Instruments and techniques are reviewed for measuring the content and potential of water in the soil, the water potential and solute potential of plant tissue, and the diffusive resistance of leaves. Emphasis is placed on difficulties encountered in the field and in the interpretation of measurements. References are given to recent technical papers and reviews dealing with the principles and construction of individual instruments.

The paper outlines ways in which agricultural scientists fail to benefit fully from collaboration with statisticians and mathematicians and from the use of computers. These result mainly from misunderstandings and wrong attitudes; the aim here is to correct them by frank examination. A statistician can easily point comment on how his attempts to cooperate with others are rendered ineffective, and many examples are quoted. He must also recognize and seek to correct his own weaknesses; the paper has no intention of minimizing these, though its chief concern is with other matters. The emphasis throughout is on statisticians and mathematicians being collaborators in research, not merely purveyors of specialized services, on computer use being coherently planned, and on scientific management being concerned with priorities.

Instruments and techniques are described for measuring the interception of light by the canopy of a crop, the CO2 exchange of individual leaves, leaf extension rate, and root distribution. Suitable systems for recording are also reviewed. Names and addresses of manufacturers are cited for a few representative models.

A continuous daily rainfall record provides much information for agriculturalists but initial questions are imprecise and are modified in the light of experience. In order to match this uncertainty the program ‘Rainfall Statistics’ is designed as a simple interactive procedure for studying quantity probabilities, dry spell frequencies and a simple soil-moisture budget. The program has various functions depending upon user requirements. Use of the program is illustrated with rainfall records from northern Thailand.

A random sample of 13 uniformity trials was examined in terms of sensitivity to choice of block size and orientation. Five trials were classified as insensitive and eight as sensitive. Trials insensitive to blocking choices showed little variation in experimental error for various sizes, shapes and orientations of blocks, but those sensitive to blocking choices showed considerable, sometimes drastic, effects of such choices. There was often an objectionable degree of inflation of experimental error due to large or improperly oriented blocks. It is proposed that the routine analysis of field experiments based on block designs be modified to reveal cases in which ineffective blocks have been used on sensitive sites.

Minimal competition was observed between adjacent plots of climbing bean varieties and maize in association, but there was a large head border effect which was not uniform among varieties. Head borders of 1 m at each end of the plot are recommended and a net plot size of 11 m2 with 3 replications for climbing bean varieties associated with maize was found to be adequate to detect significant differences in yield of beans of 24% of the mean. For climbing beans in monoculture a somewhat smaller plot (8 m2) would achieve the same level of precision. Larger plot sizes in general were needed for maize comparisons, especially in association.

Genotype-environment (GE) interactions as a component of historically rising crop yields are reviewed. For various cereals (both temperate and tropical), past selection has seemingly evoked responsive varieties, and GE effects constitute about one third of the total estimated yield increase due to G + E + GE. In potatoes and sugarcanes, more stable (less responsive) varieties, less dependent upon good environments to do well, have been prominent. Historically, responses have been largely due to unconscious selection, but GE effects could be deliberately manipulated by breeding if it were desirable to do so, for example, to produce varieties adapted to low-input agriculture (already an object of breeding research in a few crops).

When agricultural experimenters send data to a statistics unit for analysis, they do not usually include a field plan of the experimental site unless specially requested to do so. But such a plan can be a great help to a biometrician: he can use it to check that the design used in the field was correctly described, and to arrange the plot residuals in the field layout as a way of detecting a fertility trend in the site and of relating anomalous values to particular locations on the ground. This paper gives examples of how a field plan can be used to the benefit of the experimenter.

Our experience with potentially tropic-adapted grain legumes leads us to hypothesize that ‘with appropriate and sufficiently proven technology, developed specifically for the purpose, controlled environment facilities can be a powerful adjunct to field research and so assist in the attainment of crop improvement objectives in lentils’. This contribution comments on the ‘adaptation’ of grain legume crops to their environments, describes current lentil production in the USA and elsewhere, and discusses the tenets and objectives of a programme of research devoted to the ultimate release of lentil genotypes that are well adapted to the environments for which they are intended.

A field technique is described in which trenches were dug and vertical polythene partitions installed to eliminate below-ground interactions between component crops in intercropping. With pearl millet/groundnut the control intercrop (i.e. not dug out) gave a yield advantage of 22% compared with sole cropping. A treatment dug out but refilled without polythene partitions was very similar to the control intercrop, indicating that the digging procedure per se had little effect on growth. An intercrop treatment in which below-ground partitions were installed gave a yield advantage of 19% over sole cropping, suggesting that the main determinant of yield advantage was an above-ground interaction between the canopies. However, partitioning decreased millet yield and increased groundnut yield, indicating that below-ground interactions were important in determining the competitive balance of the two crops. It is concluded that the partitioning technique is promising.

The effects of date of planting and soil mulching on cowpea growth and yield were investigated for three years during the middle part of the minor season in Sierra Leone. Mulching at 8 tons/ha conserved more soil moisture than other mulch rates, with greatest difference in the top 15 cm of soil, especially during drier periods. Plant height, pod number and seed yield of cowpea planted in September were significantly greater than from other planting dates whether mulch was applied or not. The beneficial effects of mulching at 8 tons/ha on cowpea growth and yield were more dramatic when the crop received an average weekly rainfall of 100 mm before 50% flowering.

Twelve varieties of sugar-cane grown in Barbados, West Indies, during the past 40 years are compared with respect to cane yield performance in trials and in agriculture. Comparisons are based on data-pairs matched for site, season and crop, using both means and variety-on-variety regressions. Conclusions are that: trials greatly over-estimate mean agricultural differences between varieties; variety-on-variety regressions reveal numerous GE effects and account for much variance, but agree poorly as between trials and agriculture and are not useful in improving prediction; in view of the universal use of trials for prediction and decision-making, it is a matter of some practical importance that their validity should be tested in other crops.

In recent years many economists have been posted to agricultural research stations in Africa but their roles and methodologies have not been clearly defined. Maize research in Nigeria has an exceptionally good structure, good facilities and highly trained manpower, yet the innovations recommended to farmers are not entirely satisfactory. Deficiencies are the result of inadequate attention to the way innovations fit into the farm-household system, identification of farmer's objectives, the uses made of maize and problems of resource shortage. Pilot studies demonstrated how these deficiencies may be rectified by an economist, and suggest changed production recommendations and research priorities.

An account is given of large-scale trials of the use of saline water for the trickle irrigation of forest trees in the desert areas of Abu Dhabi, in eastern Arabia. Water of salinities up to an equivalent of 16,000 ppm have been used on Casuarine glauca and Eucalyptus camaldulensis. Growth rates ranged from 1.7 m in 7 years on salinities of 16,000 ppm to 10–12 m in 8 years on salinities of 2000–2500 ppm. Trickle irrigation appears to allow the use of more saline water than can be used with other systems, perhaps because plant roots are not subject to the fluctuations in moisture stress that occur under all other systems.

A technique is described that allows neutron meter access tubes to be reinstated in tilled soil without disturbance below the depth of ploughing or tine cultivation. In a test there was no evidence of greater water movement to depth around reinstated access tubes compared with newly inserted ones.

Tea yields of 1 to 4 t ha−1y−1 of dry shoot tips are much less than those of other vegetative crops like grasslands or forests growing in similar conditions, partly because plucking restricts tea biomass production, but mainly because the harvest index of tea is small. In Kenya, plucked tea produced 36% less biomass per hectare per year than unplucked tea, and 64% less wood. Only 8·3% of the total annual biomass increment was harvested. This proportion might be increased by plucking older leaves, increasing the shoot:root ratio, and by lowering the plucking table so that less wood was produced.

The ‘banjhi’ or dormancy cycle in free-growing tea trees appears to be the result of an endogenous rhythm. However, in harvested tea apical dominance causes dormant shoots to be produced under conditions favourable to growth, and the synchronized flushing cycle causes periods of apical shoot (active terminal bud) and secondary shoot (dormant terminal bud) production.

The effect on maize yields of ground rock phosphate, alone or mixed with sulphur in either of two proportions, and of single-superphosphate, was measured at five consecutive harvests. At the first harvest single-superphosphate increased yield by 1.54 t and the best of the mixtures by 0.56 t grain/ha. Ground rock phosphate alone had no effect on yield. At the third harvest there were no yield differences between sources; because yields produced by rock phosphate and the mixtures improved. The cumulative increase in yield produced by single-superphosphate was 3.58 t/ha and by rock phosphate 2.69 t grain/ha, but because the fertilizer cost ratio was 3:1, additional grain produced by rock phosphate cost half as much. The apparent P recovery ranged from 5.40 to 8.28 kg/ha, with most from single-superphosphate and least from the mixture containing most elemental-S.

The unit for separating juice from which 5 to 10 kg of leaf protein could be made in a day has been modified so as to make it more efficient and more convenient to use.

Nitrogen fertilizer requirements and potential forage yield of wheat, oats and Italian ryegrass grown under irrigation in Saudi Arabia were assessed. Mexipak wheat produced high DM yields with between 100 and 150 kg/ha N, and the mode of nitrogen application did not significantly affect DM outputs. Nitrogen at 150 kg/ha produced highest DM yields for oats, although the application of 50% of irrigation requirement as drainage water reduced yield by 8%. Italian ryegrass produced a mean yield of 14.6 t/ha in 169 days with a nitrogen application of 60 kg/ha; direct drilled on a field scale it yielded 8.76 t/DM/ha with a total application of 270 kg N/ha.