Agronomy Journal 95:82-89 (2003)
© 2003 American Society of Agronomy
SYMPOSIUM PAPERS
Improving Physiological Assumptions Of Simulation Models By Using Gene-Based Approaches
Gerrit Hoogenboom*,a and
Jeffrey W. Whiteb
a Dep. of Biol. and Agric. Eng., The Univ. of Georgia, Griffin, GA 30223-1797
b Centro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT, Int.), Apt. Postal 6-41, 06600, Mexico, D.F., Mexico
* Corresponding author (gerrit{at}griffin.peachnet.edu)
Received for publication May 1, 2001.
 |
ABSTRACT
|
|---|
Application of crop models to plant breeding has been limited, in part due to the restricted capabilities of models to accurately represent genetic differences and genotype-induced crop responses. A gene-based model, GeneGro, was developed to simulate the effects of seven genes on growth and developmental processes in common bean (Phaseolus vulgaris L.) and was published in 1996. The objective of this paper is to describe the improvements that were made in GeneGro to incorporate the effect of the Tip gene. Presence of the Tip gene in photoperiod-sensitive cultivars reduces the inhibitory effect of low temperature on photoperiod sensitivity of flowering. A mechanistic approach, further guided by information on two other genes affecting photoperiod response, i.e., Ppd and Hr, was used to incorporate the effect of the Tip gene. In the modified GeneGro, this inhibitory effect is reduced under cooler temperatures ranging from 15 to 20°C for the mean daily temperature. However, in the presence of Tip, no such reduction occurs. For the calibration data, GeneGro explained 75% of the variation in days to flower vs. 61% for the original model. For an extensive evaluation data set, the modification explained 72% of the variation in days to flower vs. 70% for the original version while days to maturity, seed yield, canopy dry mass, and harvest index showed no improvement. These results reflect two major constraints to effective use of gene-based approaches in crop modeling. The first is the lack of reliable characterizations of cultivars for genes used in the model. The second is the scarcity of data from conditions where phenotypic differences between the Tip and tip gene would be expected. It can be concluded that understanding the genetic control of quantitative traits can guide improvements to simulation models.
Abbreviations: DRPP, developmental rate of progress as influenced by photoperiod • TSPPR, temperature sensitivity of photoperiod response
|
INTRODUCTION
|
|---|
PROCESS-BASED crop simulation models are receiving increasing use in agricultural research. Many such models have been used to examine the effects of cultivar traits on growth or yield (e.g., Jones and Zur, 1984; Hoogenboom et al., 1988; Boote and Tollenaar, 1994). However, most models use genetic coefficients, which are only indirectly related to specific genes and are often difficult to determine empirically (Hunt et al., 1993; Pabico et al., 1999; Roman-Paoli et al., 2000; Welch et al., 2000b). Some advances have recently been made to estimate these coefficients based on crop performance trials (Piper et al., 1998; Irmak et al., 2000; Welch et al., 2000a; Mavromatis et al., 2001).
The gene-based simulation model GeneGro (White and Hoogenboom, 1996) was developed from the common bean model BEANGRO Version 1.01 (Hoogenboom et al., 1992, 1994) to simulate cultivar differences by representing effects of seven genes. In the model GeneGro, the genes Ppd and Hr affect photoperiod sensitivity; the gene Fd affects developmental rates before flowering; the gene Fin affects phenology and various traits related to growth habit; and the genes Ssz-1, Ssz-2, and Ssz-3 affect the coefficients for pod and seed growth (Table 1). The original GeneGro model was calibrated for 200 treatment combinations, including effects of planting date, planting density, and irrigation, representing 30 cultivars grown in 14 trials in Colombia, Guatemala, Mexico, and Florida. For this data set, the model accounted for 87% of the variation in days to flower, 85% of days to maturity, 60% of seed weight, and 30% of seed yield (White and Hoogenboom, 1996). The model GeneGro was also evaluated for 39 cultivars with an independent data set based on 14 field trials conducted in Canada, the USA, Mexico, and Colombia. This evaluation data set represented 224 treatment combinations similar to those used in the calibration data set. In this evaluation, GeneGro explained 75% of the variation in days to flower, 68% in days to maturity, and 39% in seed weight but only 10% of the variation in seed yield (Hoogenboom et al., 1997).
The original GeneGro model did not account for reported genetic variation in temperature response of common bean development, e.g., Laing et al. (1984), Wallace (1985), and White and Montes (1993), which is partially due to reduced photoperiod sensitivity at cooler temperatures (Wallace et al., 1991; Acosta-Gallegos and White, 1995). This temperature effect is caused at least in part to action of the gene Tip (White et al., 1996), which was not used in GeneGro Version 1 (White and Hoogenboom, 1996) but whose effect appears to explain major differences in phenology of photoperiod-sensitive cultivars of Andean vs. Mexican origin.
The objective of this paper is to compare the response of the gene-based simulation model GeneGro with and without the modification to incorporate the effect of the Tip gene on the phenology and growth of common bean. White and Hoogenboom (2003) identify different hierarchy levels of genetic detail that have been included in crop simulation models. Level 1 is a generic model with no reference to species; Level 2 is a species-specific model with no reference to genotypes; Level 3 is a model in which the genetic differences are represented by cultivar-specific parameters; Level 4 is a model in which genetic differences are represented by specific alleles, with gene action gene effects represented through linear effects on model parameters; and Level 5 is a model in which genetic differences are represented by genotypes, with gene action explicitly simulated based on knowledge of regulation of gene expression and effects of gene products. Based on this hierarchy of levels of genetic detail in crop models, the previous version of GeneGro (White and Hoogenboom, 1996; Hoogenboom et al., 1997) corresponds to Level 4, with cultivar differences represented through linear genetic effects. The availability of information on the action of the genes Ppd, Hr, and Tip on common bean growth and development presents an opportunity to express genetic effects in GeneGro in a more mechanistic manner, thus moving toward Level 5 of the hierarchy.
|
MATERIALS AND METHODS
|
|---|
The GeneGro model as described by White and Hoogenboom (1996) and modified by Hoogenboom et al. (1997) was used as the source for the modifications. An overall temperature effect on photoperiod response was included first, and then an effect of the Tip gene was added.
Overall Temperature Effect on Photoperiod Response
An effect of photoperiod in GeneGro is implemented by multiplying the base developmental rate by a modifier, DRPP (developmental rate of progress as influenced by photoperiod). The DRPP varies from 0 to 1 where a larger value implies more rapid development. In an important departure from most models of crop phenology, it is assumed in the model GeneGro that DRPP is a function of an inhibitory process. It is worth noting that the specification of an inhibitory system in no way implies the nature of the underlying processes of synthesis and catabolism of promoters or inhibitors. Thus, different systems of biochemical processes can be envisaged that would have the same functional response (Table 2) and that could be represented by a similar set of equations.
View this table:
[in this window]
[in a new window]
|
Table 2. Four functionally equivalent models for inhibitory processes controlling photoperiod sensitivity of flowering with two genes showing recessive epistasis.
|
|
The basic level of photoperiod sensitivity in the model GeneGro is set by a parameter, called THVAR, which depends on the genotypes Ppd and Hr. In GeneGro, this is implemented through the following equation:
 | [1] |
where PPD indicates the genotype for the gene Ppd and HR the genotype for the gene Hr. To represent an inhibitory effect of THVAR that varies with photoperiod, the developmental modifier DRPP is estimated as
 | [2] |
where THVAR is the level of photoperiod sensitivity (from 0 to 0.938) and PHOTPC is the relative photoperiod change (from 1 to 0) with respect to the critical short and long days (Fig. 1).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 1. Variation in the baseline photoperiod modifier (PHOTPC) as a function of daylength for a critical short day of 12.26 h and a critical long day of 16.88 h.
|
|
The temperature effect on photoperiod response was simulated as an additional modifier, TSPPR (temperature sensitivity of photoperiod response), which is multiplied by THVAR, so Eq. [2] becomes
 | [3] |
with TSPPR increasing from 0 to 1 with the daily minimum temperature (Fig. 2), reflecting the reported sensitivity of the photoperiod response to night temperatures (Evans, 1975). The cardinal points on the curve were based on the variation in photoperiod sensitivity observed at trials described by Acosta-Gallegos and White (1995) at Palmira and Popayan (Fig. 3).
The changes discussed in the previous section resulted in an overall reduction in the rates of development under long daylengths. An adjustment, therefore, had to be made in the equation that estimates the number of photothermal days from emergence to flowering, i.e., PHTHRS(5). Based on iterative changes using the full calibration data set, the relation was changed from
 | [4] |
where FIN and ERL are the effects of the Fin and Fd genes, respectively, to
 | [5] |
Effect of the Tip Gene
Because the Tip gene reduces the temperature sensitivity of the photoperiod response, the effect of Tip was incorporated with a conditional statement. The temperature effect was applied only if the genotype was tip tip.
 | [6] |
where TIP is the effect of the Tip gene, represented as Tip Tip = 1 and tip tip = 0.
Sensitivity Analysis
To examine the effects of introducing the temperature effects on photoperiod sensitivity, time to flower was simulated for a range of temperatures and latitudes using weather data at Palmira, Colombia (lat 3°29' N; elevation 960 m), as a base. This site was selected due to its near-isothermal climate (Table 3) and because the model was known to perform well for these environments (White et al., 1995). Temperatures were varied from 10°C below the mean temperature to 5°C above, and latitudes considered were 0, 15, 30, and 45° N. Three hypothetical cultivars with genotypes Hr Hr fd fd Fin Fin but varying genotypes of Ppd and Tip were compared. A planting date of 29 May was used to ensure that preflowering growth would correspond to the period of the longest daylengths in the year.
Comparison of Observed and Simulated Data
Data used to calibrate the model were the same as those used to develop GeneGro Version 1 (White and Hoogenboom, 1996). This data set included eight trials from five locations ranging from Canada to Colombia (Table 4) with 28 cultivars (Table 5), representing approximately 200 treatment combinations. To provide an independent evaluation, the data from 14 trials were used as described in Hoogenboom et al. (1997)(Table 4). These trials included 39 cultivars (Table 5) and various types of treatments, providing a total of 217 treatment combinations.
View this table:
[in this window]
[in a new window]
|
Table 5. Cultivars considered in this study, including origin, seed size, genotypes or traits assumed, and whether they were used to calibrate or evaluate the model GeneGro.
|
|
For cultivars where sufficient data were available, the genotypes of Tip were inferred from historical experimental data in which photoperiod response measured under normal and 18-h, artificially extended photoperiods were compared at Palmira and Popayan (lat 2°25' N; elevation 1850 m) in Colombia. These are two environments that have similar latitudes but different temperature regimes due to variation in elevation. Details of the basic evaluation methodology are outlined in White and Laing (1989) and Acosta-Gallegos and White (1995). In evaluations of approximately 4000 bean lines, White and Laing (1989) found that a value for relative sensitivity to photoperiod over 0.65 indicated full sensitivity, so this value was used as a cutoff for classifying lines as either temperature sensitive (genotype tip tip) or insensitive (Tip Tip). For lines where insufficient field data were available, genotypes were assigned based on the assumption that only fully sensitive genotypes (e.g., Ppd Ppd Hr Hr) from the Mexican highlands would be Tip Tip. The genotypes, including the eight genes, for the cultivars used either in the calibration or evaluation process are listed in Table 5.
|
RESULTS
|
|---|
We first present results of the sensitivity analyses of GeneGro Version 2 to illustrate the qualitative effect of making the degree of photoperiod sensitivity increase with higher night temperatures and then incorporating the effect of the Tip gene. For genotypes with Ppd Ppd Hr Hr Tip Tip, there was no response to temperature at the higher latitudes, and all genotypes demonstrated the same response to temperature, as would be expected (Fig. 4A). Genotypes with Ppd Ppd Hr Hr tip tip showed a U-shaped response to temperature at the higher latitudes, with an increase in the number of days to flower with an increase in latitude (Fig. 4B). However, at higher temperatures, the increased photoperiod sensitivity offset an overall increased rate of development (Fig. 4C). Introducing the temperature insensitivity conferred by Tip resulted in a photoperiod-induced delay in development across all temperature regimes.
For the calibration data set, introducing the overall temperature sensitivity of the photoperiod response decreased the accuracy of the simulations of the modified [GeneGro Version 2 without Tip (NT)] compared with the original model (GeneGro Version 1; Table 6). However, with cultivar differences in Tip included in the model (GeneGro Version 2 T), 14% more of the variation in time to flower was accounted for than with the original model (GeneGro Version 1; Table 6). For time to maturity, only a 5% improvement was found with inclusion of Tip. For other traits, the original model (GeneGro Version 1) performed as well as or better than the modified model (GeneGro Version 2 T; Table 6). Tip is assumed to modify only the sensitivity to photoperiod, thus affecting developmental rates and especially the number of days to flowering. Thus, most improvements in simulations are expected to occur in the developmental traits, rather than the growth and yield traits. The results presented in Table 5 support this expectation.
View this table:
[in this window]
[in a new window]
|
Table 6. Comparison of linear regression analysis for simulated physiological traits as a function of measured traits for calibration data sets (see Table 4) applied to GeneGro Version 1 (V1) and GeneGro Version 2 (V2), without (NT) and with (T) cultivar differences for the gene Tip.
|
|
For the evaluation data sets, a 2% improvement was found in predicting the number of days to flower with the inclusion of the Tip gene (Table 7). Root mean square error (RMSE) decreased for both the number of days to flower and the number of days to maturity. For the remaining traits, the three versions of GeneGro showed a similar accuracy (Table 7).
View this table:
[in this window]
[in a new window]
|
Table 7. Comparison of linear regression analysis for simulated physiological traits as a function of measured traits for evaluation data sets (see Table 4) applied to GeneGro Version 1 (V1) and to GeneGro Version 2 (V2), without (NT) and with (T) cultivar differences for the gene Tip.
|
|
A regression analysis was conducted that tested for an interaction between simulated values and locations to determine whether the models varied in performance across sites (Table 8). With this regression analysis, a better model performance is expressed by a larger single degree-of-freedom mean square for simulated values. If the model shows a similar response across locations, then the simulation by location term should be small. For days to flower, GeneGro Version 2 had a slightly higher mean square for the simulated values as well as the simulated by location value than the original GeneGro Version 1 model. This showed that the new GeneGro Version 2 model performed better than the original model and was also able to simulate a differential response to the environmental conditions of the evaluation data sets and especially the impact of temperature on photoperiod sensitivity. For seed yield, the original model performed slightly better than the modified model.
View this table:
[in this window]
[in a new window]
|
Table 8. Comparison of performance of GeneGro Versions 1 and 2 based on mean squares and F values for linear regression analysis for effects of location, cultivar, and simulation on measured values of days to flower (R2 = 0.69; P  0.01) and seed yield (R2 = 0.65; P 0.01) from the evaluation data (see Table 4).
|
|
|
DISCUSSION
|
|---|
The sensitivity analyses showed that the original GeneGro model could be easily extended to include responses supported by additional genetic and physiological information. However, inspection of the calibration and evaluation results indicated that the improvements were restricted only to the phenological data (Table 7 and 8). Most of the improvements were found with the calibration data set, rather than the evaluation data set. Arguably, this means that the changes in the model were no better than a curve-fitting exercise to one set of data. A broader perspective is that the results highlight the difficulties inherent in gene-based approaches and emphasize the need for further research.
Incorporating a temperature effect on photoperiod revealed a series of unknown responses that simply had to be guessed. Although the qualitative response is well documented (e.g., White and Laing, 1989; Wallace et al., 1991; White et al., 1996), previous studies did not distinguish between effects on photoperiod sensitivity per se and effects on critical daylengths. We assumed that the critical photoperiods for short day (CSDVAR) and long day (CLDVAR) were invariant and that the only effect was on the actual response to photoperiod above the critical short-day level. In addition, we used the minimum temperature to define this response although for many environments, the minimum temperature does not occur at night, but around sunrise. These areas of uncertainty are logical topics for additional research, demonstrating at the same time the application of a gene-based model as a tool to focus research priorities.
Characterization of cultivars is also problematic, despite the fact that our approach is trying to mitigate this problem. Although marker technologies offer the prospect of simplifying identification of genes related to physiological traits, e.g., Gu et al. (1998), we still lack ready sources of data on the genetic makeup of cultivars. Data availability was also a constraint. At high latitudes, germplasm are predominantly day-neutral or would be expected to carry the tip allele, ensuring that photoperiod sensitivity declines with cooling temperatures at the end of the season. At low latitudes, photoperiod effects are low due to short daylengths. To detect effects of the Tip gene, trials are needed from intermediate latitudes, e.g., 15 to 30°. For the evaluation data used in this study, only one trial fell within these limits, so additional experimental data might be needed to provide a more rigorous evaluation of the model.
Additional cultivar differences are known to exist but are not included in the model due to insufficient understanding of their genetic impact. For instance, common bean cultivars show large differences in root growth under water deficit (Sponchiado et al., 1989; White et al., 1990). It has also been found that some cultivar variations in leaf assimilation rates are associated with variation in leaf thickness caused by a genetic variation and not a response to environmental conditions (Sexton et al., 1994). The genes Dl-1 and Dl-2 can have dramatic effects on plant growth, causing severe reductions in plant height and leaf area (Shii et al., 1980), and may affect a response to soil pH (Bennett et al., 1992).
The task of accumulating sufficient genetic and physiological information to refine the model further may appear so daunting that the approach is impractical. However, we argue that its potential benefits far outweigh possible difficulties. Furthermore, the difficulties may be overestimated. The primary benefit of the gene-based approach to modeling crop growth and development is that it provides a mechanism for reducing uncertainty over cultivar differences in crop models. A second benefit is that it can provide a more objective basis for assigning research priorities and assessing progress in understanding of physiological genetics. As a result, improved understanding may be measured in terms of improved prediction. On a more applied level, the genetic characterization of the cultivars required by model GeneGro, such as shown in Table 5, may provide researchers with a more meaningful classification than maturity groups or gene pools for studies of physiological traits or yield.
To facilitate further the development of gene-based models, a well-coordinated effort is needed among modelers, agronomists, breeders, and geneticists. Management of field and genetic data needs to be integrated. At least in the early phases of model development, specific combinations of environments and genotypes may be required to ensure that specific genetic effects are readily observable. This may require producing sets of near-isogenic lines for key genes. Geneticists and breeders may find themselves working on traits such as seed size and phenology, which they would consider to be of low commercial importance or trivial as a selection trait with conventional breeding methods. However, due to the major effects of genes affecting such traits, it is essential to represent the genetic mechanisms correctly in gene-based crop simulation models.
|
REFERENCES
|
|---|
- Acosta-Gallegos, J.A., and J.W. White. 1995. Phenological plasticity as an adaptation by common bean to rainfed environments. Crop Sci. 35:199–204.
- Bennett, J.H., N.J. Chatterton, R.E. Wyse, W. Hartung, and W.D.S. Mok. 1992. High temperature stress in Phaseolus vulgaris seedlings: Genotype iron stress-response interactions. Plant Physiol. Biochem. 30:11–18.
- Boote, K.J., and M. Tollenaar. 1994. Modeling yield potential. p. 533–565. In K.J. Boote, J.M. Bennett, T.R. Sinclair, and G.M. Paulsen (ed.) Physiology and determination of crop yield. ASA, CSSA, and SSSA, Madison, WI.
- Coyne, D.P. 1970. Genetic control of a photoperiod-temperature response for time of flowering in beans (Phaseolus vulgaris L.). Crop Sci. 10:246–248.[Abstract/Free Full Text]
- Evans, L.T. 1975. Daylength and the flowering of plants. W.A. Benjamin, Menlo Park, CA.
- Gu, W.K., J.Q. Zhu, D.H. Wallace, S.P. Singh, and N.F. Weeden. 1998. Analysis of genes controlling photoperiod sensitivity in common bean using DNA markers. Euphytica 102:125–132.
- Hoogenboom, G., J.W. Jones, and K. J. Boote. 1992. Modeling growth, development and yield of grain legumes using SOYGRO, PNUTGRO, and BEANGRO: A review. Trans. ASAE 35:2043–2056.
- Hoogenboom, G., J.W. Jones, and J.W. White. 1988. Use of models in studies of drought. p. 192–230. In J.W. White, G. Hoogenboom, F. Ibarra, and S.P. Singh (ed.) Research on drought tolerance in common bean. Documento de Trabajo no. 41. CIAT, Cali, Colombia.
- Hoogenboom, G., J.W. White, J. Acosta-Gallegos, R.G. Gaudiel, and J.R. Myers. 1997. Evaluation of a crop simulation model that incorporates gene action. Agron. J. 89:613–620.[Abstract/Free Full Text]
- Hoogenboom, G., J.W. White, J.W. Jones, and K.J. Boote. 1994. BEANGRO, a process-oriented dry bean model with a versatile user interface. Agron. J. 86:182–190.[Abstract/Free Full Text]
- Hunt, L.A., S. Pararajasingham, J.W. Jones, G. Hoogenboom, D.T. Imamura, and R.M. Ogoshi. 1993. GENCALC: Software to facilitate the use of crop models for analyzing field experiments. Agron. J. 85:1090–1094.[Abstract/Free Full Text]
- Irmak, A., J.W. Jones, T. Mavromatis, S.M. Welch, K.J. Boote, and G.G. Wilkerson. 2000. Evaluating methods for simulating soybean cultivar responses using cross validation. Agron. J. 92:1140–1149.[Abstract/Free Full Text]
- Jones, J.W., and B. Zur. 1984. Simulation of possible adaptive mechanisms in crop subjected to water stress. Irrig. Sci. 5:251–264.
- Kornegay, J., J.W. White, J.R. Dominguez, G. Tejada, and C. Cajiao. 1993. Inheritance of photoperiod response in Andean and Mesoamerican common bean. Crop Sci. 33:977–984.[Abstract/Free Full Text]
- Laing, D.R., P.G. Jones, and J.H.C. Davis. 1984. Common bean (Phaseolus vulgaris L.). p. 305–351. In P.R. Goldsworty and N.M. Fisher (ed.) The physiology of tropical field crops. John Wiley & Sons, New York.
- Mavromatis, T., K.J. Boote, J.W. Jones, A. Irmak, D. Shinde, and G. Hoogenboom. 2001. Developing genetic coefficients for crop simulation models with data from crop performance trials. Crop Sci. 41:40–51.[Abstract/Free Full Text]
- Mendel, G. 1866. Experiments in plant hybridisation. (English translation in 1965, with commentary by R.A. Fisher.) Oliver and Boyd, Edinburgh, UK
- Pabico, J.P., G. Hoogenboom, and R.W. McClendon. 1999. Determination of cultivar coefficients of crop models using a genetic algorithm: A conceptual framework. Trans. ASAE 42:223–232.
- Piper, E.L., K.J. Boote, and J.W. Jones. 1998. Evaluating and improvements of crop models using regional cultivar trial data. Appl. Eng. Agric. 14:435–446.
- Roman-Paoli, E., S.M. Welch, and R.L. Vanderlip. 2000. Comparing genetic coefficient estimation methods using the CERES-Maize model. Agric. Syst. 65:29–41.
- Sexton, P.J., C.M. Peterson, K.J. Boote, and J.W. White. 1994. Early-season growth in relation to region of domestication, seed size, and leaf traits in common bean. Field Crops Res. 52:69–78.
- Shii, C.T., M.C. Mok, S.R. Temple, and D.W.S. Mok. 1980. Expression of developmental abnormalities in hybrids of Phaseolus vulgaris L. J. Hered. 71:218–222.[Free Full Text]
- Sponchiado, B.N., J.W. White, J.A. Castillo, and P.G. Jones. 1989. Root growth of four common bean cultivars in relation to drought tolerance in environments with contrasting soil types. Exp. Agric. 25:249–257.
- Vallejos, C.E., and C.D. Chase. 1991. Linkage between isozyme markers and a locus affecting seed size in Phaseolus vulgaris L. Theor. Appl. Genet. 81:413–419.
- Wallace, D.H. 1985. Physiological aspects of plant maturity, adaptation and yield. Plant Breed. Rev. 3:21–67.
- Wallace, D.H., P.A. Gniffke, P.N. Masaya, and R.W. Zobel. 1991. Photoperiod, temperature and genotype interaction effects on days and nodes required for flower of bean. J. Am. Soc. Hortic. Sci. 116:534–543.[Abstract/Free Full Text]
- Wallace, D.H., K.S. Yourstone, P.N. Masaya, and R.W. Zobel. 1993. Photoperiod gene control over partitioning between reproductive and vegetative growth. Theor. Appl. Genet. 86:6–16.[ISI]
- Welch, S.M., G. Wilkerson, K. Whiting, N. Sun, T. Vagts, G. Buol, and T. Mavromatis. 2000a. Estimating genetic coefficients for a soybean growth model. ASAE Paper 00–3043. ASAE, St. Joseph, MI.
- Welch, S.M., J. Zhang, N. Sun, and T. Yu Mak. 2000b. Efficient estimation of genetic coefficients for crop models [CD-ROM]. In W. Bowen et al. (ed.) Proc. SAAD-3 Symp., Universidad Nacional Agraria, La Molina (UNALM), Lima, Peru. 8–10 Nov. 1999. Int. Potato Cent., Lima, Peru.
- White, J.W., J.A. Castillo, and J. Ehleringer. 1990. Associations between productivity, root growth and carbon isotope discrimination in Phaseolus vulgaris under water deficit. Aust. J. Plant Physiol. 17:189–198.
- White, J.W., and G. Hoogenboom. 1996. Integrating effects of genes for physiological traits into crop growth models. Agron. J. 88:416–422.[Abstract/Free Full Text]
- White, J.W., and G. Hoogenboom. 2003. Gene-based approaches to crop simulation: Past experiences and future opportunities. Agron. J. 95:52–64 (this issue).[Abstract/Free Full Text]
- White, J.W., G. Hoogenboom, J.W. Jones, and K.J. Boote. 1995. Evaluation of the dry bean model BEANGRO V1.01 for crop production research in a tropical environment. Exp. Agric. 31:241–254.
- White, J.W., J. Kornegay, and C. Cajiao. 1996. Inheritance of temperature sensitivity of the photoperiod response in common bean (Phaseolus vulgaris L.). Euphytica 91:5–8.
- White, J.W., and D.R. Laing. 1989. Photoperiod response of flowering in diverse genotypes of common bean (Phaseolus vulgaris). Field Crops Res. 22:113–128.
- White, J.W., and C. Montes. 1993. The influence of temperature on seed germination in cultivars of common bean. J. Exp. Bot. 44:1795–1800.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
J. W. White, M. Herndl, L. A. Hunt, T. S. Payne, and G. Hoogenboom
Simulation-Based Analysis of Effects of Vrn and Ppd Loci on Flowering in Wheat
Crop Sci.,
March 19, 2008;
48(2):
678 - 687.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. D. Messina, J. W. Jones, K. J. Boote, and C. E. Vallejos
A Gene-Based Model to Simulate Soybean Development and Yield Responses to Environment
Crop Sci.,
January 24, 2006;
46(1):
456 - 466.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. W. White and G. Hoogenboom
Gene-Based Approaches to Crop Simulation: Past Experiences and Future Opportunities
Agron. J.,
January 1, 2003;
95(1):
52 - 64.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. M. Welch, J. L. Roe, and Z. Dong
A Genetic Neural Network Model of Flowering Time Control in Arabidopsis thaliana
Agron. J.,
January 1, 2003;
95(1):
71 - 81.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
