A three factor experiment with winter wheat (Triticum aestivum L.) was carried out during two years on a coarse sandy soil in Denmark. The factors comprised four irrigation strategies including no irrigation, three nitrogen levels providing 67, 83 or 100% of the recommended nitrogen rate, and two strategies for control of leaf diseases (with and without fungicides). Different varieties were used in the two years, Pepital in the first year and Hussar in the second year. Mildew dominated the trial in the first year, but was almost absent in the second year. Septoria occurred in both years, but most severely in the second year. Irrigation increased grain yield, but there were no significant differences between the three strategies, where irrigation was applied. The effect of irrigation on yield was almost solely via an effect of increased transpiration, whereas water use efficiency and harvest index was unaffected. There was a significant interaction for grain yield between irrigation and nitrogen strategies with higher irrigation effects at higher nitrogen rates. This interaction was absent at high disease levels. Increasing nitrogen rate increased grain yield in the second year, but not in the first year. This was attributed to an increasing mildew incidence with increasing nitrogen rate. Irrigation also increased mildew incidence. This caused significant interactions for grain yield between fungicide application and nitrogen rate and between fungicide application and irrigation strategy. Septoria was also significantly affected by both nitrogen and irrigation strategies, but to a lesser degree and not in a consistent manner. The main effect of disease on grain yield was through a reduction in harvest index and a reduction in grain weight. Increasing nitrogen rates slightly reduced harvest index both on a dry matter and on a nitrogen basis. The interaction effects were smaller than the effects of the main factors, and the effect of changes in irrigation or nitrogen strategy on disease incidence could not substitute for the effective disease control obtained by fungicides. The results do, however, indicate possibilities of improving the efficiency of current decision support systems for both fungicide control by better timing of application and irrigation scheduling by reducing water application.

This is part of the text of an invited lecture given by R. K. Scott on 1 December 1998, the first in a series to commemorate the memory of R. Hull, the first Director of Broom's Barn. Dr Hull devoted his career to the science underpinning the homegrown sugar industry.

Soya beans, like other legumes, contain low concentrations of the nutritionally essential sulphur amino acid, methionine. Cysteine, although not an essential amino acid because it can be synthesized from methionine, also influences the nutritional quality of soya bean products when it is only present in low levels. A low cysteine content will also aggravate a methionine deficiency. Soya bean lines deficient in 7S protein subunits have been identified. The 7S proteins contain substantially less methionine and cysteine than the 11S proteins. With the myriad of genetic null alleles for these subunits it may be possible to tailor the 7S/11S storage protein ratio and their total composition in seeds to include only those subunits with the richest sulphur amino acid composition. Cotyledon feeding experiments, using isolated soya bean cotyledons, demonstrated that addition of methionine to the culture media caused increased synthesis of both proteins and free amino acids but the mechanism by which this takes place is not clear. Biotechnological approaches to improve nutritional value of soya beans include elevated expression of genes that originate from other species which encode high-sulphur proteins. High level expression of a 2S Brazil nut albumin gene in soya bean resulted in raised methionine concentration although the Brazil nut gene is highly antigenic and therefore will not be useful in production agriculture. Modification of glycinin to increase sulphur amino acid content is possible, and these gene products are capable of normal assembly into trimers in vitro although are rapidly degraded in vivo by the asparaginyl endopeptidase responsible for post-translational modification of proglycinin. Solutions to the methionine deficiency may be anticipated from a combination of approaches followed in laboratories worldwide. Many of these approaches are not without difficulty but, despite this, the likelihood is that soya beans with improved nutritional quality (which may not be confined to sulphur-containing amino acids as other nutritionally essential amino acids are also valuable) will be available in the near future. It will be essential to confirm that the increased total methionine (or other amino acid) is digestible to the animal to at least the same degree as conventional cultivars.

The influence of herbicides at reduced rates and repeated stubble-cultivation on weeds and crop yields was estimated in five field trials with spring-sown cereals situated in the south of Sweden during the autumn of 1989 until the spring of 1997. Stubble-cultivation was accomplished during 1989–1996, while herbicides were applied at 0, 1/8, 1/4 or 1/2 of full dose during 1990–1996.

In the spring of 1997, i.e. after 7 years without herbicide application, seedling densities 3 weeks after weed emergence were 68–340/m2 at three sites and 535–610/m2 at two sites when averaged over tillage treatments.

Averaged over herbicide doses, stubble-cultivation reduced the plant density of annual broad- leaved weeds by 6–32% at three sites and increased the density by 25% at one site. At the remaining site, the density was not significantly influenced. Stubble-cultivation reduced the populations of two perennial and seven annual weed species, while one species was stimulated and nine species showed null, or inconsistent, responses. In the spring of 1997, i.e. one year after the last herbicide application, the densities of weed seedlings in 1/8, 1/4 and 1/2-doses were 34, 46 and 56% lower, respectively, than in the untreated controls.

Stubble-cultivation increased crop yields at four sites by 200 kg/ha as a mean over herbicide doses. At these four sites, averaged over 1993–1995, herbicides increased yields in plots that were not stubble-cultivated by 7, 8 and 10% in the 1/8, 1/4 and 1/2 of a full dose, respectively, relative to the untreated control. In 1996, herbicides increased yields at only two sites.

It is concluded that a fruitful way for weed management with a low input of agrochemicals is to combine the use of herbicides at reduced rates with repeated stubble-cultivation.

Data from three ryegrass spaced plant variety trials conducted at Crossnacreevy, Co. Down in each of 1989, 1990 and 1991 were investigated for evidence of spatial dependence. Spatial dependence was most evident in variates measuring the overall dimensions of the plants, especially late season variates. However, the ‘date of ear emergence’ variate, which identifies most of the variety distinctnesses that the trials are intended to detect, was not spatially dependent. On average, 45% of the residual variation in the plot means of the spatially dependent variates was estimated to be of a spatial nature. Consequently, it is likely that the efficiency of analysis of these variates would be improved by using spatial analysis instead of non-spatial analysis or by using alpha (incomplete block) designs instead of complete block designs. The spatial information was used to determine the optimal incomplete block sizes and the likely increase in efficiency due to using alpha designs instead of complete block designs. The optimal incomplete block size for a trial with 10 plants per plot was compared with the optimal incomplete block size for the same trial with single plant plots at the same spacing and was found to be smaller. The gains in efficiency were larger in the trial with 10 plants per plot than in the trial with single plant plots.

Nutritional value of most plant materials is limited by the presence of numerous naturally occurring compounds which interfere with nutrient digestion and absorption. Although processing is employed widely in removal of these factors, selection of cultivars of soya beans with inherently low levels would have a considerable impact on efficiency of non-ruminant livestock production. The review considers the role of plant breeding in achieving this objective. The most abundant trypsin inhibitors are the Kunitz and the Bowman–Birk inhibitors, containing 181 and 71 amino acids respectively. The Kunitz inhibitor is present at a concentration of 1·4 g/kg of total seed contents and the Bowman–Birk inhibitor 1·6 g/kg. A large number of isoforms of the Bowman–Birk inhibitor have been described in soya bean cultivars and it has been shown that the general properties of the inhibitor are, in fact, attributable to different isoforms. Nulls for both Bowman–Birk and Kunitz trypsin inhibitors have been identified, allowing new low trypsin inhibitor cultivars to be produced. However, research into breeding for low trypsin inhibitor cultivars currently has limited application as trypsin inhibitors contribute a major proportion of the methionine content of soya beans. Trypsin inhibitors are thought to be involved in the regulation of and protection against unwanted proteolysis in plant tissues and also act as a defence mechanism against attack from diseases, insects and animals. Hence, in breeding programmes for low trypsin inhibitor cultivars, alternative protection for growing plants must be considered. Use of soya beans in non-ruminant animal feeds is limited by the flatulence associated with their consumption. The principal causes appear to be the low molecular weight oligosaccharides containing α-galactosidic and β-fructosidic linkages; raffinose and stachyose. Non-ruminants do not have the α-galactosidase enzyme necessary for hydrolysing the α-galactosidic linkages of raffinose and stachyose to yield readily absorbable sugars. Soya beans contain between 6·8 and 17·5 g of phytic acid/kg; a ring form of phosphorus (P) which chelates with proteins and minerals to form phytates not readily digested within the gut of non-ruminants. One approach for over-coming the effects of phytic acid is through synthesis of phytase in the seeds of transgenic plants. Currently, recombinant phytase produced in soya beans is not able to withstand the processing temperatures necessary to inactivate proteinaceous anti-nutritional factors present. Soya bean lectins have the ability to bind with certain carbohydrate molecules (N-acetyl-D-galactosamine and galactose) without altering the covalent structure. Lectins are present in raw soya bean at a concentration of between 10 and 20 g/kg. Purified soya bean agglutinin is easily inactivated by hydrothermal treatment but in complex diets binding with haptenic carbohydrates may confer protection against denaturation. The majority of research into soya bean lectins is carried out using laboratory animals so very limited information is available on their in vivo effects in farm animals. This review is concerned specifically with breeding but there are other means of improving nutritive value, for example processing which may alter protein structure and therefore functionality of proteinaceous anti-nutritional factors present.

Data from a three factor experiment carried out during two years were used to analyse the effects of drought, nitrogen and disease on light interception (IPAR) and radiation use efficiency (RUE) in winter wheat (Triticum aestivum L.). The factors in the experiment comprised four irrigation strategies including no irrigation, three nitrogen levels providing 67, 83 and 100% of the recommended nitrogen rate, and two strategies for control of leaf diseases (with and without fungicides). Light interception was estimated from weekly measurements of crop spectral reflectance. This method was compared with estimates derived from crop area index measured by plant samples or by using the LAI2000 instrument. There was a good correspondence between the different methods before anthesis, but an overestimation of light interception with the methods using crop area index after anthesis due to an increase in non-photosynthetic active leaf area. Irrigation increased both IPAR and RUE. The relative increase in IPAR for irrigation was greater than the relative increase in RUE in the first year, whereas they were of similar size in the second year. The differences between the years could be attributed to changes in timing of the drought relative to crop ontogenesis. Increasing nitrogen rate increased IPAR, but caused a small decrease in RUE in both years. This reduction in RUE with increasing nitrogen concentration in leaves was also found to be significant when disease levels and drought effects were included in a multiple linear regression. Fungicide application increased IPAR in both years, but RUE was only significantly reduced by disease in the first year, where mildew dominated the trial. The data were also used to estimate the coefficients of partitioning of dry matter to grains before and after anthesis. About 40% of dry matter produced before anthesis and about 60% after anthesis was estimated to contribute to grain yield. The low fraction after anthesis is probably due to the fact that it was not possible to estimate changes in RUE with time, which may lead to biases in the estimation of partitioning coefficients.

The prediction of silage feeding value from the analysis of herbage at ensiling and the effects of rate of nitrogen (N) fertilizer, date of harvest and treatment additive at ensiling on fermentation, dry matter (DM) recovery and effluent production were evaluated in a factorial experiment. Herbage was harvested from 75 plots, laid out in three replicate blocks of 25, from the primary growth of a predominantly perennial ryegrass sward. The plots received 72, 96, 120, 144 or 168 kg N/ha and were harvested on 10, 17, 24 or 31 May or 7 June. Herbage was ensiled for 176 days in laboratory silos (6 kg capacity), either untreated or treated with formic acid or a bacterial inoculant, each applied at 3 ml/kg herbage. Increasing the rate of N fertilizer and delaying harvest date increased herbage DM yield but decreased silage digestible organic matter digestibility (D-value) and intake potential (P < 0·001).

Increasing the rate of N fertilizer increased crude protein (CP), acid insoluble N (AIN) (P < 0·001), acid detergent fibre (ADF) (P < 0·01) and acid detergent lignin (ADL) concentrations but had no statistically significant effect (P > 0·05) on pH or the concentrations of ammonia N, lactic or volatile fatty acids. Delaying harvesting date decreased CP, AIN and ADF concentrations (P < 0·001) and effluent DM losses (P < 0·001). Treatment with either formic acid or inoculant decreased silage pH and the concentrations of ammonia N, acetic acid and propanol and increased AIN, ethanol, neutral detergent fibre (NDF), ADF and cellulose concentrations, effluent volume, DM losses and intake potential. Formic acid treatment decreased D-value (P < 0·001), whereas inoculant treatment was not statistically significant (P > 0·05). For potential silage and metabolizable energy (ME) intakes and D- value, strong positive correlations (P < 0·001) were identified with herbage pH, buffering capacity and concentrations of CP, AIN, true protein nitrogen (TP), while negative correlations (P < 0·001) were identified with herbage yield and concentrations of NDF and hemicellulose. With untreated silages, strong positive correlations (P < 0·001) were identified between silage pH and herbage pH and concentrations of AIN, buffering capacity, ash, TP and negative correlations (P < 0·001) were identified with herbage DM yield and concentrations of DM and WSC. Herbage yield and concentrations of NDF, TP, DM and nitrate N provided the best linear relationship for predicting potential ME intake of the resultant untreated silages (R2 = 0·94). It is concluded that delaying harvest date had the most detrimental effect on silage feeding value, although this effect was partially overcome by reducing the rate of fertilizer N and use of either an inoculant or formic acid additive. Silage feed value was also highly correlated with the protein and fibre fractions, ash, buffering capacity and pH of the herbage at ensiling.

Two separate experiments, in clay loam and loamy sand (prepared by mixing the clay loam with washed sand), were performed to determine the effects of sowing method and sodicity on the survival, ion uptake, grain yield and yield components of spring wheat (Triticum aestivum L.) cv. Kharchia- 65. Three sodicity levels (control, exchangeable sodium percentage (ESP) 5–7; low (ESP 18–20); high (ESP 39–40)) and four sowing methods (sowing dry and pre-germinated seed and transplanting of 16 and 21-day-old seedlings) were tested. In the control and at low sodicity, sowing method had no effect on plant survival, grain and straw dry weight per plant. However at high sodicity, these variables were lower in plants established from pre-germinated seed than in plants established from dry seed, the farmers' normal practice. In contrast, transplanted seedlings showed increased survival and had significantly higher grain and straw dry weight than plants established by sowing dry seed. Differences in grain yield between sowing method and sodicity treatments were mainly due to differences in the number of grains per plant. Although increasing sodicity was associated with higher concentration of Na+, and lower concentrations of K+, Ca2+, Mg2+ and lower K+/Na+ ratio in flag leaf sap, ion concentrations were unaffected by sowing method.

It is suggested that the increased survival and yield of transplanted seedlings is due to the fact that they are not exposed to sodicity during the sensitive stages of germination and emergence. In addition, their already established roots and shoots may be better able to capture the resources required to support their subsequent growth. The decreases in grain and straw dry weight per plant, and the increases in these parameters achieved by transplanting seedlings instead of sowing dry seed, were greater in clay loam than in loamy sand. Further studies are required to determine whether the responses to transplanting observed here also occur in sodic soils under field conditions, and to investigate the technical and economic feasibility of adopting this technique in commercial agriculture.

An investigation was made during 1988 to test the hypothesis that ‘Grasslands Huia’ white clover (Trifolium repens L.) could be eliminated under close sheep grazing. The effects of grazing management, topography and fertilizer on the contribution of Huia plants to the white clover population in an 85 ha experimental upland pasture ecosystem in the southern North Island, New Zealand (lat. 40° 20′ S, long. 175° 50′ E, 125–350 m altitude) were quantified 11 years after oversowing. Replicated sampling sites (108 in total) were located on nine combinations of slope and aspect within grazing management treatments comprising rotational grazing with cattle (RC), rotational grazing with sheep (RS) and continuous grazing with sheep (CS), with high and low fertilizer treatments in each case. White clover occurrence, leaf area, phosphoglucoisomerase-2 (PGI-2) allele frequencies and the proportion of Grasslands Huia plants in the white clover population were determined at each site. White clover frequency was lower on steeper slopes. Aspect, slope and grazing management affected area of individual clover leaves. The proportion of Huia plants in the white clover population averaged 54·9, 49·0 and 33·6% for RC, RS and CS, respectively (P < 0·039, 5 D.F.). Fertilizer and topography did not affect the proportion of Huia. It was concluded that although Huia did persist after 11 years of close sheep grazing, its contribution to the total white clover population was unsatisfactory in some cases, and use of better adapted cultivars is suggested.

When crops of wheat (Triticum aestivum L.) are stressed, grain set in potentially fertile florets is reduced. Cold temperatures and boron (B) deficiency during reproductive development cause grain set failure in wheat. Patterns of grain set in cold-stressed and B-deficient wheat ears were studied under field conditions in Nepal and in controlled environments in the UK. In both B-deficient and cold-stressed circumstances, ear fertility was reduced by up to 98% but the pattern of grain set within an ear was similar. Under cold-stressed conditions, florets in the uppermost one-third of the ear were 41 to 53% less fertile than those located in the middle and basal regions. Even in the unstressed crops, the top one-third of the ear was less fertile than below by as much as 8–13%. Similarly, within a spikelet, the distal florets always had fewer grains than the proximal ones. Decreased grain set following stress markedly reduced yield per ear. We conclude that fertility should be assessed on the entire ear. The determination of competent florets should be based on the presence of well-developed ovaries, feathery stigmas and the structures of anthers (which can still be seen in the sterile florets at maturity) rather than on the length of the lemma or on judgements based on visual appearance or other subjective criteria.

Two organic fertilizers were prepared from the same initial mixture of poplar sawdust, blood and flour either by composting in a reactor or by chemical oxidation. Both processes resulted in loss of c. 30% of the organic matter. Composting required 90 days in comparison to only a few hours with chemical oxidation. Extraction of the organic residues with 1 N KOH gave solutions containing 24·6 and 15·1 g/l of humic substances respectively. These humic solutions were applied to pot-grown Lolium multiflorum Lam. at 4 and 10 mg carbon per pot to assess the short-term uptake of macro and microelements by the plants. When the plants were short of phosphorus, the humic substances from the chemically decomposed sawdust supplied at 10 mg C per pot improved total P uptake and yield. Humic substances also increased copper and manganese uptake, and by enhancing root development, also improved nitrogen uptake and biomass yield.

A field study was conducted to investigate the agronomic performance, nitrogen (N) efficiencies, yield components and yield of rice in Spain for different nitrogen treatments. The experimental variants were six preflood N rates (0, 50, 75, 100, 125 and 150 kg/ha) and three topdressing patterns: (i) non- topdressed, (ii) topdressed with 50 kg N/ha at mid-tillering stage (MT) and (iii) topdressed at panicle initiation stage (PI). The N status of the plant was measured at different growing stages to determine whether a chlorophyll meter would be useful in making N sidedress recommendations. The results showed that grain yield increased with increasing amounts of preflooding N fertilizers up to 100 kg N/ha. The main effect was on panicle number per unit ground area. The effect of additional N supply on yield components and grain yield depended on application timing. Split applications of N did not improve the agronomic efficiency but reduced days to maturity and lodging and increased the harvest index value. Split applications increased grain yield when the total N rate was 150 kg/ha with the second supply at PI. There was a significant but not very good relationship between N content and chlorophyll content (SPAD) values. The regression equation differed significantly depending on growth stage. The SPAD value may determine the need for N topdressing at MT stage, but not at PI. However, the relationship between SPAD value and the rice yield increase from N topdressing application was not very good.

The growth and yield of pinto beans (Phaseolus vulgaris L.) cv. Othello in response to a total of six sowing dates (from October to December) and irrigation was examined over two seasons in Canterbury, New Zealand. In 1994/95, two irrigation treatments (nil and full) were combined with two sowing dates (27 October and 24 November). In 1995/96, Othello was examined under two irrigation treatments (nil and full) and four sowing dates (1 November, 15 November, 29 November and 13 December). The total rainfall for the two seasons was 50% and 60% of the long-term average, respectively. The mean temperatures for the seasons were similar to the long-term average. Both irrigation and sowing date had a marked effect on growth and seed yield. Averaged over both seasons, seed yield for fully irrigated crops was 337 g/m2, c. 50% higher than the yield of unirrigated crops. The irrigated crops yielded more than the unirrigated crops because they attained greater canopy closure, intercepting 84–95% of incident radiation. They also had on the average 47% higher leaf area duration (LAD), 72% higher maximum leaf area index (LAI) and greater utilization coefficient. The mid- to late November-sown crops yielded more than the late October to early November and December-sown crops because the leaf area of the former increased most rapidly, achieved a higher maximum LAI and LAD and consequently intercepted more photosynthetically active radiation (PAR). They also had faster pod growth rates and 26% of stored assimilates contributed to pod growth compared with 13% in late October to early November and 5% in December-sown crops. The results showed that pinto beans can grow and yield well in Canterbury, and that a yield advantage could be obtained when sown in mid- to late November and with irrigation.

Laboratory incubation in the UK of an Oxisol from Burundi and an Ultisol from Cameroon with 3·1% by weight of prunings of young shoots of Calliandra calothyrsus, Cassia siamea, Flemingia congesta, Grevillea robusta, Gliricidia sepium, Leucaena diversifolia and Leucaena leucocephala resulted in increased soil pH and decreased exchangeable aluminium content. The greatest increase in pH and corresponding decrease in exchangeable aluminium occurred during the first 14 days of incubation and the decrease continued at a slower rate until 42 days incubation. The acid neutralizing effect decreased after 42 days but was still important at the last sampling time at 98 days. Polyphenol to nitrogen ratio was not well correlated with observed change in soil pH whereas the total base cation (calcium, magnesium and potassium) content proved to be a good predictor of these changes in the Ultisol, but not in the Oxisol. The proposed mechanism giving rise to acid neutralization is complexation of protons and aluminium by organic anions. The total base cation content of the prunings ranged from 0·94 to 2·25 molc/kg and the buffer capacity of the Oxisol was 48 mmol OH−/pH/kg.

Use of fertilizers and manures during 1994 were studied at the farm and catchment scale in the largely agricultural Ythan catchment, north-east Scotland, using farm level census data supplemented by questionnaire data. Grassland accounted for 40% of the agricultural land, and seven farm types represented 87% of the total land, having an average size of 90 ha. The average livestock density of 1·2 livestock units/ha was high compared to Scotland as a whole (0·5). Rates of inorganic fertilizer applied to individual crops in the area corresponded with the national average and current advisory recommendations. At the catchment scale, most fertilizer N was applied to grassland (47%), whereas spring crops received the greatest proportion of the fertilizer P (35%). The annual manure production equated to an average over the catchment of 63 and 16 kg/ha of N and P, respectively. When calculated for farm types these figures ranged from 27 and 6 kg/ha on ‘cereal’ farms to 384 and 163 kg/ha on ‘pig’ farms. The ratio of applied fertilizer N and P varied from 4[ratio ]1 for ‘general cropping’ to 10[ratio ]1 for ‘cattle and sheep (lowground)’ farms. There was no significant compensatory reduction in inorganic fertilizer applications on crops, which also had received manures.

Seed loss from sesame capsules before harvest is a major economic problem for mechanized production. Capsule dehiscence and seed loss were investigated by examining capsule anatomy of 32 sesame varieties and relating anatomical features to a measure of seed retention. This survey of sesame capsule anatomy illustrated that differences between varieties were mostly quantitative, except for the presence of cell layers over the median vascular bundle. This feature, associated with the indehiscent (idid) genotype, prevented capsule-splitting and increased seed retention. None of the dehiscent sesame varieties had capsule anatomical features that were correlated with seed retention. During capsule senescence, mesocarp cells shrank more than endocarp cells creating tension in the drying capsule wall. The tension forced capsules open along a zone of weakness between locules.

Twenty slurries, 20 farmyard manures (FYM) and 10 poultry manures were chemically analysed to characterize their nitrogen (N) fractions and to assess their potential organic N supply. The organic N fraction varied between manure types and represented from 14% to 99% of the total N content. The readily mineralizable N fraction, measured by refluxing with KCl, was largest in the pig FYMs and broiler litters, but on average only represented 7–8% of the total N content. A pot experiment was undertaken to measure N mineralization from the organic N fraction of 17 of these manures. The ammonium-N content of the manures was removed and the remaining organic N mixed with a low mineral N status sandy soil, which was sown with perennial ryegrass (Lolium perenne L.). N offtake was used as a measure of mineralization throughout the 199 day experiment. The greatest N mineralization was measured from a layer manure and a pig slurry, where N offtake represented 56% and 37% of the organic N added, respectively. Lowest (%) N mineralization was measured from a dairy cow slurry (< 2%) and a beef FYM (6%). The mineralization rate was negatively related to the C[ratio ]organic N ratio of the ammonium-N stripped manures (P < 0·01, r = −0·63).

A 5-year study was conducted from 1988 to 1992 at three sites in Niger to determine the effects of crop rotation of a cereal and legumes and nitrogen fertilizer on chemical properties of the soil (0–20 cm) and yield of pearl millet (Pennisetum glaucum (L.) R.BR.), cowpea (Vigna unguiculata (L.) Walp.), and groundnut (Arachis hypogea L.). Four N levels and rotation treatments including continuous fallow were investigated. Soil samples taken from the top 20 cm depth at the end of the experiment from treatments without nitrogen application which included continuous fallow, fallow–millet rotation, groundnut–millet rotation, cowpea–millet rotation, and continuous millet were analysed for soil pH, organic carbon, total nitrogen and exchangeable bases. Fertilizer N significantly increased yield of pearl millet, cowpea and groundnut. Continuous monocropping of pearl millet resulted in lower yields across N levels compared to legume–millet rotations. Legume yields were also consistently lower in monoculture than when rotated with millet. There was a decline in organic matter under continuous millet, cowpea–millet rotation and groundnut–millet rotation. The fallow–millet rotation supplied more mineral N than the legume–millet rotations. Nitrogen availability was greater in cowpea–millet rotation than continuous millet. Crop rotation was more productive than the continuous monoculture but did not differ in maintaining soil organic matter. The legume–millet rotation at 30 kg/ha N appears to be the most viable for millet production. Research should focus on understanding the effect of legume/cereal intercrops and rotations on soil productivity.

The objective of the study was to determine, in the absence of competition (0·74 plants/m2), the existing variation within lines of maize (Zea mays L.) for plant and ear traits (yield per plant, number of tillers, number of ears, plant height, number of nodes, ear length, and kernel row number), as well as the relationship between hybrid potential yield per plant (i.e. yield per plant in the absence of competition) with line traits. Two categories of material were tested in honeycomb trials, grown from 1989 to 1995 in the University farm of Thessaloniki, Greece. The first category were selections from long-term inbreds B73 and Mo17, derived through honeycomb design in the absence of competition, and their crosses. Inbred selections and their crosses were evaluated in different years. For the second category two sets of S5 lines, A and B, derived through controlled selfing from the commercial single- cross hybrid PR 3183, were used in random crosses between S5 plants of lines A and B and resulted in 40 triplets; each triplet included each of the selfed parent, and their hybrid. Triplets of A and B S6 lines and A × B S5 × S5 crosses were evaluated in the same experiment. Significant differentiation among lines and hybrids was found for almost all traits; variation within lines for potential yield per plant, even in case of B73 and Mo17, seemed to be transmissible to the hybrids. Although correlation values were small in the case of B73 and Mo17 selections, there was a positive correlation between hybrids and lines for their potential yield per plant. Selection for high yield within S lines gave ‘prolific’ plants, with line prolificacy being correlated with hybrid potential yield per plant.