A method for quantifying treatment comparisons for a situation in which there are too many zeros in the dataset for a conventional analysis of variance to be valid is presented. The method assumes the existence of a latent variable such that zero observations correspond to values below a threshold, and non-zero observations are transformed to fit the part of the distribution above the threshold. The method is known as Tobit analysis in econometrics. Parameters are estimated by maximum likelihood and standard errors obtained, all using standard numerical routines. Use of the method is demonstrated by analysis of a dataset of crop lodging, and it is anticipated to be widely applicable to other types of data for which high numbers of zeros prevent conventional analysis.

Field experiments were conducted over 3 years to assess the effect of a triazole fungicide programme, and additions of strobilurin fungicides to it, on nitrogen uptake, accumulation and partitioning in a range of winter wheat cultivars. Commensurate with delayed senescence, fungicide programmes, particularly when including strobilurins, improved grain yield through improvements in both crop biomass and harvest index, although the relationship with green area duration of the flag leaf (GFLAD) depended on year and in some cases, cultivar. In all years fungicide treatments significantly increased the amount of nitrogen in the above-ground biomass, the amount of nitrogen in the grain and the nitrogen harvest index. All these effects could be linearly related to the fungicide effect on GFLAD. These relationships occasionally interacted with cultivar but there was no evidence that fungicide mode of action affected the relationship between GFLAD and yield of nitrogen in the grain. Fungicide treatments significantly reduced the amount of soil mineral N at harvest and when severe disease had been controlled, the net remobilization of N from the vegetation to the grain after anthesis. Fungicide maintained the filling of grain with both dry matter and nitrogen. The proportionate accumulation of nitrogen in the grain was later than that of dry matter and this difference was greater when fungicide had been applied. Effects of fungicide on grain protein concentration and its relationship with GFLAD were inconsistent over year and cultivar. There were several instances where grain protein concentration was unaffected despite large (1·5 t/ha) increases in grain yield following fungicide use. Dilution of grain protein concentration following fungicide use, when it did occur, was small compared with what would be predicted by adoption of other yield increasing techniques such as the selection of high yielding cultivars (based on currently available cultivars) or by growing wheat in favourable climates.

The effects of additions of four organic amendments to an Oxisol on soil pH, exchangeable and soil solution Al, extractable P and maize growth were compared with those of lime and P in a glasshouse experiment. The organic amendments were: grass residues, household compost, filter cake (a waste product of sugar mills) and layer poultry manure and they were added at a rate equivalent to 20 t/ha with or without the addition of either lime (equivalent to 0, 5 or 10 t/ha) or P (equivalent to 0, 10 or 50 kg/ha). Lime applications to the control (unamended) treatment resulted in a marked reduction in exchangeable Al, and concentrations of total (AlT) and monomeric (Almono) Al in soil solution and the proportion of AlT present as Almono. In unlimed soils, amendment additions increased pH and decreased exchangeable Al in the order: poultry manure>filter cake>household compost>grass residues>control. The addition of all the organic amendments increased soil pH and reduced AlT and Almono to low concentrations regardless of whether lime was applied or not. As a result, there was no yield response to applied lime in any of the amended treatments. Olsen-extractable P concentrations in soils followed the order: poultry manure>filter cake>household compost>grass residues>control and there was no yield response to applied P for the poultry manure and filter cake treatments. The decrease in adsorption of subsequently added P induced by additions of organic amendments followed the order: poultry manure>household compost>filter cake>grass residues. It was concluded that the addition of organic amendments to acid soils is a practicable way of liming them and reducing the potential for Al toxicity and it can also reduce fertilizer P requirements.

A field experiment lasting 9 years was initiated in 1987 to study the effect of integrated nutrient management involving incorporation of wheat straw (WS) or farmyard manure (FYM) alone or in combination with inorganic fertilizer on a fine loamy mixed hyperthermic udic Ustochrept. Soil physical properties and the productivity of a rice–lentil cropping sequence were examined in a dry land farming system.

At the end of the ninth cropping season, addition of organic materials had increased organic carbon, aggregate stability, moisture retention capacity and infiltration rate of the surface soil while reducing the bulk density. Application of inorganic fertilizer alone decreased the stability of macro aggregates and moisture retention capacity but increased the bulk density values. Treatment effects on moisture retentivity were more pronounced in the higher (0·3–1·5 MPa) suction range. During the first six experimental years, sole application of inorganic fertilizers produced 10–17% higher grain yield of rice, compared to sole application of organic sources or combined organic and inorganic sources. Annual applications of wheat straw and farmyard manure gave higher grain yields of rice from the seventh year onwards. Grain yields of lentil were higher with organic sources either alone or combined with inorganic nutrients.

A significant genotype–environment interaction can limit gains in selecting superior genotypes since the best genotype in one locality may not be the best one in another locality. The efficiency of indirect selection is related to the heritability of the trait and to the genetic correlation between localities. Since a major objective of Eragrostis curvula (lovegrass) breeding programmes in Argentina is to select superior genotypes from new cultivar introductions, research on the relative effectiveness of direct and indirect selection is necessary in order to identify localities under which the efficiency of indirect selection could be maximized. To do this, experiments using a set of 18 hybrids were carried out in three localities from the semi-arid region of Argentina. Aerial biomass yield, leaf length, crown diameter, panicle number and panicle length were measured and an analysis of variance and covariance were performed. The high genetic correlation obtained suggests that the genetic mechanisms involved in the expression of these traits would be the same, or at least very similar, in Río Cuarto, Villa Mercedes and Bahía Blanca environments where the hybrids showed a stable performance. In general, the heritability obtained was higher in Villa Mercedes and Bahía Blanca than in Río Cuarto. For a given selection intensity the expected correlated response in Villa Mercedes and Bahía Blanca to indirect selection was compared with the expected gain with direct response to selection in Río Cuarto. For dry matter yield, indirect selection in Bahía Blanca did not differ from selection done in Río Cuarto (−3%). However, for performance in Villa Mercedes indirect selection was shown to be less efficient (−29%). For the other traits considered, indirect selection was variable and less efficient, ranging from −18 to −34%. The estimated decrease in efficiency of selection for all the traits considered in the locality of Río Cuarto was higher than in the other localities, suggesting that this place was not a good choice for carrying out indirect selection. Conversely, genetic progress would be faster if selection was carried out in Villa Mercedes, because the evaluations provide higher and more accurate estimates of the heritabilities than in the other localities.

A collection of 112 accessions of common beans from the north of Spain has been evaluated for 24 characters during 2 years. A combination of cluster plus canonical discriminant analysis using morphological traits to generate canonical functions was able to separate landraces from different geographic areas of northern Spain, assigning them to germplasm groups related to their centres of origin. The accessions were clustered in three groups in both years. Two groups were related to races with South American origins (races ‘Chile’, ‘Peru’ and ‘Nueva Granada’), the other group was related to races with Mesoamerican origins (races ‘Mesoamerica’ and ‘Durango’). The variation found was higher in the west of the geographic area studied decreasing toward the east. These results will permit a more efficient management and conservation of these resources, and their efficient use in breeding programmes.

Faecal indices were evaluated to assess organic matter digestibility (OMD) for small ruminants fed with Digitaria decumbens. A continuous digestibility trial was conducted with five Creole bucks and five Black-belly rams, fed with fresh cut Digitaria decumbens from 15 to 70 days of re-growth. The amount of herbage offered, the refusals and the faeces were weighed daily for each animal during 55 days. Herbage and faeces samples were kept to determine dry and organic matter (DM and OM), crude protein (CP), neutral and acid detergent fibre (NDF and ADF) and lignin contents (ADL). OMD was calculated per week and per animal. Predictive regressions of OMD were calculated for each species, from all the faecal components measured (CPf, NDFf, ADFf and ADLf), using linear or curvilinear models. Regressions from CPf had the lowest residual standard error (<0·028) when calculated for each species, compared with regressions from the other faecal components. Among the regressions from CPf, the lowest standard error of the predictive parameters was obtained with a model in the form of a-b/CPf, for each species. The latter allows prediction of OMD at pasture, consistent with theoretical energy requirement for both goats and ewes grazing a Digitaria decumbens based sward.

Data from six studies with male broilers fed diets covering a wide range of energy and protein were used in the current two analyses. In the first analysis, five models, specifically re-parameterized for analysing energy balance data, were evaluated for their ability to determine metabolizable energy intake at maintenance and efficiency of utilization of metabolizable energy intake for producing gain. In addition to the straight line, two types of functional form were used. They were forms describing (i) diminishing returns behaviour (monomolecular and rectangular hyperbola) and (ii) sigmoidal behaviour with a fixed point of inflection (Gompertz and logistic). These models determined metabolizable energy requirement for maintenance to be in the range 437–573 kJ/kg of body weight/day depending on the model. The values determined for average net energy requirement for body weight gain varied from 7·9 to 11·2 kJ/g of body weight. These values show good agreement with previous studies. In the second analysis, three types of function were assessed as candidates for describing the relationship between body weight and cumulative metabolizable energy intake. The functions used were: (a) monomolecular (diminishing returns behaviour), (b) Gompertz (smooth sigmoidal behaviour with a fixed point of inflection) and (c) Lopez, France and Richards (diminishing returns and sigmoidal behaviour with a variable point of inflection). The results of this analysis demonstrated that equations capable of mimicking the law of diminishing returns describe accurately the relationship between body weight and cumulative metabolizable energy intake in broilers.

The gastrointestinal tract (gut) contents of male Zebu cattle (Bos indicus) were studied in the southern Sahelian zone of Mali. Measurements were carried out on 39 animals in mid-dry season (MDS) and 15 at the end of the dry season (late dry season, LDS).

After overnight fasting, the animals were weighed and then slaughtered. The different gut compartments were emptied and representative samples of their fill were analysed for contents of dry matter (all) and NDF (reticulo-rumen only).

Related to fasted live weight (FLW), gut fresh matter (FM) fill (g FM/kg FLW) increased from 165·7±4·81 g in MDS to 227·5±3·05 g in LDS (P[les ]0·001), liquid gut contents (g fluid/kg FLW) increased from 138·9±4·65 to 193·5±2·49 g (P[les ]0·001) and gut dry matter fill (g DM/kg FLW) increased from 26·8±0·88 to 34·0±0·97 g (P[les ]0·001). Fresh matter content of the reticulo-rumen accounted for 77% and 80% of the entire gut FM fill in MDS and LDS, respectively. While the NDF component in reticulo-rumen fill (g NDF/kg DM) increased from 778·5±5·63 g in MDS to 836·6±6·37 g in LDS (P[les ]0·001), the ratio between dry matter and fluid in the reticulo-rumen (g DM/g fluid) decreased from 193·2±10·17 g in MDS to 169·0±5·54 g in LDS (P[les ]0·05).

The data support the notion that African breeds of Zebu cattle adjust to deteriorating feeding conditions by increasing reticulo-rumen dry matter and fluid contents. In-depth studies are needed to quantify the physiological benefits resulting from this strategy.

To clarify the role of the incisors in selective grazing by cattle and horses, experiments were conducted in July 1999, using three Holstein cattle and three thoroughbred horses to investigate bite weight and bite force using hand-constructed swards with 15 loadcells spaced 6×12 cm apart. Leaves of perennial ryegrass (Lolium perenne: PRG) were mixed with leaves of tall fescue (Festuca arundinacea: TF) or reproductive culms of reed canarygrass (Phalaris arundinacea: RCG).

There was little difference between cattle and horses in bite rate, but DM intake rate of horses was almost twice that of cattle in the PRG/TF mixture. In the PRG/RCG mixture, both animals tended to graze only leaves of PRG. Total bite force was 3270–3280 N/min in cattle and 5350–6350 N/min in horses in the PRG/TF mixture. Means of maximum values of one bite force (not total force per bite) were 88±5·5 N in cattle and 141±11·0 N in horses. These values corresponded to 1·6±0·10% of body weight in cattle and 2·2±0·16% in horses, after converting Newton force into kilogram force. Bite area was higher in horses than in cattle. Both cattle and horses did not discriminate between leaves of PRG and TF. In the PRG/RCG mixture, horses prehended both leaves and culms in 33% of total clumps grazed, compared with only 5% in cattle. Many culms escaped from the mouth at the moment of prehension. In mixtures of soft leaves and stiff stems, the lack of upper incisors in cattle may have the advantage over horses. The term ‘comb-out strategy’ is proposed for this selective foraging by cattle. DM intake per bite force was highest in TF leaves (7·5 mg DM/N) grazed by horses and similar (3·7–4·6 mg DM/N) in PRG leaves grazed by both cattle and horses.

This group, which is concerned with the applications of mathematics to agricultural science, was formed in 1970 and has since met at approximately yearly intervals in London for one-day meetings. The thirty-fifth meeting of the group, chaired by Dr David Parsons of Silsoe Research Institute, was held in the Kohn Centre at the Royal Society, 6 Carlton House Terrace, London on Friday, 11 April 2003 when the following papers were read.