This paper describes the physiological basis and validation of a generic legume model as it applies to 4 species: chickpea (Cicer arietinum L.), mungbean (Vigna radiata (L.) Wilczek), peanut (Arachis hypogaeaL.), and lucerne (Medicago sativa L.). For each species, the key physiological parameters were derived from the literature and our own experimentation. The model was tested on an independent set of experiments, predominantly from the tropics and subtropics of Australia, varying in cultivar, sowing date, water regime (irrigated or dryland), row spacing, and plant population density. The model is an attempt to simulate crop growth and development with satisfactory comprehensiveness, without the necessity of defining a large number of parameters. A generic approach was adopted in recognition of the common underlying physiology and simulation approaches for many legume species. Simulation of grain yield explained 77, 81, and 70% of the variance (RMSD = 31, 98, and 46 g/m2) for mungbean (n = 40, observed mean = 123 g/m2), peanut (n = 30, 421 g/m2), and chickpea (n = 31, 196 g/m2), respectively. Biomass at maturity was simulated less accurately, explaining 64, 76, and 71% of the variance (RMSD = 134, 236, and 125 g/m2) for mungbean, peanut, and chickpea, respectively. RMSD for biomass in lucerne (n = 24) was 85 g/m2 with an R2 of 0.55. Simulation accuracy is similar to that achieved by single-crop models and suggests that the generic approach offers promise for simulating diverse legume species without loss of accuracy or physiological rigour.

Eleven genotypes of soyabean (Glycine max) of tropical, sub-tropical and temperate origin and one accession of G. soja were grown in six locations in Australia during 1986-88, and at one location in Australia and two in Taiwan during 1989-91. Dates of sowing were varied within and among locations so as to expose plants to as many as 32 environments of widely different diurnal temperature and daylength. Times from sowing to flowering (/) were recorded, from which rates of progress towards flowering (1//) were calculated. These derived data were then related to mean pre-flowering values of temperature (T) and photoperiod (P) using a threeplane linear model developed from controlled environment data. Among genotypes, mean values of/ varied between 24-49 d and between 134-291 d in the most-and least-inductive environments, respectively. These differences were associated with variations in P from about 11 to 16 h d~', in daily mean maximum temperatures from about 17° to 36°C, in daily mean minimum temperatures from about 5° to 25°C, and in T from about 11° to 30°C, that is, a very wide range of photothermal regimes. The relations of l / / t o T and P can be described in photoperiod-insensitive genotypes by a thermal plane defined by two constants, a and />, and additionally by a photothermal plane defined by three constants, a', V and c'', in the more numerous photoperiod-sensitive genotypes. If photoperiod-sensitive genotypes are grown in sufficiently long days then a third photoperiod and temperature-insensitive plane is exposed, defined by a constant, d'\ this plane indicates the maximum delay in flowering of which the Present addresses: §Queensland Department of Primary Industries, PO Box 591, Ayr, Queensland 4807, Australia; "11 Edward Crescent, Byford, Western Australia 6201. 254 R. J. SUMMERFIELD et al.genotype is capable. The constants a', b', c' and d' define the delay in flowering caused by photoperiod-sensitivity genes. The two intercepts between the three planes define, respectively, the critical photoperiod, P c , above which increase in daylength delays flowering, and the ceiling photoperiod, P ce , above which there is no further delay. The values of the six constants for any genotype can be estimated from observations of/in several natural environments. Comparisons between years in Australia and between Australia and Taiwan show that these genotypic constants can predict \/f, and so the time taken to flower, given data on latitude, sowing date and daily values of maximum and minimum air temperatures. This model is more accurate than an alternative logistic model; we also believe that all six constants in the three-plane rate model described here have biological meaning. They indicate separate genetic control of flowering responses to P and T and could form a rational basis for the genetic characterization and analysis of these responses in the soyabean germplasm. Pronostico del momento defloration II RESUMENSe cultivaron once genotipos de poroto de soja (Glyane max) de origen tropical, subtropical y te...

In recent years, many sorghum producers in the more marginal (<600 mm annual rainfall) cropping areas of Queensland and northern New South Wales have used skip row configurations in an attempt to improve yield reliability and reduce sorghum production risk. This paper describes modifications made to the APSIM sorghum module to account for the difference in water usage and light interception between alternative crop planting configurations, and then demonstrates how this new model can be used to quantify the long-term benefits of skip sorghum production. Detailed measurements of light interception and water extraction from sorghum crops grown in solid, single and double skip row configurations were collected from on-farm experiments in southern Qld and northern NSW. These measurements underpinned changes to the APSIM-Sorghum model so that it accounted for the elliptical water uptake pattern below the crop row and the reduced total light interception associated with skip row configurations. Long-term simulation runs using long-term weather files for locations near the experimental sites were used to determine the value of skip row sorghum production as a means of maintaining yield reliability. These simulations showed a trade-off between long-term average production (profitability) and annual yield reliability (risk of failure this year). Over the long term, the production of sorghum in a solid configuration produced a higher average yield compared with sorghum produced in a skip configuration. This difference in average yield is a result of the solid configuration having a higher yield potential compared with the skip configurations. Skip configurations limit the yield potential as a safeguard against crop failure. To achieve the higher average yield in the solid configuration the producer suffers some total failures. Skip configurations reduce the chance of total failure by capping the yield potential, which in turn reduces the long-term average yield. The decision on what row configuration to use should be made tactically and requires consideration of the starting soil water, the soil’s plant-available water capacity (PAWC), and the farm family’s current attitude to risk.

Eight genotypes of cultivated mung bean, black gram and rice bean (Vigna mungo, Vigna radiata ssp. radiata and Vigna umbellata, respectively) were sown at six sites in Australia on various dates in order to provide a range of photothermal environments. In addition, four accessions of the related wild species Vigna radiata ssp. sublobata were sown on five occasions. Times from sowing to first flowering (/) varied between environments from 34 to 317 d; pre-flowering temperature and photoperiod means ranged from 12.7° to 29.1°C and from 11.8 to 15.5 h d~'. No effect of photoperiod was detected on rate of progress towards first flowering (1 If) in four genotypes, but in each case a significant positive relation was detected between l//*and mean temperature. These simple thermal time relations did not differ significantly among these four genotypes; the common base temperature was 7.9°C. In two genotypes observations were well described by a thermal response plane when the mean photoperiod was less than 13 h d" 1 (p < 0.01) but photoperiods greater than 14 h d~' delayed flowering. In each of the remaining genotypes the observations were best described by photothermal planes, that is, \lf was modulated by temperature and photoperiod. Predictions from the models based on our data were in good agreement with the times to first flowering observed in three genotypes in an earlier controlled environment study. 6201. 32 R. H. ELLIS et al. Pronostico del liempo hasta la floracion. IV RESUMENSe sembraron ocho genotipos de Vigna mungo, Vigna radiata ssp. radiata y Vigna umbellata en distinta fecha en seis emplazamientos en Australia a fin de proporcionar una variedad de ambientes fototermicos. Ademas, en cinco ocasiones se sembraron cuatro accesiones de las especie salvaje relacionada Vigna radiata ssp. sublobata. Los tiempos desde la siembra hasta la primera floracion (/) variaron con los distintos ambientes, de 34 d a 317 d, las medias de temperatura previa a la floracion y fotopen'odo fueron de entre 12,7°C a29,l°C, y de 11,8a 15,5 hd~'. En cuatro genotipos no se detecto efecto del fotopen'odo sobre el fndice de avance hacia la primera floracion (1 If), pero en cada uno de los casos se detecto una significativa posicion positiva entre \lfy la temperatura media. Estas relaciones de tiempo termico simple no presentaron diferencias significativas entre estos cuatro genotipos: la temperatura de base comun fue de 7,9°C. En dos genotipos las observaciones estuvieron bien representadas por un piano de respuesta termicas cuando la temperatura media fue inferior a 13 h d~ (p < 0,01), pero fotopen'odos superiores a 14 h d~ retrasaron la floracion. En cada uno de los genotipos restantes las observaciones estuvieron mejor representadas por pianos fototermicos, o sea que 1//Tue modulada por la temperatura y el fotopen'odo. Los pronosticos de los modelos basados en nuestros datos se correspondieron con los tiempos hasta la primera floracion observados en tres genotipos durante un estudio anterior en ambiente controlado.

No abstract

Six genotypes of cowpea (Vigna unguiculata) of diverse origin were sown on various dates at five locations in Australia in order to provide a range of photothermal environments. Times from sowing to first flowering (/) ranged between 36 d and 145 d; pre-flowering temperature and photoperiod means varied from 15.7° to 29.2°C and from 11.6 to 15.5 h d~'. In five genotypes there was no effect of photoperiod on rate of progress towards flowering (\lf), but the relation between l//and mean temperature was always positive. The base temperatures (at which 1//= 0) varied between 8.1° and 10.4°C. The rankings of parameter estimates among four photoperiod-insensitive genotypes common to this study and earlier research in controlled environments were almost identical, and there was generally good agreement between field observations and predictions from controlled environments once hourly temperatures were used to describe the natural environment. When cowpea plants were exposed to temperatures below 3°C, flowering was delayed beyond expected values, presumably as a result of chilling damage. In one genotype, rate of progress towards flowering was affected by both temperature and photoperiod, and relations between l//and the photothermal environment were described by a two-plane linear model of similar form to that determined in an earlier controlled environment study. These latest findings support the utility of such linear models for the prediction of crop phenology in the field and for the genetic characterization of photothermal flowering response in annual crops. §Present address: Queensland Department of Primary Industries, PO Box 591, Ayr, Queensland 4807, Australia. 18 R. H. ELLIS et al. Pronostico del liempo hasta la floracion. Ill R E S U M E NSe sembraron seis genotipos de caupi (Vigna ungiculata) de diverso origen en distintas fechas en cinco emplazamientos en Australia, a fin de lograr una variedad de ambientes fototermicos. Los tiempos desde la siembra hasta la floracion (f) fueron de 36 d a 145 d, las medias de temperatura y fotoperiodo fueron de 15,7°C a 29,2°C, y de 11,6 a 15,5 h d~'. En cinco genotipos el fotoperiodo no tuvo efecto en el indice de avance hacia la floracion (1 /f), pero la relacion entre 1 /f y la temperatura media siempre fue positiva. Las temperaturas de base (con las cuales 1/f = 0) fueron de entre 8,1°C y 10,4°C. La distribution de las estimaciones de los parametros de cuatro genotipos insensibles al fotoperiodo en comiin entre este estudio e investigaciones anteriores en ambientes controlados fue casi identica, y en general existio una buena correspondencia entre las observaciones de campo y los pronosticos provenientes de ambientes controlados cuando se comenzo a usar temperatura/hora para describir el ambiente natural. Cuando las plantas de caupi fueron expuestas a temperaturas inferiores a los 3°C, la floracion se retraso mas de lo esperado, presumiblemente debido a los danos producidos por el frfo. En un genotipo, el indice de avance hacia la floracion fue afectado tanto por la tem...

Variation in time from sowing to flowering (/) was examined for 44 cultivars of soyabean, mungbean, black gram, ricebean, cowpea, chickpea, lentil and barley, when grown in up to 21 diverse environments obtained by making one or more sowings at each of six locations spanning tropical, sub-tropical and temperate climates in Australia. The utility of simple linear models, relating rate of development (1//) towards flowering to mean photoperiod and temperature prevailing between sowing and flowering, was evaluated. The models were highly efficient, explaining most (86.7%) of the variation observed across species, cultivars and environments. They were particularly efficient in describing responses where cultivars were relatively welladapted, in agronomic terms, and least efficient where cultivars were exposed to unfavourable temperature and, to a lesser extent, photoperiod. Opportunities for exploiting the models in applied crop improvement include their use in interpretation of G X E interaction, genotypic characterization and selection of parental genotypes, selection of test environments, designing screening procedures, and more efficiently matching genotypes to target environments. The main strengths of these linear, additive rate models in crop improvement are their wide applicability across species and genotypes, their relative simplicity, and the requirement for few genotype-specific response parameters. Their main weakness is their lack of precision in describing responses when plants are exposed to unfavourable photothermal extremes, albeit in circumstances that are sometimes unrealistic for cropping those particular genotypes.Predicciones del tiempo que ha de transcurrir hasla la flotation. VI. . 90 R. j . LAWN et al. RESUMENSe examinaron las variaciones en tiempo entre la siembra y la floracion de cuarenta y cuatro cultivos de sqja, vigna mungo, vigna radiata, arroz, caupi, garbanzos, lentejas y cebada cuando se cultivaban en 21 ambientes diferentes, obtenidos mediante la practica de una o mas siembras en cada una de las seis ubicaciones-que iban de tropicales y subtropicales a climas templados en Australia. Se evaluo la utilidad de modelos lineales simples, que relacionan la velocidad de desarrollo (1/f) hacia la floracion con el fotoperiodo y la temperatura predominante medios entre la siembra y la floracion. Los modelos fueron altamente eficaces, explicaron la mayon'a (86,7%) de las variaciones observadas en las distintas especies, cultivos y ambientes. Fueron particularmente eficaces para describir rcacciones donde los cultivos sc habian adaptado relativamente bien, en terminos agronomos, y menos eficaces donde los cultivos habian sido expuestos a temperaturas desfavorables y en menor medida, a fotoperiodos desfavorables. Las oportunidades de explotacion de los modelos en mejoras aplicadas de los cultivos incluyen su uso en la interpretation de la interaction G X A, la caracterizacion genotipica y la seleccion de genotipos padres, la seleccion de ambientes de prueba, el disefio de procedimientos de cri...

Four genotypes of'desi' and two of'kabuli' chickpea (Cicer arietinum) were sown at six locations in Australia on various dates between 1986 and 1988, giving 22 combinations of site and sowing date with diverse photothermal environments. Times from sowing to first flowering (/) varied from 30 to 162 d, mean pre-flowering temperatures from 10.8° to 29.3°C and mean photoperiods from 11.3 to 15.6 h d~'. There was no evidence that any observation had been obtained in photoperiods shorter than the ceiling photoperiod (P cc ) or longer than the critical photoperiod (P c ). This suggests that, in typical agricultural environments, chickpea crops experience photoperiods (P) which should satisfy the condition P ce < P < P c . In one 'kabuli' and two 'desi' genotypes, l//was influenced by both temperature and photoperiod. The coefficient of determination (/? 2 ) for a linear, additive rate of development model ranged from 0.74 to 0.80, with no significant difference in either temperature sensitivity or photoperiod sensitivity among these three genotypes. In the remaining three genotypes, no significant response to temperature was detected between 10.8° and 29.2°C, so rate of progress to flowering was influenced solely by photoperiod. There was no significant difference in the sensitivity of \ If to P among these three genotypes. The linear, additive rate model found here to be so satisfactory as a predictive tool for phenology, is also shown to have much wider general application. Prediction del liempo defloration RESUMENSe sembraron cuatro genotipos 'desi' y dos genotipos 'kabuli' de garbanzo (Cicer arietinum) en seis emplazamientos en Australia, en diversas fechas entre 1986 y 1988, obteniebdose asi 22