In experiments conducted between 1982/83 and 1984/85 on the winter barley cv. Igri, nitrogen, including slow release forms, applied at several growth stages in the autumn, increased vegetative growth, including vegetative tiller numbers, nitrogen content and greening, particularly in the early sown crops. By spring, these effects were less apparent and only very infrequently did autumn nitrogen applications result in an increase in fertile tiller number and yield, although the autumn nitrogen treatments had been tested in combination with several different sowing dates and spring nitrogen levels. Studies on leaf production over the winter showed that the additional leaf produced after nitrogen application was subsequently lost through leaf senescence (especially in the severe winter of 1984/85) and therefore the treated and untreated plots often recommenced growth in the spring with similar amounts of live leaf tissue. It is argued that the use of autumn nitrogen on winter cereals in the north of Scotland cannot be supported agronomically and may also be undesirable for environmental reasons.

Experiments in 1985 and 1986, at Woburn Experimental Farm in Bedfordshire, tested the effects of fungicides, applied in autumn, and a growth regulator, applied at GS3O–31 or GS32–33 in spring, on winter barley grown on two contrasting soil types in each year. Leaf diseases did not become severe in any of the experiments but take-all (Gaeumannomyces graminis var. tritici) was prevalent in 1985. Triadimenol (‘Baytan’) was more effective than flutriafol (‘Ferrax’) in decreasing the severity of takeall and its activity against the disease was related to earliness of sowing.

Mean responses in grain yield to the fungicide treatments were mostly small and not significant but did not conflict with the hypothesis that crops on lighter soils benefit more from autumn fungicides than those on heavier soils. Mean effects of the growth regulator sprays were also small but they interacted with both soil type and season. Over the 2 years the later spray applied to crops on the heavier soil gave the largest mean response. Sprays applied to crops on lighter soils were often detrimental to yield, especially in 1985.

Soil chemistry in relation to yield of spring barley (Hordeum vulgare, cv. Golden Promise) was investigated in three field experiments in north-east Scotland in 1982. Previous treatments had produced a range of soil pH values from 4·0 to 6·0 (measured in 10mM-CaCl2 solution). The experiments demonstrated the values for soil acidity components below which yields were reduced. Large additions of phosphate fertilizer allowed barley to withstand larger concentrations of soluble and exchangeable aluminium (the principal acid soil toxin), thus decreasing the critical soil pH from 5·2 to 4·8 (corresponding to about 5·9 and 5·5 if measured in a water suspension), and increasing the critical soluble Al concentration from 10 to 50 UM and the critical Al saturation of the cationexchange complex from 0·025 to 0·1.

To assess the progress made in modern breeding of open-pollinated sorghum cultivars, 24 contemporary and improved sorghum cultivars from ICRISAT (International Crops Research Institute for the Semi-Arid Tropics, India) were compared with 40 sorghum landraces collected in Ethiopia. The experiment was performed at Bet Dagan, Israel, in 1988 under near-potential (dripirrigated) and dryland (stored soil moisture) growing conditions. Data on phenology, biomass, harvest index, grain yield and its components were collected. ‘Drought susceptibility index’ (S) was calculated for each genotype as the reduction in yield from irrigated to dryland conditions relative to the mean reduction for all genotypes.

On average, cultivars produced about three- to fourfold more grain than landraces. Cultivars had the same mean growth duration and biomass but were 40% (irrigated) to 27% (dryland) shorter in mean plant height than landraces. This yield advantage was fully accounted for by a proportionately greater mean kernel number per panicle and harvest index, irrespective of the water regime. ‘Drought susceptibility index’ was near average for all cultivars and much more variable among landraces. Dryland yield became progressively dependent on the yield potential as the yield potential was genetically improved in the cultivars.

It was concluded that, irrespective of the water regime, the large yield improvement in modern open-pollinated cultivars of sorghum, as compared with landraces, has been achieved only through greater dry matter partitioning to the panicle accompanied by a reduction in plant height. The genetic improvement of yield in open-pollinated cultivars and hybrids of sorghum is discussed.

The effect on soil fertility and crop performance of different organic fertilizers; paddy straw (PS), farmyard manure (FYM), water hyacinth compost (WHC) and tank silt (TS), at different rates of application and in combination with N fertilizer, was studied in a rice-based cropping system on an acid lateritic soil at Kharagpur, India, during 1985/86. Organic manuring of wet-season rice (first crop) with 5 t PS/ha 10 days before transplanting and 10 t FYM or 10 t WHC/ha at transplanting increased grain yield as much as the application of 30 kg N/ha. Increasing the rates of FYM and WHC application up to 15 t/ha increased yield but increasing the rate of PS beyond 5 t/ha did not. Response to increasing amounts of N was not linear; there was a significant increase up to 90 kg N/ha and a decrease when N was applied in conjunction with organic fertilizers. There was a significant increase in the N uptake of the rice but a decrease in the recovery of applied fertilizer N with the application of increasing rates of organic and N fertilizer.

The organic C content of the soil after the rice harvest increased significantly after PS application, whereas there was more available N after WHC and FYM. Increasing the rate of application of PS up to 15 t/ha increased organic C but not available N. Mineral N fertilizer had little effect on fertility build-up. Grain yields of wheat and gram (Cicer arietinum), grown after rice without any additional fertilizer, increased significantly. The residual N effect of the previous crop on wheat or gram yield was small and adding fertilizer directly is considered essential for higher productivity in these crops in a rice-based cropping system.

The effect of phosphorus fertilizer on the performance of an improved semi-dwarf (100 cm), long duration (170 days), photosensitive rice variety CR 1016, grown under conditions of intermediate deepwater (15–50 cm) and simulated flash flooding (80 cm), was investigated at Cuttack, India in 1985, 1987 and 1988. Growth and yield of rice subjected to overhead flash flooding (complete submergence) at the vegetative stage for 10 days was decreased significantly compared with partial submergence in water of intermediate depth. The adverse effect of complete submergence was due to greater tiller mortality and lower number and/or weight of panicles at maturity. Rice plants fertilized with P tolerated flooding better and produced significantly higher grain yields than when no P was applied. There was no significant effect of P application on the crop grown under natural unsubmerged conditions, except when it was raised in the same plots as the previous year and may therefore have benefited from residual effects. Therefore, application of 8.7–17.5 kg P/ha along with 60 kg N and 16.7 kg K/ha at sowing was essential to overcome the deleterious effects of submergence on rice grown under waterlogged conditions.

Quinoa produces a cereal-like grain with a higher protein content and a better balanced amino acid composition than the major cereals. It is cultivated at high altitudes in the Andes, and is believed to have potential for temperate regions. In the development of quinoa as an arable break crop, sowing date, sowing rate and row spacing are identified as agronomic variables having a high priority for investigation. The variety Baer, from high latitudes at sea-level in Chile, and the variety Blanca de Junin, from the inter-Andean valleys of Peru, were sown on 25 March, 14 April and 7 May 1982 at spacings between rows of 0·8 and 0·4 m and sowing rates within rows of 0·2, 0·4 and 0·6 g/m, at Cambridge, England. These varieties were chosen for their strongly contrasting origins and plant types. In another experiment sown on 15 March 1984, Blanca de Junin was replaced by another valley variety, Amarilla de Marangani, somewhat better adapted for cultivation in England. Between-row spacings of 0·4 and 0·2 m, and sowing rates of 15, 20 and 30 kg seed/ha were used. Weed competition was more intense after later sowings, causing the plots sown in May to be abandoned. At the higher sowing rates, plants were shorter, a higher proportion were stunted, branching was reduced and maturity was earlier. These effects were more marked in the valley varieties, particularly Blanca de Junin. Increases in within-row density caused greater increases in competition effects than corresponding reductions in row width, except for the effect on the proportion of branched plants. The highest grain yield, 6·96 t/ha, was obtained with Baer sown in March in rows 0·2 m apart at 20 kg seed/ha. However, Amarilla de Marangani produced a higher yield at 30 kg seed/ha than at 20 kg seed/ha, which is surprising since valley varieties are normally sown with low target population densities or in intercrop, and Amarilla de Marangani was, in other respects, less tolerant of competition than Baer. The implications of these results for the cultivation of quinoa in mechanized farming systems, both in temperate and in highland tropical regions, are discussed.

Hand harvesting is a major constraint to lentil production in North Africa and West Asia. This study, in north Syria, compared hand harvesting, cutting by mower (double-knife) and cutting with angled blades on two lentil cultivars differing in standing ability and using two sowing methods (broadcast and drilled) both with and without the use of a heavy bar for field levelling in the 1984/85 season. Seven treatments were selected for testing in five locations in the 1985/86 season; and in the 1986/87 and 1987/88 seasons, agronomic comparisons of mowing v. hand harvesting were conducted on five farmers' fields.

Both machine methods of harvesting resulted in significant harvest losses compared with hand harvesting. The angled blades performed well on a ridged broadcast crop, but tended to mix soil with the harvested crop. The loss of straw associated with harvesting by mower was reduced by levelling the seedbed after sowing. The superiority in seed yield of cultivar 78S26002 over the local cultivar increased from 9% when hand harvested to 39% with mowing because of its lower likelihood of lodging. In the 1986/87 and 1987/88 seasons, the seed yield from a hand harvest was 1650 kg/ha compared with 1508 kg/ha following harvest by mower, representing a loss of 8·6% from mechanization. The corresponding straw loss was 16·6% of the mean from a hand harvest of 2140 kg straw/ha. However, the harvest losses from mechanical harvesting by mower were compensated for by the reduced labour costs compared with hand harvesting.

Lentils were grown in 1984/85, 1985/86 and 1988/89 in Canterbury, New Zealand. Results showed that lentil canopies were capable of intercepting a maximum of 95% of incident solar radiation at a leaf area index of 7. Autumn sowings attained canopy closure, but late spring sowings did not. At the highest population density used (500 plants/m2), only 65% of incident solar radiation was intercepted by a late spring-sown crop. In all three growing seasons there was a highly significant positive relationship between cumulative absorbed radiation and dry matter (DM) production. Over all seasons, 1·6 g DM/m2 was produced per MJ of absorbed photosynthetically active radiation (PAR). The final utilization coefficient (u) was not affected by sowing date but in a late-sown, diseased crop, u was 38% less than in a disease free crop. In 1984/85, dryland crops had a utilization coefficient of 1·39 while irrigated crops had a u of 1·54. Growth efficiency of lentils was low, at 2·5%.

Experiments to evaluate the suitability of lentil lines using an innovative dual-season approach for improved and reliable production under the harsh environmental conditions of the highlands of west Asia were conducted in Balochistan for 3 years (1986/87, 1987/88 and 1988/89) by the Arid Zone Research Institute. Two ICARDA-bred lines, ILL5865 and ILL5677, when assessed using a wide range of seeding times, appeared to be sufficiently flexible and well adapted to the conditions of extreme cold and aridity to be able to replace the ultra small-seeded local landrace. These ICARDA lines have seed sizes approximately double those of the local landrace and results of consumer acceptability studies suggest that the lines will have desirable marketing qualities. These two lines were selected for further larger-scale agronomy trials at multiple sites in Balochistan in subsequent seasons. Other lines, ILL5690, ILL1939 and ILL5730, have also shown potential, but only at certain seeding times and, as such, will be retained for further assessment in ongoing germplasm evaluation trials.

Initial findings indicate that there are now good prospects for identifying dual-season lentil lines suitable for introduction into highland farming systems. However, further research will be required before new lentil lines will be an attainable reality for a large proportion of the dryland farming community of west Asia.

The intake of temperate forages by herbivores can be determined using the C33 alkane naturally present in the forage and dosed C32 alkane. To determine whether the technique can be used with tropical forages, the concentrations of C33 alkane were determined in seven species.

The leaves of Brachiaria decumbens, Digitaria decumbens, Pennisetum glaucum and Stylosanthes scabra contained sufficient C33 alkane (> 50 mg/kg DM) for the estimation of dry matter intake. Low concentrations of C33 alkane were found in the leaves of Setaria sphacelata, Sorghum sp. and Leucaena leucocephala. Setaria sphacelata was found to contain high concentrations of C27 alkane and therefore intake could be estimated by using C28 as the dosed alkane although the intake would possibly then be underestimated by c. 8%. Leaves of Sorghum sp. contained > 50 mg C31 alkane/kg DM and, with C32 as the dosed alkane, intake would possibly be underestimated by c. 5%. Leucaena leucocephala contained insufficient alkanes to estimate forage intake. The concentration of n-alkanes in Pennisetum glaucum and Sorghum sp. leaves decreased with age.

It was concluded that some tropical forages contain insufficient C33 alkane for the measurement of intake by the double alkane technique. In these species, a shorter chain length alkane can sometimes be used, but only with a reduction in accuracy in estimating intake.

Herbage samples were collected in late June, and again in early September (1987) from permanent mountain meadows situated in North-western Spain (León). Botanical composition was determined by grouping the species into grasses, legumes and herbaceous ‘weeds’. Rumen degradability of the dry matter (DM), crude protein (CP) and neutral detergent fibre (NDF) of the botanical groups in each harvest season was determined by the nylon-bag technique.

Grasses and legumes differed in their degradation characteristics, and the comparisons between forage species were different for the two harvest seasons. There were no significant differences between September grasses, June legumes and September legumes in the potential degradabilities of the DM (86·9, 87·1 and 88·4%) and CP (94·2, 92·4 and 93·8%). The NDF of grasses harvested in September was degraded to a greater extent (80·6%) than that of legumes from both harvests (70·9 and 73·6% in June and September respectively). However, June grasses showed significantly lower potential degradabilities for the DM (77·0 %), CP (84·3%) and cell wall (65·9%). Legumes were degraded at a faster rate than grasses (mean DM degradation rates of 0·143 and 0·057 respectively), and lag times were longer for grasses (4·7 h) than for legumes (2·5 h). Rumen degradation characteristics of the herbaceous ‘weeds’ were estimated, but it should be remembered that many other factors may limit their utilization by ruminants.

It was concluded that the major constraint to the nutritive value of these permanent swards would be the maturity of the grasses harvested in June, which markedly reduced the rumen degradability of the forages.

Various physiological characteristics of castrate male sheep fed ad libitum were studied between 1986 and 1987 in Australia. The sheep were between 20 and 45 kg live weight (LW) before and during fattening. Sequential data on body composition, estimated from TOH space (starting at 10 kg), established that the weight gain contained 19% fat and 13% protein below c. 26 kg LW and 55% fat and 10% protein above. For a given feed intake, the rate of fat gain was constant but the rate of body protein gain was 45% lower above 26 kg LW. Voluntary feed consumption per unit weight peaked at 26 kg LW and declined thereafter. There was an increase at 45 kg LW in weight of rumen digesta per unit feed intake and in digestibility; metabolizable energy (ME) per kg feed was increased as a result by 6–9%. Efficiency of use of ME, determined by calorimetry, increased from 44% in the lower weight range to 53% in the higher; utilization of nitrogen decreased from 37 to 28%. Partition of ME and energy gain suggested that the energetic efficiency of neither fat nor protein deposition changed between the two weight ranges. Change in the amount of DNA (‘cell number’) rather than g tissue/mg DNA (‘cell size’) accounted for weight change in three muscles and most digestive organs. Exceptions were the rumen, in which cell number and size both contributed to gain and the large intestine, in which cell size diminished during growth. It was concluded that fattening in sheep reflects diminution of the protein content of weight gain due to a decrease in the efficiency of protein utilization at the tissue level.

The Agricultural and Related Industries Study Group of the Operational Research Society held a one-day meeting on ‘Modelling Agriculture under Pressure’ on 22 May 1991 at the Open University, Milton Keynes, UK. The group promotes the use of the scientific method in solving management problems. The aim of this meeting was to look at different ways of modelling the effect of changes on farm economics and profitability. Three papers discussed partial-equilibrium economic models and national policies. The other three papers concerned farm profitability. The speakers generated lively discussions. One of the most memorable replies was from Peter Midmore to the question of what he would use instead of input-output analysis: ‘1.2’.