Crop Development Rate Research Articles

Crop yield reduction in the tropics under climate change: Processes and uncertainties

Many modelling studies examine the impacts of climate change on crop yield, but few explore either the underlying bio-physical processes, or the uncertainty inherent in the parameterisation of crop growth and development. We used a perturbed-parameter crop modelling method together with a regional climate model (PRECIS) driven by the 2071–2100 SRES A2 emissions scenario in order to examine processes and uncertainties in yield simulation. Crop simulations used the groundnut (i.e. peanut; Arachis hypogaea L.) version of the General Large-Area Model for annual crops (GLAM). Two sets of GLAM simulations were carried out: control simulations and fixed-duration simulations, where the impact of mean temperature on crop development rate was removed. Model results were compared to sensitivity tests using two other crop models of differing levels of complexity: CROPGRO, and the groundnut model of Hammer et al. [Hammer, G.L., Sinclair, T.R., Boote, K.J., Wright, G.C., Meinke, H., and Bell, M.J., 1995, A peanut simulation model: I. Model development and testing. Agron. J. 87, 1085–1093]. GLAM simulations were particularly sensitive to two processes. First, elevated vapour pressure deficit (VPD) consistently reduced yield. The same result was seen in some simulations using both other crop models. Second, GLAM crop duration was longer, and yield greater, when the optimal temperature for the rate of development was exceeded. Yield increases were also seen in one other crop model. Overall, the models differed in their response to super-optimal temperatures, and that difference increased with mean temperature; percentage changes in yield between current and future climates were as diverse as − 50 % and over + 30 % for the same input data. The first process has been observed in many crop experiments, whilst the second has not. Thus, we conclude that there is a need for: (i) more process-based modelling studies of the impact of VPD on assimilation, and (ii) more experimental studies at super-optimal temperatures. Using the GLAM results, central values and uncertainty ranges were projected for mean 2071–2100 crop yields in India. In the fixed-duration simulations, ensemble mean yields mostly rose by 10–30%. The full ensemble range was greater than this mean change (20–60% over most of India). In the control simulations, yield stimulation by elevated CO 2 was more than offset by other processes – principally accelerated crop development rates at elevated, but sub-optimal, mean temperatures. Hence, the quantification of uncertainty can facilitate relatively robust indications of the likely sign of crop yield changes in future climates.

DEVELOPMENT OF DOUBLE HAPLOIDY IN UMBELLIFERAE

The umbelliferae species includes vegetables (e.g. carrot, celery, parsnip) as well as herb/spice crops (e.g. dill, caraway, anise). When compared to other major crop species, there has been very little breeding or genetic improvement of the umbelliferae, especially the herb/spice crops. Therefore, there is a potential for variability in crop performance and content of active ingredients. Haploidy technology can be used to develop uniform, true breeding lines as well as to accelerate breeding programs. Screening the umbelliferae species for microspore culture response has yielded embryos from anise, angelica, caraway, carrot, dill, lovage, fennel, and parsnip. Haploid and double haploid (DH) plants have also been regenerated from anise, caraway, carrot, dill, lovage, fennel, and parsnip. Field trials of the DH lines were conducted in 2003 and 2004 for anise, dill, and fennel. These trials showed that there were significant differences between the DH lines and their corresponding parental lines in terms of seed viability, plant height, rate of crop development, seed yield, and oil quantity.

Development and assessment of a coupled crop–climate model

AbstractIt is well established that crop production is inherently vulnerable to variations in the weather and climate. More recently the influence of vegetation on the state of the atmosphere has been recognized. The seasonal growth of crops can influence the atmosphere and have local impacts on the weather, which in turn affects the rate of seasonal crop growth and development. Considering the coupled nature of the crop–climate system, and the fact that a significant proportion of land is devoted to the cultivation of crops, important interactions may be missed when studying crops and the climate system in isolation, particularly in the context of land use and climate change.To represent the two‐way interactions between seasonal crop growth and atmospheric variability, we integrate a crop model developed specifically to operate at large spatial scales (General Large Area Model for annual crops) into the land surface component of a global climate model (GCM; HadAM3). In the new coupled crop–climate model, the simulated environment (atmosphere and soil states) influences growth and development of the crop, while simultaneously the temporal variations in crop leaf area and height across its growing season alter the characteristics of the land surface that are important determinants of surface fluxes of heat and moisture, as well as other aspects of the land‐surface hydrological cycle. The coupled model realistically simulates the seasonal growth of a summer annual crop in response to the GCM's simulated weather and climate. The model also reproduces the observed relationship between seasonal rainfall and crop yield. The integration of a large‐scale single crop model into a GCM, as described here, represents a first step towards the development of fully coupled crop and climate models. Future development priorities and challenges related to coupling crop and climate models are discussed.

Cotton growth and yield after seed treatment with mepiquat chloride in the tropical winter season

Cottonseed ( Gossypium hirsutum L.) was treated with mepiquat chloride (MC) to examine its potential for the management of vigorous early growth experienced on cotton grown in the tropical winter season. As there is little published research from the field on this topic, we aimed to determine the effect of seed treatment of cotton with MC grown to maturity. In two field experiments over two seasons the seed surface was treated. Seed surface treatment was compared with pre-sowing priming with MC in experiment 2. Lint yields were high, 2059 and 2270 kg/ha averaged over all treatments in experiments 1 and 2. In both experiments, seed surface treatment with 2 g MC/kg reduced ( P < 0.05) plant height for up to 30 days after sowing and there were no significant changes to rate of crop development, time to maturity, lint yield or fibre quality. In experiment 2 treating with MC reduced plant establishment in proportion to the amount of MC applied ( P < 0.01). Pre-sowing priming caused approximately twice the reduction in plant height per gram of MC than surface treatment and significant differences persisted longer, even until maturity using 8 g MC/kg seed. Priming with ≥4 g MC/kg seed had additional effects ( P < 0.05): (1) reduced node development between first square and first flower, (2) delayed the time to first square, first flower and maturity by 4 days, (3) increased production of bolls on monopodial branches, and (4) reduced lint yield compared to the same concentrations applied to the surface. Fewer smaller plants having smaller bolls were implicated in this response. It was concluded that the relative benefit of seed treatment with MC at the concentrations and methods that did not reduce lint yield appears questionable as height reductions per gram of MC/ha were less than measured at this location with the same amount applied as a foliar spray.

A generalized nonlinear tempeature response function for some growth and developmental parameters in kiwifruit (Actinidia deliciosa (A. Chev.) C. F. Liang &amp; A. R. Ferguson)

Temperature is a major factor that affects metabolic processes in living organisms. Thermal time has been widely used to account for the effects of temperature on crop growth and development. However, the thermal time approach has been criticized because it assumes a linear relationship between the rate of crop growth or development and temperature. The response of the rate of crop growth and development to temperature is nonlinear. The objective of this study was to develop a generalized nonlinear temperature response function for some growth and developmental parameters in kiwifruit (Actinidia deliciosa (A. Chev.) C. F. Liang &amp; A. R. Ferguson). The nonlinear function has three coefficients (the cardinal temperatures), which were 0ºC, 25ºC, and 40ºC. Data of temperature response of relative growth rate, relative leaf area growth, net photosynthesis rate, and leaf appearance rate in kiwifruit (female cv. Hayward) at two light levels, which are from published research, were used as independent data for evaluating the performance of the nonlinear and the thermal time functions. The results showed that the generalized nonlinear response function is better than the thermal time approach, and the temperature response of several growth and developmental parameters in kiwifruit can be described with the same response function.

Yields and Economics of High Tunnels for Production of Warm-season Vegetable Crops

Crop development rates, yields and production economics for muskmelon (Cucumis melo), pepper (Capsicum annuum) and tomato (Lycopersicon lycopersicum) grown in high tunnels [4.3 m wid × 2.5 m high × 29 m long (14 × 8 × 96 ft)] were compared to standard low tunnels over several cropping seasons in a temperate production area. The polyethylene-covered high tunnels protect several rows of crop for the duration of the cropping season. Air temperatures in the high tunnels were controlled by raising the sides of the tunnel. Low tunnels cover only a single row and must be removed soon after the crop is established to prevent overcrowding or overheating. When the low tunnels were in place, rates of accumulation of growing-degree days (GDDs) and early crop growth were comparable in the two tunnel systems. However, once the low tunnels were removed, the accumulation of GDDs in the high tunnels exceeded the standard system. The crops in the high tunnels matured 1 to 2 weeks earlier and produced substantially greater fruit yields before frost than in the low tunnel treatments. The high tunnels provided little frost protection and were of limited utility for extension of the growing season. The high tunnels were much more costly to purchase and construct than the low tunnels but were durable enough to be used for multiple cropping seasons. Based on wholesale commodity prices, it would take 2 to 5 years for the enhanced gross returns obtained with the high tunnels to cover their higher capital costs.

Efeito da radiação solar e temperatura na definição do número de grãos em milho

O objetivo deste trabalho foi estudar o efeito da radiação solar e da temperatura entre a emergência e o pendoamento na taxa de crescimento da cultura, no número de grãos por unidade de taxa de crescimento, e nas relações dessas variáveis com o rendimento final de grãos de híbridos de milho. Os ensaios foram conduzidos nos anos agrícolas de 1994/95, 1995/96 e 1996/97 com os híbridos comerciais C-901, XL-560 e XL-678, em 1994/95, e os híbridos C-901, XL-212 e XL-370, nos outros anos. Os tratamentos consistiram em diferentes datas de plantio de setembro a dezembro, em 1994/95, e de agosto a dezembro, em 1995/96 e 1996/97. No período entre emergência e espigamento, as plantas acumularam maior quantidade de massa seca quando a radiação solar incidente foi mais elevada. No entanto, como a temperatura média do ar exerce efeito na duração do período, a taxa de crescimento foi mais dependente da temperatura do que da radiação. Também neste período, o efeito da temperatura foi inversamente proporcional ao do número de grãos por unidade de taxa de crescimento, possivelmente por causa do maior tempo disponível para interceptar a radiação solar. Decorrente da associação entre temperatura e radiação, o coeficiente fototérmico foi positivamente associado ao rendimento de grãos.

Effect of varying seeding date on crop development, yield and yield components in canaryseed

The effects of varying seeding date on crop development, yield and yield components in canaryseed (Phalaris canariensis L.) have not been previously reported. In 1996 and 1997, a seeding date study was conducted at Swift Current, SK, which included barley (Hordeum vulgare L.), canaryseed and wheat (Triticum aestivum L.) sown at three dates in separate tilled fallow and untilled wheat stubble sites. Terminal summer drought occurred in both years of this experiment. Cumulative degree days (DD0) to reach maturity did not differ significantly among seeding dates for barley, or for wheat in 1997, while cumulative degree days to reach maturity decreased by 60 DD0 with delayed seeding for wheat in 1996. In contrast, cumulative degree days to reach maturity in canaryseed increased by 70 DD0 in 1996 and by 90 DD0 in 1997 with delayed seeding. Delaying seeding from the early to the late date decreased canaryseed yield by 29%, while barley and wheat yields decreased only 14 and 11%, respectively. Panicle density in canaryseed was reduced 24% between the early and late seeding dates, while barley and wheat spike densities were reduced only 2 and 6%, respectively. The large yield reduction in canaryseed was likely due to slowed crop development with delayed seeding, which intensified late-season drought stress. The slowed crop development with delay in seeding date in canaryseed may be due to vernalization requirement in this crop. In the semiarid prairie region, canaryseed should be seeded early to maintain a rapid crop development rate to minimize yield loss due to drought stress. Key words: Canaryseed, Phalaris canariensis L., seeding date, drought stress

The role of ecophysiological models in QTL analysis: the example of specific leaf area in barley

Crop modelling has so far contributed little to the genetic analysis of a quantitative trait. This study illustrates how a simple model for crop phenological development, which assumes that crop development rate is affected by daily effective temperature, can assist the identification of Quantitative Trait Loci (QTLs), using specific leaf area (SLA) in barley as an example. The SLA was measured in a field experiment six times during the growing season of 94 recombinant inbred lines (RILs) derived from a cross between cultivars Prisma and Apex. Of the six measurements, one was conducted at the same physiological age for all RILs (at flowering), four were undertaken at specific chronological days prior to flowering, and the last one was taken at 14 days after flowering. When the measured SLA was directly used as the quantitative trait, one to three QTLs were detected for SLA at each measurement time. The major dwarfing gene denso segregating in the population was found to affect SLA strongly at all measurement times except at flowering. If SLA of the different RILs was corrected for differences in physiological age at the time of measurement, by the use of the crop development model, QTLs were detected for SLA at only three stages. Furthermore, the effect of the denso gene was no longer significant during the preflowering stages. The effect of the denso gene detected in the first instance was therefore the consequence of its direct effect on the duration of the preflowering period. This demonstrates the important role that crop development models can play in QTL analysis of a trait that varies with developmental stage. Potential uses of ecophysiological crop growth models in QTL analysis are briefly discussed.

A Risk Management Perspective on Integrated Weed Management

Abstract The variability inherent in agriculture influences many crop production decisions made by farmers, including weed management. This paper addresses how farmers perceive the variability, or risk, associated with integrated weed management systems in terms of yield, economic returns, and time and labor management. This paper addresses how key biological time constraints such as periodicity of weed emergence, rate of crop growth and development, and critical periods of weed control can influence the outcome of integrated weed management systems. A key component to developing successful integrated weed management systems lies in the ability of the crop producer to align individual time and labor management issues with existing biological time constraints.

Rice flowering in response to diurnal temperature amplitude

Temperature is an important determinant of crop development rate (DR). Some reports have shown that the response of DR to temperature is nonlinear over a wide range. Others indicate that DR is affected by the amplitude of the diurnal temperature fluctuation. However, the nonlinearity between DR and temperature and the temperature amplitude effect are often confounded, as the amplitude effect can be explained by the nonlinearity. This paper investigates the effect of temperature amplitude per se on development to flowering in rice ( Oryza sativa L.), using controlled-environment data for 16 cultivars. The response of DR to diurnally constant temperatures was quantified by a nonlinear model and subsequently used to predict days to flowering under diurnally fluctuating temperatures. The often large discrepancies between predicted and observed days indicated an effect of the diurnal temperature amplitude on DR. This effect was significant at different probability levels in 11 cultivars, but not in the other five P > 0.10). The trend of this effect, where significant, also strongly differed among cultivars: temperature amplitude accelerated flowering in one cultivar but delayed flowering in others. The nonlinearity between DR and temperature and the effect of the temperature amplitude can explain a large part of the often observed variation among environments for thermal unit requirements for flowering.

A simulation study of crop growth and development under climate change

Climate changes of the order predicted by Global Circulation Models have important implications for arable crop production. We have studied the impact in Scotland using simulation models for three crops of contrasting developmental type: faba or field bean, potato, spring and winter wheat. The models used were the FABEAN, SCRI water-constrained potato model and AFR-CWHEAT2 models respectively. Consideration has been made of the natural year-to-year variation in weather which causes yield variability by using 100 years of input weather data produced by a weather generator. The models were run for four Scottish sites and five Scottish soils. Based on GCM predictions, we used eight scenarios of future climate which combine both temperature and rainfall changes. Current temperature ( T 0) and rainfall ( R 0) were used as a baseline, and each of T 0 + 1°C, T 0 + 2°C, T 0 + 3°C were used with rainfall unchanged at R 0, and increased by seasonally adjusted amounts ranging from 0 to 1.5 mm per wet day. Possible enhancements due to CO 2 fertilisation were not included in the study. Increased temperatures increase crop development rate, which shortens the growing season for wheat and faba bean, but, given a fixed harvest date, lengthens the season for potatoes. Yields of potato increased by up to 33% over all our sites and scenarios, whereas wheat yields decreased by 5–15% and faba bean by 11–41%. Rainfall increases of the amount suggested here do not affect the yield of potatoes or spring wheat, but winter wheat yields are reduced, due to leaching, and faba bean yields increase through alleviation of water shortage. Faba beans also show a reduction in yield variability as a result of increased rainfall. Changes in variability in wheat and potato were less pronounced and tended to reflect the increase in variability which was assumed to accompany the increased rainfall. Predictions for the changes in the frequencies of high and low yields are also presented. The results give an indication of the level of changes in crop production which would be expected in these future climates.

The application of agricultural land rating and crop models to CO2 and climate change issues in Northern regions: the Mackenzie Basin case study

The Mackenzie Basin in northwestern Canada covers approximately 1.8 million km2 and extends from 52°N to 70°N. Much of the Basin is currently too cool and remote from markets to support a viable agricultural sector, but the southern portion of the Basin has the physical potential to support commercial agriculture. This case study employed agricultural land rating and crop models to estimate the degree to which a CO2 -induced global warming might alter the physical potential for commercial agriculture throughout the Basin. The two climate change scenarios considered in this analysis would relax the current constraints imposed by a short and cool frost-free season, but without adaptive measures, drier conditions and accelerated crop development rates were estimated to offset potential gains stemming from elevated CO2 levels and warmer temperatures. In addition to striving for a better understanding of the extent to which physical constraints on agriculture might be modified by climate change, there is a need to expand the research context and to consider the capacity of agriculture to adapt to altered climates.

A nonlinear model for crop development as a function of temperature

The Beta function, commonly used as a skewed probability density function in statistics, was introduced to describe the effect of temperature on the rate of crop development. The framework is set by three cardinal temperatures, namely the base ( T b), the optimum ( T o) and the ceiling ( T c) temperature. The model parameters T b and T c and three other coefficients μ, α and β can be used to derive the value of T o and the maximum development rate. Parameter α also characterizes the curvature of the relationship with temperatures between T b and T o, and parameter β describes the curvature between T o and T c. The model has one parameter less than the Rice Clock Model (RCM); and in contrast to the RCM, it ensures that the maximum development rate occurs exactly at T o. The model accurately described the response to temperature of several developmental processes, and was superior to two widely used thermal time approaches in predicting rice flowering time.

Growth, yield and composition of four winter cereals. I. Biomass, grain yield and yield formation

A field experiment with 3 cultivars of each of 4 winter cereals (wheat, rye, triticale and barley), sown at about 320 plants/msuperscript 2, was conducted on a fertile clay soil in the central Netherlands. The N fertilizer was split-dressed: 120 kg/ha in total for wheat and triticale, and 60 kg/ha for rye and barley. The fewest shoots/msuperscript 2 were found in triticale (828/msuperscript 2), and the most in barley (1477/msuperscript 2). The average decrease in number of shoots during shoot/ear development was 51% in wheat, 54% in rye, 49% in triticale and 67% in barley. The rate of crop development was largest in rye and barley; they flowered and matured earlier than wheat and triticale. Rye was the first to attain a closed canopy (LAI &gt;3), and had the lowest maximum LAI and shortest leaf area duration. Wheat and triticale stayed green longest. Average specific leaf weight was 4.3 mg/cmsuperscript 2 in wheat and triticale, and 3.7 mg/cmsuperscript 2 in rye and barley. The growth rate of the grains at the linear stage was fastest in barley (1.89 mg grain-1 d-1) and slowest in rye (0.89 mg grain-1 d-1). Total aboveground DM production was 18 790 kg/ha in wheat, 15 230 kg/ha in rye, 18 300 kg/ha in triticale and 12 460 kg/ha in barley. Grain yields (t/ha) were 8.74 (wheat), 6.64 (rye), 8.20 (triticale) and 6.62 (barley). Cultivar differences in grain yield and in yield components were mostly smaller in wheat and rye than in triticale and barley. The harvest index was highest in barley (53.3%) and lowest in rye (43.7%); in some plant species there were marked differences between cultivars, but in others there were not.

Agronomic Performance of Winter versus Spring Wheat

AbstractRecent advances in crop management have enabled the production of winter wheat (Triticum uestivum L.) in most of the Canadian prairies. The agronomic performance of dominant hard red winter and highq‐uality hard red spring wheat cultivars was compared in semiarid and dry subhumid regions of Saskatchewan. In 15 trials, winter wheat (WW) outyielded spring wheat (SW) by an average 36%, and in three trials, ‘Norstar’ WW yielded 26% higher than the high‐yielding semidwarf SW, ‘HY 320.’ Protein yield was higher for WW in two of 15 trials and protein concentration was always higher for SW. Higher grain yield of WW was attributed mainly to the production of a higher kernel number per square meter (KNO). Crop development rate, aerial dry matter production, evapotranspiration (ET), and water use efficiency (WUE), were measured for Norstar WW and ‘Katepwa’ SW in six trials grown between 1986 and 1988. The period between Zadoks growth stage (ZGS) 21 and 65 was 4 to 14 d longer for Norstar than for Katepwa. Average daily air temperature between ZGS 21 and 65 was 2.6°C higher for Katepwa, and daily air temperature during this period was negatively correlated with KNO in both cultivars. Early season dry matter production was highest for Norstar. These early season cultivar differences gradually disappeared and, with the exception of two trials, total dry matter was similar by crop maturity. Harvest index was significantly higher for Norstar in five of six trials. Seasonal ET patterns were similar for Norstar and Katepwa, except in three of six trials where ET in early May was significantly higher for the WW cultivar. The WUE for dry matter production, grain yield and grain protein yield was consistently higher for Norstar.

Vernalization and Phyllochron in Winter Wheat

AbstractPhyllochron (degree‐days/leaf) has been used to describe phasic leaf development in dynamic wheat (Triticum aestivum L.) crop development models. Vernalization controls the rate of crop development in winter wheat. The objective of this experiment was to determine whether vernalization affected the thermal time required for winter wheat to produce leaves. ‘Stephens’ winter wheat plants were grown in a controlled environment chamber and in outdoor pots, using both pre‐vernalized and non‐vernalized germinating seed. In the outdoor pots plants became vernalized naturally and vernalization before sowing had no effect on growth, development, or flowering. In the controlled environment, up to the time of flowering of the vernalized treatment, there were no differences in phyllochron or total leaf number per main stem between the plants seeded using pre‐vernalized and non‐vernalized germinating seed. The non‐vernalized plants continued to produce tillers until the experiment was terminated, however, while tillering ceased at stem elongation in the vernalized plants. These results support the use of a constant value for phyllochron in simulating leaf development in both vernalized and non‐vernalized seedling winter wheat plants. However, they also show the need to adequately model the effect of vernalization on development (i.e., the cessation of leaf production and tillering in conjunction with flowering) in order to simulate the development of the crop leaf canopy.

Phenological Disruption and Economic Consequence of Injury to Alfalfa Induced by Potato Leafhopper (Homoptera: Cicadellidae)

A field study to characterize the effect of potato leafhopper, Empoasca fabae (Harris), feeding on alfalfa developmental physiology was begun in 1984 (one trial) and continued in 1985 (two trials). The rate of crop development was significantly reduced, with heavily infested plots (200 potato leafhoppers/m2) maturing about 30% later than uninfested plots. This phenological delay resulted in a concomitant reduction in daily accumulation of dry matter, crude protein, and digestible energy per hectare, but much of the loss was compensated for with additional regrowth time. To account for the disparity in maturity among plots, estimated nutrient yields at the predicted date of first bloom were compared with calendar date nutrient yields (all plots harvested when uninfested plots reach first bloom). These comparisons indicate that the economic consequence of potato leafhopper feeding was more severe with alfalfa managed on a calendar-date basis because time for compensatory regrowth was not available. Values for injury per insect were estimated from linear equations that predicted crop phonological delay based on potato leafhopper density. Economic losses associated with delays were based on slopes of nutrient yield accumulations per day of regrowth. Separate economic injury levels were calculated for potato leafhopper based on losses in digestible energy, crude protein, and dry matter.

Validation of a simple potato growth model in the North Central United States

A simple potato growth model was validated in 1985 and 1986 at two dryland sites, Grand Forks, ND and Rosemount, MN, and one irrigated site, Hancock, WI. At each site, Russet Burbank potatoes were grown to maximize yield. Leaf area index and leaf, stem, root, and tuber biomass data were collected from replicated experiments every two weeks. Soil matric potentials, minimum and maximum temperatures, and solar radiation also were measured. Among sites, crop development rates and solar radiation receipts were similar, but final tuber yields differed by as much as 35 Mg/ha. Model predictions of leaf, stem, and tuber dry weights were similar to observed data both in terms of maximum values attained and seasonal development except for the experiment in Grand Forks in 1986 where the model over-estimated final tuber yield by 23 Mg/ha (115%). Soil aeration or salinity, factors not currently modeled, may account for this discrepancy. Modeled yield responses to defoliation were compared with field data from one of the experimental locations. At levels of defoliation below fifty percent, proportional yield reductions were similar between observed and modeled crops.

Growth, water use and yield of barley in Mediterranean-type environments

SummaryRelations between yield, water use and pre-anthesis growth were analysed for cropsof barley grown for three seasons at several sites in northern Syria. The relations obtained were compared with those for other cereal crops grown in similar regions of Mediterranean climates.Phosphorus fertilizer application increased the rate of crop development from emergence to floral initiation and advanced anthesis by up to 11 days. Grain and total shoot dry-matter yields were increased by fertilizer (nitrogen + phosphorus) applications at all sites in all years, in most cases without increasing total evapotranspiration. The increased dry matter at anthesis was produced without having used a larger proportion of the total evapotranspiration in the whole season. Consequently, the ratio of grain yield to total above-ground dry-matter yield (harvest index) and kernel weight were also relatively stable between sites and years, despite some very low amounts of post-anthesis water use. Grain yield appeared to be largely determined by anthesis and there were strong linear relationships between grain yield or total dry-matter yield and number of kernels. Differences in water use efficiency of crops given fertilizer between sites and years were closely related to the differences in amounts of winter growth.Some responses differed from those predicted from models of growth, water use and yield developed in other regions with similar climates. It is concluded that agronomists and breeders should increase amounts of early growth thereby increasing grain and dry-matter yields. Future research emphasis should also be on the development of dynamic simulation models of pre-anthesis growth and water use.