Much of the error in field experiments comes from environmental differences, which traditional methods (e.g. blocks or rows and columns) sometimes fail to control, especially if the fertility pattern is complex or unexpected. In 1937 Papadakis proposed an alternative approach in which the performance of each plot was judged by that of its neighbours. If a plot was situated in an area of generally high cropping, its own high crop was partly discounted but if surrounding plots were yielding poorly, it was given special weight. Adjustment was by the analysis of covariance. The method has not been much used, partly because it still lacks a mathematical basis and partly because it is suspected of bias in the estimation of contrasts between treatments.

It has been studied here by simulation, i.e. bodies of data have been generated on the computer with known characteristics and then analysed by Papadakis's method to see if a correct answer was returned. Also, an improvement in the covariate is suggested and an iterative procedure developed to ensure more consistent results.

It emerges that the suspicion of bias was unjustified. Also, the method can be extremely effective in reducing experimental error. In all its variant forms it has difficulty with outside plots, and for that reason is more effective on compact areas than on strips. Also, interference between plots is especially serious, because the neighbours affected are those needed to form the adjusting covariate. Discontinuities arising from past use of the land can cause problems but a method is suggested for minimizing their effect.

The results of field experiments in 1983/84 and 1984/85 were used to test the hypotheses (i) that cultivars and management systems which result in high biomass of winter barley will also produce high grain yield and (ii) that greater uniformity of tiller and ear size is associated with greater yield. In a set of cultivars, grain yield per plant was significantly correlated with biomass per plant despite no correlation with any individual yield component. Biomass per shoot was significantly correlated with grain yield per shoot. Treatments to increase shoot uniformity (the mass of individual tillers relative to that of the main stem) generally increased biomass per shoot and grain yield per shoot. The re-examination of data in the literature where yield and biomass were available confirmed the association of biomass and grain yield both on a per plant and per shoot basis. It is argued that harvest index is not a character which can be easily targeted for manipulation by growers, whereas crop management systems to increase biomass should be easier to specify.

To date, little molecular genetic research has been conducted in Lolium and the current study is the first investigation of diversity in ribosomal DNA (rDNA) among perennial ryegrass cultivars. Ribosomal DNA probes derived from the intergenic spacer (IGS–M) and transcribed (COD–M) regions of mustard (Sinapis alba) were used to examine 35 perennial ryegrass (L. perenne L.) cultivars for genetic diversity using RFLP methodology. Analysis of digested genomic DNA on Southern blots revealed only 10 cultivars with unique RFLP patterns. Genetic diversity was exclusively associated with the IGS–M probe, as the cultivars were all indistinguishable when examined with the COD–M probe. This low incidence of distinctions was not expected, given that the 35 cultivars were known genetically distinct entities, representing the very wide range of the major phenotypic and physiological characteristics present in commercial stocks. Furthermore, the forage cultivars were proportionately less polymorphic in comparison to the amenity cultivars. This was also unexpected as amenity cultivars are recognized as a more closely clustered group than the forage cultivars, based on morphological characteristics. The potential use of rDNA for assessing cultivar diversity and breeding histories in perennial ryegrass is considered.

Guinea grass (Andropogon gayanus Kunth cv. Bisquamulatus) and Buffel grass (Cenchrus ciliaris L. cv. Biloela) are grown as temporary fodder crops in the Sahelo-Sudanese region of Africa. Differences in their responses to drought conditions may be important in optimizing cattle feeding programmes in integrated farming systems. An experiment conducted in a wind tunnel equipped with a tank forming a weighable lysimeter compared net photosynthesis (A), stomatal conductance (gs) and leaf water potential (Ψ leaf) of both species in a mixed sward during three simulated environmental conditions: (1) rainy season, (2) drought period within the rainy season and (3) dry season.

With a soil water reserve (SWR in % volume) of 8% during the drought period and the dry season, the Ψ leaf and A parameters were significantly lower (P<0·001) when compared to those recorded with a SWR of 32% in the rainy season. Mean gs values were significantly different during each environmental condition (P<0·001). No significant difference (P>0·05) was found in gs and A between the two species during the rainy and the dry season or during a diurnal cycle, although a significant difference (P<0·05) in Ψ leaf existed between the species during the dry season. During this season, Guinea grass was dormant while Buffel grass was growing slowly.

When the drought period was imposed during the rainy season (SWR declined from 25 to 5%), A was reduced from 19·8 to 7·7 μmol m−2 s−1 for Guinea grass and from 16·8 to 3·7 for Buffel grass. Ψ leaf in both species dropped below −2 MPa while gs declined slowly from 0·70–0·80 to 0·55–0·60 cm s−1. A was significantly higher (P<0·05) in Guinea grass during the last days of the drying cycle, but declined rapidly when drought was extended up to the wilting point. Two days after rewatering, all the parameters of both plants were restored to previous levels.

It is concluded that Guinea grass was as drought-resistant as Buffel grass and that no competition for water existed between the species in the mixed sward. The study also gives relationships between the physiological parameters of both species and the soil water content during a drying cycle.

Nitrate leaching was measured over a 3-year period from rotationally grazed perennial ryegrass (Lolium perenne L.) pasture receiving 200 kg fertilizer-N/ha and from similarly grazed ryegrass/white clover (Trifolium repens L.) pasture that received no N fertilizer. The results are discussed together with those from the same plots in the preceding 3 years when they were stocked continuously. Under both managements, the numbers of grazing sheep were adjusted on the basis of the quantity of herbage available on the plots. During the whole 6 years, mean nitrate concentrations in soil water collected by porous cup samplers remained below the European Union limit of 11·3 mg N/l except for the fertilized grass plots in year 5 of the study. Quantities of nitrate leached ranged from 6 to 34 kg/ha per year from the grass/clover plots and 2·46 kg/ha from the fertilized plots. Leaching losses from both types of pasture were positively correlated with the numbers of lamb grazing days in the later part of the grazing season. This relationship and the high spatial variability associated with the measurements indicated that N derived from excreta was the main source of leached nitrate. It was concluded that, where pastures of equal productivity are compared, similar quantities of N are likely to be leached from grass/clover swards as from grass swards receiving N fertilizer.

In order to understand better some of the reasons for differences between forage plant species in digestibility, different parts of nine plant species in either milled or chopped (1 cm lengths) form were examined for in vitro digestibility and in milled form for neutral detergent fibre and lignin. The nine species were: Trifolium repens L., Medicago sativa L., Desmodium intortum (Mill.) Urb., Brassica napus L., Lolium perenne L., Festuca arundinacea Schreb., Chloris gayana Kunth, Cenchrus ciliaris L. and Zea mays L. In each case early harvesting was compared with later harvesting in each of two years. The plants were grown in spring–summer in a heated glasshouse.

The forage at the early harvest was, on average, 1–4% units more digestible in vitro than that at the later harvest and generally slightly lower in lignin and neutral detergent fibre content. However, the stems of Z. mays were higher in neutral detergent fibre at the early than at the later harvest.

The leaf sheaths of L. perenne and F. arundinacea were more digestible than the leaf blades. L. perenne was more digestible than F. arundinacea in both sheaths and blades. The sheaths and blades of C. gayana and C. ciliaris were less digestible and had a higher neutral detergent fibre content than those of L. perenne and F. arundinacea. The leaf blades, excluding the midribs, and the stems and leaf sheaths of Z. mays were all rather high in digestibility when milled and moderately low in neutral detergent fibre and lignin; the leaf blade midribs of Z. mays were less digestible and higher in neutral detergent fibre than the stems and similar to the stems in lignin content. The leaflets of T. repens had an appreciably lower neutral detergent fibre content than the stolons and petioles and a rather lower lignin content in dry matter and yet were, if anything, less digestible than the stolons and petioles. The stolons of T. repens were much more digestible than the stems of M. sativa and D. intortum. The digestibility of D. intortum was low in all the plant parts examined, leaflets, petioles and stems. In both D. intortum and B. napus, the leaflets or leaf blades were much lower than the stems in neutral detergent fibre and lignin and yet they were no more digestible than the stems when milled.

The digestibility of chopped leaflets and leaf blades was similar to that of milled leaflets and leaf blades, but chopping rather than milling reduced the digestibility of stems (particularly of those of Z. mays), petioles, the leaf blade midribs of Z. mays, and, to some extent, leaf sheaths.

To appreciate more clearly some of the physical characteristics of forages which may be important in relation to digestibility and structural integrity, different parts of eight plant species were examined for the proportion of thick-walled, thin-walled and epidermal cells, the thickness of the cell walls and the diameter, length and volume of the cells. The eight species were: Trifolium repens L., Medicago sativa L., Desmodium intortum (Mill.) Urb., Lolium perenne L., Festuca arundinacea Schreb., Chloris gayana Kunth, Cenchrus ciliaris L. and Zea mays L. Early harvesting was compared with later harvesting in each of two years. The plants were grown in a heated glasshouse in spring–summer.

The plant parts with the lowest proportion of thick-walled cells (3–6% of cross-sectional area) were the legume leaflets and those with the highest proportion (47–57%) were the leaf blades and stems of C. ciliaris. The plant parts with the highest proportion of thin-walled cells were the legume leaflets and petioles and the Z. mays stems and leaf sheaths. The walls of the cells categorized as thick-walled were thinnest (0·9 μm) in L. perenne leaf blades and T. repens leaflets and thickest (2·0–2·3 μm) in the leaf blade midribs, leaf sheaths and stems of Z. mays and in the stems and petioles of T. repens. The thinnest outer walls of epidermal cells (0·9 μm) were recorded for the leaf blades of L. perenne.

The largest cells within the categories and plant parts examined (1 100000 μm3) were thin-walled cells in the stems of Z. mays. The longest cells recorded (180 μm) were thin-walled cells in the petioles of T. repens. The thick-walled cells were particularly small (1800–2600 μm3) in L. perenne leaf blades and sheaths and in T. repens leaflets. The largest thick-walled cells in the study were in the stems and petioles of T. repens. The epidermal cells of D. intortum leaflets, petioles and stems were particularly small (2000–3000 μm3).

In order to understand more about the nature of forage particles before they enter the rumen and about the degradation of cell walls in the rumen, particles of different plant parts produced by sheep when eating each of eight plant species were described in terms of volume and exposed surface area and in terms of numbers of cells and volume and surface area of cell wall. Estimates were made of the extent to which the walls would be degraded in the rumen. The eight species were: Trifolium repens L., Medicago sativa L., Desmodium intortum (Mill.) Urb., Lolium perenne L., Festuca arundinacea Schreb., Chloris gayana Kunth, Cenchrus ciliaris L. and Zea mays L. In each case early harvesting was compared with later harvesting in each of two years. The plants were grown in a heated glasshouse.

The plant parts which were chewed to the smallest particle size during eating, and which produced particles with the largest proportion of total cell wall area exposed on the outside of the particle (6–10%), were the legume leaflets and the grass leaf blades and sheaths. The proportion of total cell wall area exposed on the outside of chewed particles derived from stems was only 3–4%. The number of thick-walled cells per mm3 of chewed particle was greatest in particles derived from the leaf blades and stems of C. ciliaris, the leaf blades and leaf sheaths of L. perenne and the stems of C. gayana. The average number of thick-walled cells per particle was 860000 in particles derived from the stems of C. gayana and C. ciliaris and 260000 in particles derived from L. perenne leaf blades and sheaths. The volume and surface area of walls of thin-walled cells per mm3 were greatest in particles derived from legume leaflets. The total surface area of walls of epidermal cells per mm3 was greatest in particles derived from leaflets and leaf blades and least in those derived from stems. The estimated percentage of wall thickness of thick-walled cells which would be degraded in the rumen was relatively high in the case of leaf blades, sheaths and stems of C. ciliaris, leaf blades of Z. mays, leaf sheaths of L. perenne and F. arundinacea and stems of C. gayana, relatively low in the case of petioles of T. repens and stems of D. intortum and M. sativa and close to zero in legume leaflets. The estimated percentage of wall thickness which would be degraded did not seem to be related to the concentration of lignin in the walls, either in grasses or in legumes.

A field trial on alley-cropping was conducted at the University of Ibadan research farm in the 1990/91 cropping season to assess the contributions of arbuscular mycorrhizal (AM) fungi and hedgerow woody legumes to the yield and nutrient uptake of cassava (Manihot esculenta Crantz) as an intercrop in an infertile soil. The trial also investigated the influence of AM fungi on the interplanting of a non-nodulating woody legume Senna siamea (syn. Cassia siamea) with a nodulating woody legume (Leucaena leucocephala).

AM contributions to cassava were greater than the hedgerow contributions, which demonstrated that AM associations are an essential component in the nutrition of cassava. In contrast to cassava, AM inoculation only influenced the leaf dry weight and uptake of nutrients of non-interplanted woody legumes but not the above-ground biomass and P uptake of interplanted woody legumes. However, non-inoculated interplanted Leucaena benefited more from indigenous AM fungi than the competing Senna. The negative contributions to the nutrient uptake (K, Ca and Mg) of cassava by hedgerows and the lack of response to AM inoculation in interplanted hedgerow woody legumes could be attributed to root competition among the different plant species growing in close proximity to each other. The present results show that cassava benefits more from AM association than Leucaena which in turn benefits more than Senna in an alley-cropping system.

Two experiments were carried out in order to study the effect of the ash content of poultry litter (PL) on its nutritional value. In Expt 1, broiler PL was incorporated in the proportions of 30 and 45% into diets based on wheat straw supplemented with 20% maize grain. The PL was either fresh (containing 19% ash) or had been heated in a stack (giving 26% ash). There were also two additional diets containing 45% fresh PL to which either ground limestone or wood ash had been added to provide the same ash content as the diets containing 45% stored, high ash PL. An in vivo digestibility and nitrogen balance trial was carried out. Storing the PL, which resulted in heating and an increase in ash content, reduced dry matter intake significantly. The largest and most significant effect on the digestibility of dry matter, organic matter and crude protein can be ascribed to the freshness of the PL. Stored, heated PL had a highly negative effect on crude protein digestibility and on nitrogen retention. No negative effect of adding extrinsic ash was recorded. In Expt 2, the nutritional value of layer PL (LPL), compared with that of broiler PL (BPL), was examined using ten beef cows before and after calving. The diet was composed of 45% PL, 35% wheat straw (WS) and 20% maize grain. Daily dry matter intakes were 12·7 and 16·2 g/kg liveweight (LW) before calving and 15·5 and 22·8 g/kg LW after calving, for cows fed LPL and BPL, respectively. Owing to the combined effect of the lower dry matter intake, lower gross energy content and lower digestibility of LPL, the digestible energy intake of BPL diets was almost three-fold, and nitrogen absorption (g/kg LW0·75) on the BPL was 2·4-fold higher than on the LPL diets. Because the ash intake of both diets was similar, it is suggested that the main inhibiting factor of intake after calving was the rumen's ability to mobilize the large amounts of ash-rich complexes of ash and organic matter to the reticular omasal orifice and, consequently, out of the rumen.

To determine the genetic variability and population structure of the β-lactoglobulin (LGB) locus in the Argentinian Holstein breed, milk samples from 12 herds at four locations in Córdoba, Argentina, were analysed. Allelic and genotypic frequencies of LGB variants were compared by contingency tables. A G-test was applied to detect Hardy–Weinberg equilibrium. LGB A allele frequency was 0·43 and B allele frequency was 0·57. Allelic and genotypic frequencies indicated great homogeneity among populations. Most of the populations were in Hardy–Weinberg equilibrium. According to the fixation index, populations were panmictics. The partition of the F-statistic demonstrated that mating was at random, although there was a slight excess of heterozygotes, and there were no genetic differences among populations. The Shannon–Weaver index showed that more than 96% of the total genetic diversity for this locus is attributable to the differences within each herd. The diversity among herds within locations and among locations accounted for <4% of the total variation. Phenograms obtained by the Nei procedure for genetic distances disclosed the separation of the herds into two groups, one with an average frequency of LGB B of 0·61 and the other with an LGB B frequency of 0·53. Locations defined two clusters, but the divergence among populations was not important. These results indicated that the Argentinian Holstein breed from Córdoba constitutes one population.

A set of four silages, comprising unwilted and wilted silages, both with and without bacterial inoculation, was prepared from perennial ryegrass swards on each of eight harvesting occasions during the 1994 growing season. The four silages, within a single harvest, were offered as the total diet at the maintenance level of feeding to 16 wether sheep in an 8-period change-over study with experimental periods of 3 weeks duration, in order to determine whole tract digestibility and urinary excretion of purine derivatives (PD). Silage dry matter (DM) and nitrogen (N) degradabilities were also examined in the rumen of four rumen-fistulated steers offered a medium quality silage and a concentrate supplement at a ratio of 60[ratio ]40 (silage[ratio ]concentrate) on a DM basis. The data presented are the mean results obtained across the eight harvests.

There were no significant effects of inoculation on silage pH and ammonia-N/total-N across the unwilted and wilted materials. Inoculation of either the unwilted or wilted silages had no significant effect on microbial activity in the rumen of sheep, as indicated from excretion of urinary PD, or on silage DM or N degradability assessed with rumen-fistulated steers. However, inoculation significantly increased the digestibility of organic matter (1·7%, P<0·01), N (4·6%, P<0·001), energy (2·3%, P<0·01), neutral detergent fibre (1·5%, P<0·05) and acid detergent fibre (2·8%, P<0·001) in the silage. These increases in digestibility following inoculation, in general derived equally from both the unwilted and wilted silages. The results indicate that the increase in silage nutrient digestibility following inoculation probably reflects a more extensive digestion in the abomasum and intestine, rather than in the rumen.

Wilting of grass prior to ensiling resulted in silages with significantly lower pH (5·4%, P<0·05) and ammonia-N/total-N (42·7%, P<0·001) across the untreated and inoculant-treated materials. Wilting also significantly increased (P<0·05) urinary PD-N output of sheep by 6·9% and silage DM degradability by steers by 1·6–2·4% across the untreated and inoculant-treated silages. However, there were no significant effects of wilting on silage N degradability and silage nutrient digestibility.

Four grass silages, namely unwilted and wilted silages each with and without bacterial inoculation, were prepared on each of eight harvesting occasions during the 1994 growing season (Yan et al. 1998). The four silages, made at each harvest, were offered ad libitum to 48 lactating dairy cows in a 2 (control v. additive)×2 (unwilted v. wilted)×8 (harvest (period)) change-over design experiment with 3-week experimental periods. The animals also received 6·9 kg dry matter (DM) of a concentrate supplement daily during the first period and then the allowance of the supplement was reduced by 0·43 kg DM at the start of each subsequent period. The supplement was based on barley, molasses and soyabean meal. The data presented are the mean results obtained across the eight harvests.

No significant interactions were found between inoculation and wilting for silage intake, performance and silage eating behaviour. Inoculation significantly increased milk yield (2·4%, P<0·05) and lactose output in milk (2·3%, P<0·05) across the unwilted and wilted silages. These increases following inoculation in general arose equally from both the unwilted and wilted silages. However, inoculation of either the unwilted or wilted silages had no significant effects on silage DM intake, milk composition, or silage eating behaviour. Wilting significantly increased silage DM intake (20·1%, P<0·001), milk yield (2·8%, P<0·05) and the concentrations (P<0·001) and yields (P<0·01) of both fat and protein across the untreated and inoculant-treated silages. The increases following wilting derived equally from both the untreated and inoculant-treated silages. Wilting also significantly influenced silage eating behaviour with a significant decrease following wilting in number of meals per day (6·5%, P<0·05) and an increase in eating rate (kg DM per eating hour) (29·7%, P<0·001) across the untreated and inoculant-treated silages. These results indicate that inoculation of the unwilted and wilted silages produced similar increases in milk production. Large positive responses in silage DM intake and performance were recorded following wilting with both the untreated and inoculant-treated silages.