for further evaluation of rectal grab sampling under grazing conditions is highlighted.

Tiller weight, lamina length, lamina growth rate, appearance interval and lamina utilization were examined in Lolium perenne, Agrostis stolonifera, Holcus lanatus and Poa trivialis. Marked tillers were measured in situ over 14–21 day periods in a continuously grazed permanent pasture under steady state management in Devon, UK, in 1985, 1986 and 1987 on plots receiving either zero (0N) or 400 kg nitrogen (400N) fertilizer/ha per annum.

L. perenne was incapable of reducing its individual tiller weight or lamina length to the same extent as in the other three species during the grazing season. Lamina appearance interval was longer in L. perenne than in the other three species in most of the observation periods in both the 400N and 0N plots. Overall the lamina appearance intervals were similar between A. stolonifera, H. lanatus and P. trivialis. Lamina extension rates were greater in L. perenne than in either A. stolonifera or P. trivialis in most observation periods in both plots. H. lanatus had lower lamina extension rates than L. perenne in c. 50% of the observation periods in both plots, at other times the extension rates of the two species were similar, with one exception in early spring in the 400N plot when H. lanatus had a higher extension rate. There was no clear pattern in seasonality as to when H. lanatus had a lower extension rate than L. perenne. H. lanatusachieved either similar or higher lamina extension rates than either A. stoloniferaor P. trivialis in c. 50% of the observation periods, respectively; the periods when higher rates were observed in H. lanatus in the 400N plot occurred mainly in the spring and early summer. A. stolonifera and P. trivialis achieved similar lamina extension rates in most observation periods in both the ON and 400N plots. The specific differences observed in the lamina extension rates were emphasised when converted to growth rates by multiplying the length increments by the average weight per unit length of the expanded lamina for each species.

Lamina utilization, in terms of percentage of lamina length removed by grazing, was similar between the four grass species in most observation periods in the ON plot. A. stolonifera lost more lamina length than either L. perenne or P. trivialis in May and June and H. lanatus lost a greater percentage of lamina length than either L. perenne or P. trivialis in May in the ON plot. In the 400N plot L. perenne lost a greater percentage of lamina length than any of the other species in June and more than P. trivialis in May. A. stolonifera lost a greater percentage of lamina length than any of the other species in July, and more than either L. perenne or P. trivialis in May. This greater severity of grazing in A. stolonifera was associated with a high incidence of tillers being grazed to stubble. H. lanatus lost more lamina length than L. perenne in May. When the loss of lamina length through grazing was converted to losses in terms of weight of tissue removed, then L. perenne consistently lost more than either A. stolonifera or P. trivialis in both the 400N and 0N plots.

The implications of these differences in tissue production and utilization on competitive interactions between L. perenne and the other three species are discussed.

Thirty-five populations of Medicago polymorpha were collected from throughout Sardinia (Italy) in 1989 with a view to developing pasture species suitable for improving degraded grasslands in the northern Mediterranean basin. Herbage and seed production were compared with the Australian cultivar, Circle Valley, over 2 years at Bonassai, north Sardinia. Regeneration in the 2 years after sowing and seed size were also estimated. All variables were related to collection site parameters using multivariate analysis.

Herbage production varied between 2 and 8 t dry matter/ha, and up to 1·5 t/ha of seed was produced. K-means clustering of agronomic and morphological variables indicated that there were two groups present; one similar to variety polymorpha and one to variety vulgaris. There were two other single entry clusters, one of which contained cv. Circle Valley. Principal component analysis of the environmental variables indicated that cluster 1 (vulgaris) was more likely to come from mountainous areas where winter temperatures are low, and cluster 2 (polymorpha) from coastal areas where temperatures are mild. Regeneration of cluster 1 was better than that of cluster 2, which in turn was better than Circle Valley, indicating that populations in cluster 1 are better adapted to the management system imposed at Bonassai.

The results indicate that M. polymorpha has considerable potential to improve the grasslands of Sardinia. However, it is unlikely that imported cultivars will be successful, and it seems important that the selection of local populations should continue. Commercial seed production in Sardinia is likely to be a problem, and grazing management under the conditions of communal ownership may have to be reviewed. It is important that future research and development involves farmers and other industry groups.

The idea that the addition of compound feed to grass at ensilage can result in adequate preservation of the crop without the need for an acid additive, and that less compound feed would then need to be given when the silage is subsequently fed, was tested with dairy cows. Two diets were compared: (1) a conventional grass silage, to which acid had been added at the time of ensiling, plus concentrates supplied at the time of feeding, and (2) grass silage, to which a compound feed had been added at the time of ensiling, plus a lower rate of concentrates. Each diet was fed to 12 lactating cows for 82 days.

The total intake of dry matter was 17·3 kg per cow per day on diet 1 and 18·3 kg on diet 2. Daily milk yields per cow were 24·5 and 26·1 kg, respectively, with similar milk composition. The estimated diet costs per kg of milk were 9·3 p on (1) and 8·3 p on (2). It is concluded that the addition of compound feed to grass at ensilage is worth considering on dairy farms as an alternative to conventional grass silage, but that a more comprehensive study is needed.

The buoyancy and distribution within the rumen of feed particles were measured to assess the extent to which particle buoyancy modulated the movement of particles within and from the reticulorumen. Two experiments conducted in north Queensland in 1988 involved swamp buffaloes (Bubalus bubalis)and crossbred cattle (Bos indicus × B. taurus) fed rice straw with various supplements of energy and protein, lablab (Lablab purpureus), verano (Stylosanthes hamata cv. verano) and two grasses, sorghum (Sorghum bicolor × S. sudanense) and pangola (Digitaria eriantha). Apparent passage rate of particles from the reticulorumen bore an inverse curvilinear relationship to particle size. Passage rate of the smallest particles, retained by a cloth of aperture 50 μm, was faster in buffaloes. Mean particle size of total rumen digesta was smaller in buffaloes than in cattle in Expt 1 with rice straw, but there was no difference between the forage diets offered in Expt 2. The large particle content in digesta (dry matter per kg) in the dorsal rumen was greater than that in the ventral rumen, which in turn was greater than that in the reticulum. The sedimentation rate of particles was measured by a procedure designed to interfere as little as possible with microbial fermentation. A curvilinear relationship was evident between sedimentation rate and particle size; the medium particles sedimented faster than the fine and large particles. Sedimentation rates of some particle fractions were greater in buffaloes than in cattle. Between diets there were marked differences in sedimentation rates of large particles, but these differences decreased with smaller particle size. Large particles from the dorsal sac sedimented faster than those from the ventral sac, and there was no obvious dorsal raft. There was no apparent relationship between particle buoyancy and ease of particle escape from the reticulorumen. The results are discussed in terms of current theories of the physical factors involved in the control of voluntary feed consumption by ruminants.

The growth of sugarbeet (Beta vulgaris L.) seedlings was investigated under laboratory conditions in 1990 and 1991. Three phases were identified during heterotrophic growth in the dark, based on the changes in dry weight, carbon and nitrogen contents. The perispermic reserves were first transferred to the young seedling. All the components of the seedling (cotyledons, hypocotyl, radicle) began to elongate and their dry weights increased. The seedling separated from the seed at the end of this first period, 4 days after imbibition had started at 20 °C. During the second period, there was a major redistribution of material between organs, mainly from cotyledons to the hypocotyl. The different organs elongated rapidly. From the start of the third period, 6 days after sowing at 20 °C, the organs began to lose weight and gradually ceased to elongate. The dry weights of the whole seedlings and of each organ were proportional to the initial seed weights. The elongations and changes in dry weights obtained at different temperatures coincided when time was expressed as thermal time using a base temperature of 3·5 °C. The growth of the seedling after illumination depended on the heterotrophic growth duration. If emergence occurred after 100 °Cd, the relative growth rate of the seedling decreased. In field conditions, plant variability at the end of sugarbeet establishment is due not only to variations in seed weights but also to environmental conditions that control emergence delay.

Low, medium and high planes of nutrition were imposed on Friesian calves from 1 to 13 weeks of age (Period 1) and factorially arranged with low and high planes of nutrition in the 13–25 week period (Period 2). In the final (residual) period, from 25 weeks to slaughter, all animals were given the same high plane diet comprising 3·5 kg of concentrates and grass silage ad libitum. The different planes of nutrition were achieved by offering restricted allowances of milk replacer or concentrates with grass silage ad libitum. Representative animals were slaughtered at 1, 13 and 25 weeks of age while the remaining animals were slaughtered at mean liveweights of 503 and 553 kg. Over 2 years, a total of 120 Friesian steers were taken through to the final slaughter point, with 60 additional animals being slaughtered at interim stages. The animals were housed in non-bedded accommodation throughout the study and the calves were born in January of each year. The ME intake of the low plane animals in Period 1 was 0·55 that of the high plane animals and produced liveweight (LW), empty body (EBW), carcass (CW), non-carcass (NCW) and alimentary tract (ALW) weight gains which were 0·60, 0·46, 0·47, 0·44 and 0·55 of the gains of the high plane animals. In Period 2 the low plane animals had ME intakes which were 0·69 of the high plane animals and produced proportional gains of 0·57, 0·48, 0·45, 0·51 and 0·56 respectively. Previously restricted animals had higher rates of gain subsequently and the indices for weight compensation by the final slaughter point in LW, EBW, CW, NCW and ALW for the low plane treatments in Period 1 were 0·72, 0·82, 0·89, 0·73 and 0·97 while indices for low v. high in Period 2 were 0·82, 0·82, 0·79, 0·86 and 0·95. With lower plane animals, food intakes/liveweight0·75 were significantly higher during the compensating period. Calves which had been restricted in either of the early periods had significantly improved food conversion efficiency during the compensating period but lifetime food efficiency was not affected by plane of nutrition in early life.

Pure chicory (Cichorium intybus) and perennial ryegrass (Lolium perenne) forages were cut and fed fresh at Palmerston North, New Zealand, during March 1993 to castrated male red deer kept indoors in metabolism crates. Chicory contained lower levels of dry matter, higher levels of ash, and had a higher ratio of readily fermentable: structural carbohydrate than perennial ryegrass. Apparent digestibility of organic matter was highest for chicory (0·81 v. 0·72), but cellulose apparent digestibility was highest for perennial ryegrass (0·71 v. 0·59). Relative to perennial ryegrass, the rumen fluid of deer fed chicory contained higher concentrations of protozoa, ammonia and total volatile fatty acids (VFA) but had a lower pH at 15.00 h. Chicory-fed deer had higher rumen VFA molar proportions of n-butyrate and a higher acetate: propionate ratio. Total eating time and chews during feeding/g dry matter intake were similar for deer fed the two forages, but deer fed chicory spent much less time ruminating (33 v. 270 min/day) and had fewer rumination boluses (38 v. 305/day). It was concluded that the low rumination time may indicate rapid disintegration of chicory in the rumen to < 1 mm critical particle size, and that particle breakdown and rumen fractional outflow rate should be measured in future experiments with deer fed on chicory.

Low, medium and high planes of nutrition were imposed on Friesian calves from 1 to 13 weeks of age (Period 1) and factorially arranged with low and high planes of nutrition in the 13–25 week period (Period 2). In the final (residual) period, from 25 weeks to slaughter, all animals were given the same high plane diet. The different planes of nutrition were achieved by offering restricted allowances of milk replacer or concentrates with roughage ad libitum. The roughage was grass hay during the milkfeeding period and grass silage thereafter. Representative animals were slaughtered at 1, 13 and 25 weeks of age while the remaining animals were slaughtered at mean liveweights of 503 and 553 kg. The ME intake of the low plane animals in Period 1 was 0·55 that of the high plane animals and produced daily accretions of protein, fat and energy in empty body which were respectively 0·43, 0·29 and 0·36 of the gains of the high plane animals. The ME intake of the low plane in Period 2 was 0·69 of the high plane and daily accretions of protein, fat and energy in empty body were 0·46, 0·59 and 0·54 of the gains of the high plane animals.

Within Period 2 compensatory protein growth was observed in both carcass and non-carcass components in response to nutritional restriction in Period 1. Compensatory protein growth was also obtained in the carcass component in the residual period as an indirect effect of nutritional restriction in Period 2.

Nutritional restriction in early life produced relatively small non-significant direct effects on body and carcass composition and while early restriction subsequent produced a considerable degree of compensatory growth, there were no significant differences in carcass measurements or chemical composition of the carcass by the final slaughter weight.

The relationships between body weight, scrotal characteristics and environmental effects (year, birth type and age) and fertility were studied in a total of 254 Horro rams. Fertility varied between years (P < 0·01), but not with body weight at mating within the range (30–55 kg) observed, nor with type of birth (both P > 0·05).

Testicular diameter and circumference were strongly correlated to body weight (r = 0·95 and 0·89 respectively; both P < 0·001), but weight of testes only weakly so (r = 0–28; NS). Scrotal circumference was more highly correlated to body weight in ram lambs (r = 0–85; P < 0–001) than in adults (r = 0·51; p = 0·06).

It is concluded that adequate levels of fertility (82 ± 1·3%) are expressed by 30 kg Horro rams, and that when feed supplies are limited there are no benefits to be gained from maintaining rams of this breed at higher weights.

The nutritional potential of a number of raw tropical seeds was assessed in a series of feeding trials with rats. Seed lectin reactivity was also monitored, α-amylase and trypsin inhibitory activities were determined in some of the seeds.

Abelmosclius esculentus, Chenopodium quinoa, Delonix regia, Macroptilium lathyroides, Papaver sonmiferum, Parkia biglandulosa, Sesbania arabica, Terminalia catappa, Vigna subterranea, Vigna umbellata and Vigna unguiculata seeds supported moderate rat growth. The seeds contained only low levels of essentially non-toxic lectin, moderate amounts of trypsin inhibitors and negligible quantities of a-amylase inhibitors and they have great potential as dietary protein sources for man and animals.

Artocarpus altilis, Canavalia ensiformis, Canavalia maritima, Dioclea grandiflora, Phaseolus acutifolius, Phaseolus coccineus and Phaseolus vulgaris cv. Processor, cv. Rosinha G2 and cv. Carioca 80 seeds were toxic. These seeds contained high levels of potentially toxic lectins. Other antinutritional factors may also have contributed to the high oral toxicity of some of these seeds.

Albizia adinocephala, Albizia lebbeck, Bauhinia violacea, Cassia nodosa, Cassia tora, Dioclea sclerocarpa, Entada phaseoloides, Enterolobium cyclocarpum, Leucaena leucocephala and Moringa oleifera seeds were also highly toxic but had only low levels of essentially non-toxic lectins suggesting that the toxicity was due to other anti-nutritional factors.

Bauhinia reticulata, Macrotyloma uniflorum and Tamarindus indica proteins were poorly digested and utilized. The seeds contained low levels of lectins which agglutinated only rat and cattle erythrocytes which had been pre-treated with suitable proteases. Brownea macrophylla had a similar lectin reactivity.

Whole-body oxygen (O2) consumption was measured continuously in pigs from the product of blood flow through the pulmonary artery and arterio-venous (AV) difference of blood O2 content across the lungs (pulmonary AV procedure). Blood flow in the pulmonary artery, a measure of cardiac output, was monitored continuously using a transit-time ultrasound flow probe surgically implanted around the artery. The O2 content of pulmonary artery blood, which represents mixed venous blood O2 aturation (SvO2) was recorded continuously by fibre-optic oximetry. The relative constancy of arterial blood oxygen saturation (SaO2), obtained by catheterization of the saphenous artery, obviated the need for continuous monitoring. Complete recovery from surgical procedures, as indicated by the absence of infection and a return to pre-surgery food intake, occurred in 3–5 days. The combination of transit-time ultrasound and fibre-optic oximetry allowed the pulmonary AV procedure to provide continuous data on the oxygen consumption of growing pigs at high temperatures.