The Lolium-Festuca complex of related species provides a wide range of variation for the development of versatile grasses adapted to meet the changing requirements of livestock farming. Various approaches have been used in order to harness the full potential of this variation. Tetraploid hybrids between the two ryegrass species L. perenne and L. multiflorum have resulted in stable and successful commercial varieties. Hybrids between more distantly related ryegrass and fescue species have demonstrated useful combinations of traits from both parents. Chromosome doubling is necessary in these hybrids to restore fertility in synthetic amphiploids. However, genetic instability can still be a problem and research is in progress to obtain tighter control of chromosome pairing. An alternative approach is to introduce a few specific traits from fescues into ryegrasses, or vice versa, using introgressive breeding procedures. This review outlines the progress already achieved in exploiting the ryegrass–fescue complex by using these various approaches and assesses the potential of hybrid derivatives.

Three experiments carried out at Bridget's Experimental Husbandry Farm, Hampshire between 1984 and 1987 on clay loam and silty clay loam soils over chalk investigated the response of winter wheat grown after grass to spring-applied fertilizer nitrogen. The influence of method of establishment and the use of a nitrification inhibitor were also studied. Yields were similar whether the crops were established by direct drilling or following ploughing. Crops required between 88 and 209 kg N/ha nitrogen fertilizer for optimum yield. There was no consistent difference in nitrogen requirement nor were there consistent differences in nitrogen offtake (which is the amount of N in kg/ha removed in harvested grain) or apparent fertilizer recovery following ploughing or direct drilling. The use of a nitrification inhibitor increased grain yield when applied in the autumn without fertilizer, but not when applied with fertilizer in the spring.

Spring barley and spring oats were strip-seeded into crops of the white clover cultivar Alice at Hereford in 1987 and cultivar S184 at Aberystwyth in 1988. Drilling was done with or without a band-spray of glyphosate, a moderate or low (1988 only) dose of paraquat or into plots where the clover had been killed by herbicide 2 months previously.

In the first experiment, initial cereal emergence was sparse; growth was suppressed in the unchecked (unsprayed) clover base but was vigorous in the clover-free plots; the cereals also became dominant in the swards sprayed with herbicide. Whole-crop yields in mid-August were 13, 3–4 and 8–10 t DM/ha in the plots in which clover was killed, unchecked or checked with herbicide. Clover contributed 4–12% of the harvested herbage in the latter treatment. This treatment also yielded c. 70 % as much N, carbohydrate and fibre as that without clover. Grain yields exceeded 7 t/ha without clover but were only 0·3 t/ha for barley and 1·1 t/ha for oats with unchecked clover; in the checked clover plots, barley yielded 60% and oats 78% as much as on the clover-free plots. Four and 5 weeks after whole-crop harvest, residual clover growth was 27 and 39% of that on unchecked plots for oats and barley, respectively.

In the second experiment, the cereals emerged thickly but were later dominated by the clover, and an equitable balance was achieved only with the larger dose of paraquat. However, the oat cultivar Emrys was suppressed less than the tall barley cultivar Dandy; the short barley cultivar Digger was the most suppressed. Mean whole-crop yields were 11 t/ha in the treatment without clover, about half this in the unchecked bases and c. 9 t/ha with the larger dose of paraquat. Differences in chemical composition reflected much larger clover contents in 1988 than in 1987. Yields of N and water-soluble carbohydrate were at least as large or larger with moderate paraquat than for the clover-free plots. Grain yields ranged from 5·6 to 6·9 t/ha for the three cereal cultivars without clover but were negligible to very small in the unchecked and band-sprayed treatments, and were 3·4 and 2·0 t/ha for oats and barley, respectively, with the larger dose of paraquat. Residual stolon weights, 70–80 days after whole-crop harvest, greatly exceeded initial values in all treatments. They were smallest following the larger dose of paraquat, and larger in Digger than in Dandy, which in turn was larger than in Emrys.

It is concluded that the concept of growing cereals in a clover base shows potential as a low input–moderate output system of cereal production. However, further longer term work is needed on regulation of the cereal–clover balance, on the release and uptake of N and the environmental effects of the technique.

Apex and ear development in nine spring barley genotypes were studied at two sites in southern Spain in 1988. The genotypes were grouped into medium and early types, depending on the time from sowing to anthesis. Medium genotypes had significantly more spikelet primordia than early ones at both sites. Spikelet mortality was similar between genotypes and sites, resulting in a higher final number of grains/ear in the medium than the early genotypes.

There was a positive correlation between the maximum number of spikelet primordia and the duration of three stages of apical development: lemma-, stamen- and awn-primordia. Negative correlations were observed between spikelet mortality and the duration of certain developmental phases of the ear.

The growth and seed production potential of white clover (Trifolium repens L.) was studied in field plots in 2 years, with and without a companion grass (Lolium perenne L.). Three rates of grass suppressant, fluazifop-butyl (Fusilade), applied in early June, were compared on plots sown with the companion grass. Two clover varieties, Aberystwyth SI84 and Olwen, were compared.

Sowing a companion grass reduced the amount of clover in the sward in both the first and second harvest years, but did not affect the vigour of individual clover growing points or the number of clover inflorescences when the crop was ready to be harvested for seed.

Fluazifop-butyl, at the highest rate used (125g/ha), reduced the rates of leaf emergence and extension, and the numbers of reproductive tillers, in Lolium, without reducing the rate of leaf emergence in clover, although the weight per clover leaf was somewhat reduced. In the first harvest year, the numbers of SI84 inflorescences in June and July were increased by fluazifop-butyl, but the effect had worn off by harvest (24 August). Seed yield potential was not increased by fluazifop-butyl at the rates used.

Olwen produced heavier leaves, but fewer leaves and fewer growing points/0·1 m2 than SI84. In the first harvest year, Olwen had fewer and smaller inflorescences than S184.

It proved feasible to measure the weight of new leaf produced per week and the nett additions to stolon and inflorescence dry matter (DM). The estimate of nett addition to stolon DM was increased by increasing the number of visible internodes included in the calculation, up to the maximum number recorded (nine). The rate of leaf emergence was slower on young branches than on main axes.

Field experiments were conducted at Fagu and Shimla in India in 1983 and 1984 to study the effect of thiourea (0 and 5 kg/ha) on tuber yield of potatoes and uptake of P with different rates of fertilizer P (0, 22, 33, 44 and 66 kg/ha). Application of phosphorus and thiourea, both alone and in combination, increased tuber yield and uptake of P significantly at both the locations in both years.

Experiments were conducted in controlled environment cabinets and in the field at Cambridge, UK, in 1987–90 and in Valencia, Spain 1989–90 with several varieties of potato. The initiation of leaf primordia and flowers was recorded in sprouts stored for long periods at 13°C with constant illumination and in sprouts exposed to different photoperiods and periods of cold storage. In field experiments, the effects of a wide range of planting dates, differences in physiological age, the time at which seed tubers were sprouted and of imposing an 8 h photoperiod were investigated.

More than 40 leaf primordia were initiated in the sprouts of Home Guard and Arran Comet stored at 13 °C, but, in other varieties, fewer leaves were initiated. Flowers were initiated in the sprouts of some varieties. In field-grown plants, the number of nodes to the first flower differed between varieties and was increased by ageing sprouts before planting. Stems from physiologically old seed of all varieties had more nodes below ground than young seed but the number of aerial leaves to the first flower increased only slightly with increasing physiological age. The number of aerial leaves to the first flower was lower and less variable in Estima than in Maris Piper.

Initiation of flowers usually occurred before tuber initiation and within two weeks of emergence but was earlier in some varieties than others. There was little influence of photoperiod on the time to initiation of flowers and tubers in most experiments but, with low levels of radiation, flowering was completely inhibited in the variety Maris Piper.

The effect of bee visits to the flowers on flower longevity and capsule, seed and oil production in linseed (Linum usitatissimum L.) cv. Antares, was investigated in two very different growing seasons (1988 and 1989); plots caged to exclude insects, plots caged with a honeybee colony and plots uncaged and open-pollinated by honeybees and bumblebees were compared. The experiments were conducted at Rothamsted Experimental Station Farm in Hertfordshire.

Although honeybee visits to the flowers resulted in earlier petal fall, they did not increase capsule or seed production. It is concluded that the crop is self-fertile, self-pollinates well, and that inadequate pollination is not a factor limiting yield.

Dormancy in seeds of Adansonia digitata (L.) can be attributed partly to the testa and partly to the pulp. Several methods, such as wet heat treatment, total or partial seed decoating and scarification of seeds with concentrated acids, herbicides, fungicides and growth regulators, were tested in 1988 and 1989 to break dormancy. The most effective was scarification with HNO3 or H2SO4, which gave 86 and 98% germination, respectively, while total decoating, partial decoating, wet heat treatment and ethrel gave 85, 80, 75 and 70%, respectively. Seeds treated with herbicides and fungicides did not germinate. Inhibition of germination caused by the pulp was overcome by thorough washing of seeds and soaking in water for 3 days.

In pot and field experiments at Xichang, China, in 1984–85, (NH4)2MoO4 applied alone or with KH2PO4 to Vicia faba var. major increased total N, soluble sugar and chlorophyll contents and photosynthetic rate. The growth of root nodules and N fixation were stimulated, and more assimilate was distributed to seeds. Treatment with (NH4)2MoO4 combined with KH2PO4 gave the highest total N, soluble sugar and chlorophyll contents, the most root nodules, fruiting branches, fertile pods and seeds/plant and the highest 100-seed weight, and consequently produced the highest seed yield.

Ammonia losses from surface-applied cattle slurry were measured under field conditions using a wind tunnel system that allows variables affecting ammonia loss to be examined under controlled conditions. The experiments were carried out on a sandy soil with seven different surface covers. This report considers the effect of wind speed, temperature and water vapour deficit on the ammonia loss over a series of 6-day periods. During October 1986 to November 1989 42 treatments were examined, using slurries taken from the same slurry tank to provide slurries of similar chemical composition.

When temperatures were near zero, the rate of ammonia loss was generally low. The accumulated loss over 6 days was high, however, because the rate of loss was constant throughout the period. In these experiments the soil was saturated with water and partly frozen, and the infiltration of slurry into the soil reduced. At 19 °C initial loss rates were high but, after 12 h, almost no further loss occurred. Apart from these extremes, the ammonia loss rates within the initial 24 h were significantly affected by temperature and wind speed.

Ammonia volatilization after 6 h was exponentially related to temperature (r2 = 0·841) but the correlation weakened with time after slurry application. An increase in ammonia volatilization with increasing water vapour pressure deficit was considered to be an effect of temperature.

The ammonia loss rate increased when wind speeds increased up to 2·5 m/s. No consistent increase in ammonia volatilization was found when the wind speed increased from 2·5 to 4 m/s. Ammonia loss after 24 h increased with increasing initial pH of the slurry.

A two-stage pattern for ammonia volatilization from slurry is proposed. During the first stage (the initial 24 h) ammonia loss rate is high due to an elevated pH at the slurry surface followingv application, and temperature significantly affects the loss rate. In the next stage, pH declines and the rate of ammonia volatilization decreases. During this stage other factors, including the dry matter content of the slurry, control the rate of ammonia loss.

Losses of ammonia (NH3) through volatilization from the excretal returns from sheep grazing a range of swards in S.E. England were determined over two seasons. Four swards were examined: grass + 420 or 0 kg N/ha per year, grass/white clover and clover monoculture. These were grazed continuously to sustain a constant sward height of 6 cm. The patterns of NH3-N concentration above the sward changed with treatment and throughout the season: there were, on occasion, irregularities in the concentration profiles which may have been due to differential behaviour of paniculate and gaseous forms. Significant losses of NH3 occurred on all treatments, although on some occasions the low input systems appeared to be acting as sinks rather than sources of NH3. There were trends for the losses to be greater during the middle of the grazing season. In contrast to previous studies there was little relationship between seasonal N losses and changes/differences in dietary N content. The overall losses, whether on a per unit area or per animal basis, were generally in the order, clover > grass + 420 kg N > grass+ 0 N > grass/clover. Mean annual losses from grass + 420kg N and grass/clover were 94 and 1·2 kg N/ha, respectively. For a comparable fertilizer N treatment, losses were lower than previously found with swards which were rotationally grazed by cattle. The present losses, although dependent upon level of N input, whether from fertilizer or from biological fixation, were relatively small, ranging from 05 to 11·2 kg N/ha per year, and represented small proportions of either the N input to the system or the excretal returns from the animals.

Prairie grass (Bromus catharticus) straw (13·7 g N/kg DM) was fed ad libitum to six goats and seven sheep kept in metabolism cages at Palmerston North, New Zealand in 1986 and 1987.

Goats had greater voluntary feed intake (VFI) (56 v. 36 g DM/kg W0·75 per day), greater apparent DM digestibility (36·8 v. 32·6%) and a larger rumen pool of DM and liquid (W0·75) than sheep. Goats also had greater apparent digestibility of fibre, especially of lignin, and greater rumen fractional degradation rates (FDR) of cellulose, hemicellulose and lignin. Goats had a higher rumen ammonia concentration (115 v. 80 mg N/l), lower rumen pH (6·73 v. 6·90), a smaller proportion of large particles and greater proportion of small particles in rumen contents than sheep. Passage through a 1 mm sieve was established as the threshold particle size to have a high probability of leaving the rumen in both goats and sheep. Voluntary water intake/unit DM consumed was lower in goats than in sheep and, whilst rumen fractional outflow rate (FOR) of water and particulate matter also tended to be lower in goats, the difference was not significant. Irreversible loss rate of rumen NH3 and the amount of N calculated as recycled to the rumen, both expressed as mg N/kg W0·75 per day, were greater for goats than for sheep.

It was concluded that the ability of goats to maintain a higher rumen NH3, concentration than sheep, their larger rumen pool and the higher proportion of small particles in rumen contents were all contributing factors to their greater VFI and fibre digestion of this low-quality roughage.

Twenty-eight male castrate kids and 28 male castrate lambs were assigned to a comparative slaughter experiment to compare energy and protein utilization. The experiment was conducted from January to March at Lincoln, New Zealand. Ten animals of each species comprised the initial slaughter group while the remaining 18 animals were randomly allocated to five feeding regimes. Animals were offered high-quality meadow hay for 13 weeks and then slaughtered. The energy retained in the body was regressed against metabolizable energy intake (MEI) to obtain estimates of maintenance energy requirement (MEm) and efficiency of energy utilization for growth (kg).

Kids tended to have a higher MEm (0·44 v. 0·37 MJ ME/kg W0·75 per day in kids and lambs, respectively) and a higher kg (0·28v. 0·22) though these differences were not significant. There was no difference between the species in the composition of gain of the fleece or hair-free body with a protein energy to gross energy ratio in the gain of 0·20. The fleece or hair-free body composition was similar for both species (initial 0·51v. 0·48; final 0·36 v. 0·32; protein energy:gross energy, kids and lambs, respectively). Overall, the values for kg in both species were low and could not be explained solely by the composition of the gain.

The efficiency of use of apparently absorbed amino acid (AA) for fleece or hair-free body protein deposition was higher for kids (0·52 v. 0·29) and, when the fleece or hair was included, the value increased markedly for lambs (0·50 v. 0·41, kids v. lambs, respectively). The protein energy apparently absorbed in the small intestine/MEI was not high or different between species (mean 0·18). The efficiency values for both species are low and not explained by the supply of protein relative to energy, which was not high.

It was concluded that, with high-quality forage used to promote live weight gain, there was little difference between kids and lambs in the utilization of forage energy, and that the efficiency of utilization of absorbed protein for protein deposition was lower than expected in both species.