The growth and development of maize and soya bean when grown in mixtures in different proportions and arrangements were studied at Samaru (11° 11' N, 38° E) and Yandev (07° 23' N, 09° 10' E) in the savannah belts of Nigeria. The crops were sown in both pure and mixed stands. The pure-stand densities were 513 plants/m2 for maize and 20·15 plants/m2 for soya bean. Mixtures were made in three proportions: additive (100:100), containing 100% of the sole crop population of each crop; semi-additive (67:67), containing 67% of each crop's sole population; and replacement (67:33 and 33:67), containing 67% of one crop's population to the other's 33%. The mixtures were formed either on the same ridge (in-ridge) or on different ridges (alternate ridge).

Maize grain yield at Samuru was not significantly reduced in mixtures except where its proportion was reduced to 33% of its sole crop density. Seed yields in soya bean were significantly reduced in all mixtures, and when associated with maize density of not greater than 67%, soya bean was capable of 50% of its sole crop yield. At Yandev, late planting coupled with inadequate rainfall in the month following planting depressed seed yields in both crops, the sole crop yields being only 1·53 t/ha in maize and 0·24 t/ha in soya bean. Also maize appeared to be adversely affected in the mixtures at this location, the yields being significantly reduced in the mixtures irrespective of planting ratios.

All mixtures, except 100:100 at Yandev, proved beneficial by producing land equivalent ratio (LER) values greater than one, while 67:67 proportions made higher gross returns than any of the sole crops at Samaru. In this experiment moderately high populations of each crop were found to be necessary to obtain beneficial yields from the mixtures. In this regard semi-additive mixtures, with 67% of each crop's pure stand population, appeared most appropriate.

Incidence of maize streak was significantly lower in alternate ridge mixtures; otherwise the crop in the mixtures seemed to be little affected by pattern of plant arrangement; while soya bean was quite sensitive, its flowering and podding being significantly delayed in in-ridge mixtures at Yandev. The legume yields in this mixture were also slightly, though not significantly, adversely affected. Thus alternate ridge arrangement was more favourable.

Growth curves fitted to factorial data can be modelled using an extrinsic time variate or using the mean responses within the levels of a subset of factors. Where factors can be partitioned into a set of ‘treatment’ factors and a set of ‘environment’ factors, fitting growth curves to the mean effects of environments allows the effects of treatments to be assessed relative to a uniform background growth rate. This leads to a test of a null hypothesis of equal treatment effects in all environments, given that the mean growth rate and stage of development in all environments is equal. The approach is exemplified using data from a glasshouse tomato crop experiment testing variety, nutrient and sowing date factors. Variety and nutrient treatment effects were of direct interest but sowing dates were intended to generalize results by providing a range of growing environments. Treatment effects are analysed by modelling running cumulative yield totals by growth curves and regressing variety and nutrient growth variates on the mean growth variate within each sowing date. In the discussion the case of more than one environmental factor is considered.

The post-weaning growth, frame size and body condition of Hereford (H × H), Brahman × Hereford (B × H), Simmental × Hereford (S × H) and Friesian × Hereford (F × H) heifers were evaluated on three pasture systems in each of five years. One group of heifers was grown from weaning (at 8 months of age) to 18 months on the high-pasture system, which included supplementation with grain, while the other two groups were grown to 30 months on pastures of medium or low quality. Pasture effects were large compared with differences between years, with average daily gains between weaning and 18 months of 552, 275 and 97 g/day on high, medium and low pastures, respectively. Genotype had a significant effect on all measurements. Pasture × genotype interaction was significant in most instances. While crosses grew faster and were heavier than H × H on all pastures, S × H were heaviest by 18 months on high pasture, with B × H heaviest at all ages up to 30 months on both medium and low pastures. There was little difference in live weight between B × H and F × H on high pasture, or between S × H and F × H on medium pasture, whereas F × H was heavier than S × H on low pasture. There were concomitant differences in measures of skeletal size, but when adjusted allometrically to a common live weight, pasture effects were removed, while differences between genotypes remained. At a common live weight F × H had the largest height, length and pelvic area, but shared the lowest body condition with S × H. Pelvic height was greatest among B × H heifers, but this cross had similar body length, condition and subcutaneous fat cover to H × H. The pattern of genotype × pasture interaction observed in this study was similar to that observed by Darnell, Hearnshaw & Barlow (1987) for growth rate among contemporary steers located in different environments.

The interaction between Phytophthora infestans and potato viruses X (PVX) and Y (PVY) was studied in the potato cultivar Kufri Chandramukhi. Prior infection by PVX and PVY reduced the susceptibility of the plants to P. infestans. The reduction in susceptibility was more marked when leaflets were inoculated with P. infestans 72 h after inoculation with PVY. The same type of reaction was observed in whole plants when the interval between PVY and P. infestans inoculation was 30 days. No significant differences were observed when the sequence of inoculation was reversed.

The Southdown is a terminal sire breed producing about 205 registered rams for commercial use in Britain each year. It is strongly associated with its native area of Sussex, with an hierarchical breed structure very similar to that of 30 years ago. The number of breeding ewes has declined from 5475 in 1957 to (minimum estimate) 1202 in 1985.

French and New Zealand imports in the last 20 years have resulted in present-day British Southdown breeding rams being, on average, only 76% of British descent. If further importations are to be made, semen from rams without foreign blood should be taken for storage first, in the interests of genetic conservation.

The required production of Southdown rams for commercial purposes is achieved with a rather small population size. It would be misleading to describe the Southdown as a ‘rare breed’ without taking account of its commercial function and breed structure.

The potential of selected cultivar mixtures to decrease lodging and consequently improve yields in winter barley was examined in field experiments at two locations over 2 years. The experiments were designed to identify growing conditions and varietal characteristics associated with mixture effects, i.e. the deviations from arithmetic pure cultivar means, and to compare the latter with effects of growth regulators. Pathogens were controlled chemically.

Use of mixtures tended to reduce initial lodging. With increasing lodging, desired mixture effects prevailed in mixtures of cultivars highly susceptible with those highly resistant to lodging. On average, the yield advantage of mixtures compared with pure stands was small; it was greater when lodging decreased yield of a pure cultivar. Reductions in lodging prevailed in two-component mixtures; adding more components was of no advantage.

When lodging developed steadily, effects of mixtures on grain yields were comparable to those of growth regulators. When lodging was caused suddenly by a thunderstorm, only growth regulator applications resulted in yield improvements. Growth regulators resulted in yield advantages in two of the four trials only.

Experiments were conducted on the seed rates of the components of oat-vetch mixtures combined with varying rates of nitrogen fertilizer in a rainfed area of north-east Portugal in 1982–3 and 1984–5.

Forage oats grown alone showed a very high yield response to N and rainfall ranging from 30 to 18·2 t D.M./ha, but had a very low crude protein concentration (37–54 g/kg D.M.) and a poor mineral nutrient composition.

The increased use of vetch in the seed mixture had a buffering effect on dry-matter (D.M.) yield but this depended very much on crop N nutrition. For N-deficient conditions vetch inclusion increased the D.M. yield and had a strong positive effect on total forage crude protein (CP) and mineral composition. When crop N nutrition was good, very high D.M. yields (> 12 t D.M./ha) were possible using a high proportion of oats, with vetch making a reduced contribution to the yield and nutritive value of total forage.

Seed rates suggested in the literature for Mediterranean conditions are adequate for low N inputs and low available soil N, but not for high D.M. yields which can only be obtained by using a higher proportion of oats and adequate N fertilizer.

In two trials the effect on D.M. intake of applying flavouring agents to grass silage at feeding time was examined using a total of 89 British Friesian cows in mid to late lactation. The flavouring agent used in Expt 1 was Simax 100 (5 ml/cow per day), while in Expt 2 Palatol (20 ml/cow per day), molasses (0·7 kg D.M./cow per day) and molasses + Talin (0·7 kg D.M./cow per day) were added to both lowand high-quality grass silage.

Both Simax (significantly) and Palatol increased silage D.M. intake by an average 0·67 kg/day. The total D.M. intake of the cows fed the molasses and molasses + Talin was similar to that obtained by those receiving the silage treated with Simax and Palatol. However, in the case of the cows receiving the molasses and molasses + Talin, the increased D.M. intake was obtained mainly from the additional molasses and not by an increased silage intake. This result indicates a very low substitution rate for molasses.

Although D.M. intake was increased by the addition of flavouring agents, milk production remained unaltered because of the relatively late stage of lactation of the cows in both trials.

This paper presents a series of 20 two-replicate row-and-column designs for field trials comparing between 25 and 100 varieties on small plots.

The in sacco, in vitro and in vivo effects of feeding proteins partially coated with calcium soaps of longchain fatty acids were examined.

In sacco, 84–90% of whey powder and soya-bean meal coated with calcium salts of fatty acids remained after 20 h incubation in the rumen of sheep. In vitro tests revealed no effects on volatile fatty acid or ammonia production.

In vivo sheep balance studies, where soya-bean meal coated with calcium soaps was substituted for soya-bean meal, showed no effects on ammonia or volatile fatty acid production in the rumen due to the calcium soap coated proteins. No changes were observed in digestibilities of dry matter, nitrogen or acid detergent fibre; total fatty acid digestion increased. Nitrogen balance was improved slightly in sheep fed the protected protein.

It appears that proteins coated with calcium soaps are not degraded in the rumen and thus energy and non-degradable protein can be supplied to ruminants by this route.

Five experiments in harvest years 1980–2 tested the effects of straw inoculum and fungicides on leaf blotch (Rhynchosporium secalis), growth and yield of winter barley. Inoculating plots with rhynchosporium-infected straw increased the rate of disease development and greatly decreased seedling growth but inoculating plots with wheat straw or sterilized barley straw had no effect. In 1982, plots inoculated with infected straw gave less grain than uninoculated plots, but in all years fungicide sprays applied in winter or early spring had mostly small effects on grain yield that were only poorly related to their effects on leaf blotch and seedling growth.

Experiments of balanced design in harvest years 1981 and 1982 were used to measure interactions between plots of winter barley with different amounts of leaf blotch, caused by the splash-dispersed pathogen Rhynchosporium secalis. On the appropriate transform scales (logarithms of counts and logits of percentages), the effects of extreme treatments on neighbouring plots were up to 30% of the effects of the same treatments on the plots to which they were applied. Powdery mildew (Erysiphe graminis f.sp. hordei) was commonly least severe in plots with most leaf blotch except soon after fungicide sprays had been applied which, although chosen to decrease leaf blotch, also had short-lived effects on mildew. Consequently, contrasts in mildew between differently treated plots changed sign during the season. The effects of the same treatments on neighbouring plots similarly changed with time but not necessarily in phase with their direct effects. Analyses of the rhynchosporium data that recognized the effects of neighbouring treatments typically had much smaller residual mean squares than analyses that ignored neighbour effects but assumed randomized block designs.

Treatments had mostly small effects on grain yield but these data from two of the experiments showed marked positional variation. Individual plots yields from one of these experiments, testing five treatments, are quoted in the appendix so that they are available to others with an interest in alternative methods, such as nearest-neighbour models, to adjust for local correlations between plots.

The effects of different severities of continuous grazing imposed during the spring, followed by a uniform continuous grazing management from mid-summer onwards, were examined in relation to changes in tiller demography and leaf growth of perennial ryegrass (Lolium perenne). Two grazing severities were imposed, based upon maintaining a constant sward height of either 35 mm (severe) or 75 mm (lenient). From 1 June onwards, a uniform grazing pressure with a sward height of 60 mm was imposed.

Differences in the age class structure of the tiller populations developed during the spring between the two grazing treatments, principally through the suppression of daughter tiller development under the more lenient grazing. The demographic differences between the treatments were further increased during the remainder of the grazing season, under the common grazing management, largely because of the greater production of secondary daughter tillers in the sward which had been leniently grazed. The demographic differences between the swards had little effect on net pasture production during the treatment season, with lamina growth rates being virtually unaffected across a wide range of steadystate, continuous-grazing pressures. However, longer-term effects on pasture development, due to the demographic differences between the swards, did appear in the following season; the more severe grazing pressure treatment resulted in earlier inflorescence development in the subsequent season.

Gnotobiotic pigs were fed control and soya-bean protein containing diets in an attempt to identify direct effects of this protein on intestinal structure and function.

Feeding diets containing soya-bean protein for 4 days to 21-day-old gnotobiotic pigs increased crypt depth in the lower half of the small intestine. This increase was accompanied by a corresponding reduction in lactase activity, brought about mainly by a decrease in the apparent rate at which this enzyme appeared in the luminal membrane of developing enterocytes. Sucrase and maltase II and III activities increased slightly in tissue taken from pigs fed soya-bean protein. Alanine transport measured in the presence and absence of Na remained unaffected by change of diet.

The structure and function of control gnotobiotic pig intestine differed from that found in normal unweaned piglets. The possible source of these differences and the probable role played by enteric microflora in amplifying initial effects of soya-bean protein on intestinal function is discussed.

Field experiments during the wet seasons of 1981–1985 determined the response of soya bean to the application of P, K and Mo at four locations in the Nigerian savanna. Soya bean was grown at factorial combinations of four rates of P (0, 13·2, 26·4 and 396 kg P/ha) and two rates of K. (0 or 41·5 kg K/ha) in 1981, but in subsequent years two or three rates of Mo (0, 0·5 or 1·0 kg ammonium molybdate/ha) were included in a randomized complete block design with three replications at all locations.

Significant yield increases due to the application of P were recorded in 10 out of 11 trials. Yield increases due to 13·2 or 26·4 kg P/ha were from 29 to 210% over the control. The effects of K and Mo application were not significant in 10 out of 11 trials and eight out of nine trials, respectively. Whereas P x Mo interaction effects were significant in three trials only, highest yields were recorded due to application of 26·4 or 39·6 kg P/ha in combination with 0·5 kg ammonium molybdate/ha.

Application of P increased the percentage of P in soya-bean leaves significantly and a range of 0·275–0·330% P appears to be the critical nutrient range of P in soya-bean leaves at full bloom stage. Application of K had no effect on K concentration in leaves.

Variations in floral morphology have been observed by most breeders involved in cowpea (Vigna unguiculata (L.) Walpers) improvement. Usually, the most successful pod set is achieved following hand pollination when plants with large flowers are used as female parents because these are easy to manipulate. In breeding work involving wide crosses, unconscious selection of female parents on the basis of flower size could lead to correlated changes in pod and seed characteristics which may or may not be desirable. This would depend on the heritability of flower size and its correlation with the pod and seed characteristics.