The effects on the potato crop of the herbicides metribuzin, linuron and metobromuron, singly and in combination with each other and with hand weeding, were investigated and compared with cropped and uncropped weedy controls. Weed competition was determined by measuring weed dry weight. When the three herbicides were applied singly metribuzin gave the lowest weed dry weight, while the combination of metribuzin, metobromuron and hand weeding gave the lowest weed dry weight per unit area of all treatments in both seasons. However, the best potato yield was obtained from the combination of metribuzin and hand weeding. Metobromuron consistently reduced yield through reduction of tuber size irrespective of its effect on weed competition. Metribuzin was the best herbicide for the control of broad leaved weeds and many grasses.

A grazing management study involving the new white clover cultivar Osceola grown in association with Bahiagrass cv. Pensacola was conducted on wet and dry sites with and without pesticide/herbicide treatment. A combination of heavy grazing pressure and a short deferment period before grazing during the summer and autumn resulted in the largest soil seed reserves and seedling recruitment by the white clover component. However, maximum survival of plants was recorded from a combination of a short deferment period and moderate grazing pressure.

A slight increase in plant survival and seedling recruitment was shown on the wet site. Survival potential was least in pasture treated with pesticide and herbicide. Heavy grazing pressure and short deferment periods appear essential for the survival of the clover component. Annual seedling recruitment from soil seed reserves was the primary mechanism of white clover survival from spring to spring in pasture associations based on tropical bahiagrass.

Two field experiments were conducted to evaluate the efficiency of rate, mode and time of application of manganese sulphate to irrigated wheat on manganese-deficient loamy sand (Ustipsamments) soils recently brought under rice. A significant increase in manganese uptake by wheat and marked increases in wheat grain and straw yields occurred following both soil (5, 10 and 20 kg Mn ha-1) and foliar (0.5, 1.0 and 2.0% MnSO4 solution) application of manganese, but foliar application was more effective. A series of sprays initiated before the wheat was first irrigated (26 days after sowing) gave significantly better results than application begun after the first irrigation (32 days after sowing); four sprays of 0.5% MnSO4 solution produced the largest grain yield, followed by three of a 1% solution.

The response of sesame to fertilizer was studied in 11 trials at four different sites, in which phosphorus was applied at rates from 0 to 26 kg ha−1 to the seed bed and nitrogen at rates from 0 to 60 kg ha−1 after thinning. Yield responses, found only to the main effects of nitrogen and phosphorus, are discussed in relation to soil type and previous cropping history of the sites and used to calculate gross cash returns from fertilizer use. Seed oil content was affected by environment, but responses to nitrogen and phosphorus were inconsistent.

If corn-soyabean intercrops are to replace corn grown alone, maintenance of yield and improvements in forage quality are important. Total yields and inter-crop competition in corn-soyabean intercropping experiments have been discussed previously (Putnam et al., 1985). In Part II yield composition, components and protein content are presented for intercrops of varying planting patterns and corn densities. Most intercrop patterns produced more forage than sole crops compared at the same yield ratio of corn-soyabean as in the intercrop harvested mixture. At high plant density ear yields in the intercrop were increased while in sole crops they were reduced or unaffected. The increase was a result of an increased number of ears, weight per ear and kernel, and number of kernels per row in the intercrops. Increases in crude protein content over sole crop corn of 11–51% were shown for the various intercrop treatments. Crude protein yields in the intercrops were greater than either sole crop in 1981 and greater than corn and similar to soyabean in 1982, and greater than the crude protein yields expected from an equal planted area of sole crop corn and soyabean in both years. Intercrop forage quality was improved over sole corn both by the addition of protein from soyabean pods and an increased ear: stover ratio of the intercropped corn.

Smallholders in south-east Tanzania grow the groundnut Red Mwitunde (Arachis hypogaea ssp. hypogaea), a long-season variety with seed dormancy, although earlier maturing varieties with no seed dormancy (ssp. hypogaea and fastigiata) usually yielded better in trials, particularly when late sown. Small-seeded, short-season varieties established better in the field than Red Mwitunde and other large-seeded varieties, especially under moisture stress. Trials with selective fungicides to estimate yield losses from leafspots (mainly Cercosporidium personatum) and rust (Puccinia arachidis), jointly and separately, showed no consistent varietal differences.

Brassica carinata (an uncultivated species) was compared with other oilseed species already under cultivation in India, on irrigated and rainfed soils. B. carinata yielded very well in a wide range of environments partly because it had a large number of primary and secondary racemes. It also showed better environmental adaptation and substantial resistance to pests and disease.

A series of population density and row spacing trials were carried out with sorghum in conditions of increasingly severe water stress from 1980 to 1984. In conditions of reasonable water availability, increasing density resulted in increased leaf area indices, dry weight production and grain yields. In conditions of severely limiting water availability, increasing density resulted in developmental delays, density-dependent mortality and reduced plant dry weights, with little increase in dry weight production per hectare. In these conditions grain yields were reduced by increasing density. The proportion of dry weight allocated to grain declined with increasing density and decreasing plant size, a consequence of the requirement for a minimum plant size to support grain production. When plant size was reduced below a critical range, disproportionately large reductions in grain yield resulted.

Optimum densities for grain production varied from below 10 000 plants ha−1 in dry conditions to over 120 000 plants ha−1 in moist conditions. Medium density treatments produced the largest grain yields on a long term basis, but the low density treatment showed a reduced risk of crop failure and greater yield stability.

Variety × spacing and variety × sowing date trials in south-east Tanzania showed that the long-season spreading bunch groundnut variety Red Mwitunde (A. hypogaea ssp. hypogaea var. hypogaea) was less tolerant of delayed sowing and more tolerant of wide spacing than Natal Common, a short-season upright bunch variety (ssp. fastigiata var. vulgaris). The same variety × spacing and variety × sowing date interactions were detected in groundnuts intercropped with cassava (Manihot esculenta). These observations help to explain the farmers' preference for Red Mwitunde despite the higher trial yields of Natal Common.

Eighteen arabica coffee introductions from the USDA germplasm collection were planted at 23° S latitude in Brazil from 1977 to 1982 to assess their yield potential and frost hardiness, compared with the cultivar Catuai Amarelo which is extensively planted in Brazil. Nine of the introductions had similar yields to Catuai Amarelo. Erecta T966 and Harrar yielded as well as Catuai Amarelo and were also more frost hardy. Tall, early maturing coffee varieties showed the least frost injury.

The productivity, land equivalent ratios (LERs) and light use efficiency of a pear and radish interculture system were assessed. Pear yield was unaffected by intercropping. Relative yields for the radish component varied between 0.5–1.01 depending upon the yield index and spatial arrangement employed. This gave LER values for the system of 1.5–2.01. The overall trans-missivity of the pear canopy was 73%. A 47% reduction in photosynthetically active radiation (PAR) gave a yield reduction of 65% in terms of number of saleable radish, but did not affect total dry matter productivity. Reductions in radish yield directly beneath pear trees was thought to be due to other factors besides PAR. The total dry matter productivity of a system containing five successive radish crops was estimated at 26.25 tonnes ha−1 yr−1.

Plant growth, development and yield were monitored in sorghum/cowpea intercropping trials in Botswana during the drought seasons of 1982/3 and 1983/4. The cowpea proved to be more competitive than sorghum in these arid conditions, resulting in substantial increases in sorghum mortality and developmental delay, and substantial decreases in sorghum leaf area indices, dry weight production and grain yield. Intercropping in standard rows at medium plant densities resulted in Land Equivalent Ratios considerably less than 1.0, in contrast to earlier trials in Botswana in higher rainfall years which demonstrated intercropping advantages. Low density, wide row intercropping resulted in some small intercropping advantages, suggesting that this treatment may have potential in Botswana's variable climatic conditions.

Stands of pearl millet were grown in three controlled environment glasshouses in which were imposed different combinations of atmospheric saturation deficit (1.5 to 2.3 kPa) and soil water content (fully irrigated and not irrigated). Consistent differences in saturation deficit (SD) were maintained throughout the experiment (100 days) but a high water table restricted differences in water supply to the first 40 days. Responses to SD and soil water were observed in some variables but not in others. Developmental processes such as the rate of leaf appearance were unaffected, whereas the efficiency for conversion of intercepted solar energy decreased by 24% and the rate at which the canopy expanded by 50%, in the driest compared to the wettest conditions.

Two experiments were conducted over a two year period at the International Rice Research Institute (IRRI) in the Philippines to evaluate the effect of a soyabean green manure intercrop on the grain yield of maize. Three rows of soyabeans were interplanted at 50 cm spacing between maize rows 1.5 m apart. Forty two days after planting the two outside soyabean rows were ploughed into the maize rows in a traditional ‘hilling up’ operation. At the zero and lower nitrogen rates, the soyabean green manure resulted in significantly larger maize yields than when maize was grown without green manure, providing the equivalent of 28 kg N ha−1 on the zero nitrogen plots. A response to green manure was also seen at 25 and 50 kg N ha−1. Allowing the central soyabean row to grow to maturity provided additional yield and protein.

Nine rice genotypes were classed as tolerant, moderately tolerant, susceptible and highly susceptible, depending on their growth retardation and the intensity of lime-induced iron chlorosis. All except genotypes IET 7972 and IET 7973 showed varying degreees of chlorosis and corresponding decreases in shoot dry matter production. Chlorosis was negatively correlated with the orthophenanthroline reactive Fe2+ iron content of the leaves, the concentration of which accounted for 82% of the variability in chlorosis. However, total iron concentration in the leaves and the Fe/Mn, Fe/Cu and K/Fe ratios were not significantly correlated with dry matter accumulation in the plants.

Reduction of leaf area in sorghum without tillers (uniculm sorghum) might result in conservation of water at early stages of growth and hence in stability of grain yield under dry conditions. In two experiments in south-east Queensland, Australia, tillers were removed by hand to examine the growth of uniculm sorghum. Tiller removal promoted root development at the flag leaf stage but significantly reduced shoot dry matter and lowered grain yield by about 20% in a wet season. The saving in soil water as a result of tiller removal was relatively small but statistically significant.

The ureides allantoic acid and allantoin form a bright pink compound with acidic p-dimethylamino cinnamaldehyde. It is possible to estimate the ureide content of a soyabean petiole segment by crushing it on filter paper and developing the colour with this reagent. There is a strong correlation between the spot test and estimates from colorimetric determination of ureides. In field grown soyabeans, there was a positive correlation between the petiole ureide content during pod fill and the integrated seasonal acetylene reduction activity. The spot test can be used to identify plants with low nitrogen fixation activity.

The effects of sowing pattern and seed rate on the green-chop yield of forage sorghum-sudangrass hybrid (cv. Pioneer 988) and forage sorghum (cv. Abu Sabeen) were investigated. In general, more green-chop yield was obtained from seed of both cultivars drilled in regular rows, double rows and triple rows than when traditionally sown as broadcast seed. Yields were greater when plants were sown at 70 kg seed ha−1 than when traditionally sown at 35 kg seed ha−1. These yield advantages were sustained through three successional harvests. Pioneer 988 generally yielded more than Abu Sabeen, especially when sown in January. Both cultivars gave minimum yield when sown in April, July or October.

This review paper deals with some of the general statistical principles of crop rotation experiments. The basic terminology of the subject is defined. Attention is paid to the fundamental design principle that each phase of a rotation should each year be present in an experiment involving that rotation. The inclusion of ‘auxiliary’ treatments - usually fertilizer treatments - is discussed. Consideration is given to matters influencing the choice of size and shape of plots. The statistician's role in planning a rotation experiment is portrayed as onerous. Emphasis is put on the importance of care and good practice in the management of a long-term agricultural experiment. Complications arising in the statistical analysis of data from a rotation trial are sketched.