Source: UNIVERSITY OF PUERTO RICO AT MAYAGUEZ submitted to NRP
IMPLEMENTATION OF SIMULATION MODELS FOR PRODUCTIVITY PREDICTIONS IN THE CARIBBEAN
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
SPECIAL GRANT
Reporting Frequency
Annual
Accession No.
0204340
Grant No.
2005-34135-16063
Cumulative Award Amt.
(N/A)
Proposal No.
2005-04614
Multistate No.
(N/A)
Project Start Date
Sep 15, 2005
Project End Date
Sep 14, 2007
Grant Year
2005
Program Code
[AH]- (N/A)
Recipient Organization
UNIVERSITY OF PUERTO RICO AT MAYAGUEZ
P. O. BOX 9000
MAYAGUEZ,PR 00681
Performing Department
Crops and Agroenvironmental Sciences
Non Technical Summary
The use of simulation has become a standard procedure in many disciplinary areas, including agriculture. Simulation models allow researchers to analyze physiological and management aspects and to extrapolate results across different conditions. The CERES-Maize and CROPGRO models are being used to investigate if a region is appropriate for large-scale production of maize and beans, which are important crops in the Caribbean. The CERES-Maize and CROPGRO models released do not include genetic coefficients (parameters that describes the genotype x environment interaction) estimates for maize and bean cultivars recommended for the Caribbean. The correct genetic coefficients (GC) estimation requires field trials conducted with different management practices or different locations. In this study, GC for three open pollinated maize cultivars and a commercial genetic modified hybrid and a group of recombinant inbred bean lines were estimated at two locations in which long-term climatic data is available. Once all the data required by the models is available it can be used to predict yield and optimum management practices of maize and bean cultivars accurately in regions where both crops could be grown on a large scale. In west and southwestern Puerto Rico, the government declared two agriculture reserves. With the use of the simulation techniques researchers may simulate maize and bean yield and to identify the feasibility of the productions of those crops on agricultural reserves and other areas without involving in traditional research
Animal Health Component
25%
Research Effort Categories
Basic
75%
Applied
25%
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
2031410106020%
2031510106020%
2051410102030%
2051510102030%
Knowledge Area
205 - Plant Management Systems; 203 - Plant Biological Efficiency and Abiotic Stresses Affecting Plants;

Subject Of Investigation
1510 - Corn; 1410 - Beans (dry);

Field Of Science
1060 - Biology (whole systems); 1020 - Physiology;
Goals / Objectives
1. Perform a sensitivity analysis of the Beangro 4.0 (CROPGRO-bean) model. 2. Determine genetic coefficients for maize hybrids and bean cultivars in the Caribbean. 3. To use CERES-Maize and Beangro (CROPGRO-bean) models to predict yield response for different management practices of maize and beans in Puerto Rico.
Project Methods
The use of simulation has become a standard procedure in many disciplinary areas, including agriculture. Simulation models allow the researcher to analyze some physiological and management aspects without traditional experiments, and to extrapolate results across different conditions and locations with the use of appropriate inputs files (weather, soils, and genetic coefficients). The CERES-Maize and CROPGRO (DSSAT V4.0) models will be used to investigate if a region is appropriate for large-scale production of maize (Zea mays L.) and beans (Phaseolus vulgaris L.), which are important crops in the Caribbean. Various steps need to be accomplished for models to be suitable for accurate predictions. First, the researcher needs to know which inputs are more important to the model to produce a better prediction. Therefore, a sensitivity analysis of CROPGRO model was performed by systematic changes in management practices using local soil and weather data. Once the sensitivity analysis is done, researcher knows which input need particular attention when is been estimated. CERES-Maize and CROPGRO models released include genetic coefficients for several cultivars, but genetic coefficients (GC) for maize and bean cultivars recommended for Caribbean conditions require to be estimated. The correct GC estimation for new bean cultivars and maize hybrids require field trials conducted with different management practices. In this study, GC for the maize and bean cultivars were estimated at two locations in which long-term climatic data is available. Genetic coefficients for three open pollinated maize cultivars and a commercial genetic modified hybrid and a group of recombinant inbred bean lines (derived from a cross between ICA Pijao x Montcalm and Jamapa x Calima) were estimated in the field. Both crops were managed toward optimal conditions to guarantee that plants express it genetic potential. Four and seven GC were estimated for maize and bean, respectively. Phenological and growth stages of maize and bean were recorded and data used to calculated the corresponding GC. Once all the data required by the models is available it can be used to predict yield and optimum management practices of maize and bean cultivars accurately in regions where both crops could be grown on a large scale. In west and southwestern Puerto Rico, the government declared two agriculture reserves. With the use of the simulation techniques researchers may simulate maize and bean yield and to identify the feasibility of the productions of those crops on these reserves.

Progress 09/15/05 to 09/14/07

Outputs
Field experiments were conducted during a two-year period with the principal goal of utilizing the CERES-Maize and CROPGRO models as a tool to predict better management practices for producing maize and common bean (Phaseolus vulgaris L.) on a bigger scale in Puerto Rico. Duration of phenological stages and genetic coefficients were estimated for three locally adapted open pollinated maize cultivars: "Mayorbela improved", "Diente de Caballo", "Churo", and a genetically modified hybrid (GMO). Also, genetic coefficients, yield components, and other agronomic traits of two common bean populations 'ICA Pijao x Montcalm' and 'Jamapa x Calima' were estimated. Measured genetic coefficients for maize were similar to published values. 'Mayorbela improved' yield and overall performance was similar to that of GMO and higher than that of the other cultivars evaluated. The data measured under field conditions and published values of the three open pollinated maize cultivars were utilized to validate the CERES-Maize model. Even though phenology and yield components were not simulated accurately, yield predictions were fairly accurate for 'Mayorbela improved'. Narrow sense heritabilities were estimated for common bean seed yield components and phenological and agronomic traits used by CROPGRO model. Heritabilities for seed yield and mean number of pods per plant were intermediate for both populations. On the other hand, the heritabilities of mean number of seed per pod and mean seed weight were high. The magnitude of the heritabilities for determinate and indeterminate lines in the ICA Pijao x Montcalm was similar for that of the seed yield components. However, the heritability of emergence to flowering (EM-FL) of indeterminate lines of the ICA Pijao x Montcalm population was lower than the heritability of EM-FL of the determinate lines.

Impacts
Field estimation of genetic coefficients facilitates the use of simulation models to predict maize and bean better management practices for acceptable yields and to reduce production costs for the conditions prevailing in four valleys, reserved by a state law, to be used exclusively for agriculture. For beans, the magnitude of the heritabilities of all the traits studied is sufficiently large to permit selection in a bean breeding program. Moreover, there is sufficient genetic variability among the lines in both populations to possibly permit the identification of molecular markers (QTL) associated with the genetic coefficients used in the crop model.

Publications

  • Saliceti Rivera, M. 2005. Evaluacion de dos poblaciones de habichuela (Phaseolus vulgaris L.) para ser utilizada en el modelo de simulacion CROPGRO. MS thesis. Department of Agronomy and Soils. University of Puerto Rico.
  • Esteves Nieves, M., 2007. Determinacion de los coeficientes geneticos en maiz para ser utilizados en el modelo CERES-Maize. MS thesis. Department of Agronomy and Soils. University of Puerto Rico.


Progress 09/15/05 to 09/14/06

Outputs
A second field trial was established at Lajas substation to estimate genetic coefficient for three open pollinated maize cultivars, 'Mayorbela improved', 'Diente de Caballo', 'Churo', and a genetically modified hybrid for Bt planted on a Vertic Haplustert March 2, 2005. Maize fertilization, pest and disease control, weed control, drip irrigation were managed toward optimal conditions to guarantee the expression of genetic potential. Maize was planted at l9 cm between plants and 90 cm between rows. Growing Degree Days between phenological stages were calculated by using the standard method. The hybrid showed the earliest time to reach silking date (52 dap) and physiological maturity (90 dap) and the lowest leaf number, with 16. 'Mayorbela' and the 'Diente de Caballo' took the longest to reach maturity (95 dap). The hybrid produced the highest yield with 3,845 kg/ha. In both years, yields were lower than average commercial yields, however maize was planted at low plant population. It is important to point out that P1 estimates obtained in 2005 (cultivars stayed longer in the juvenile phase) were higher than those obtained in 2004. No cultivar reached a LAI of 1.5, meaning that they may not have reached full development. No significant differences were found among cultivars for total LAI and the interaction between time of sampling and cultivar was significant. No significant differences were found among maize cultivars for total dry matter accumulation non was there significant interaction between time of sampling and cultivars. During the upcoming months we plan to validate the CERES-Maize and CROPGRO-dry bean crop models. Also, there is a genetic relationship among the dry bean lines in the 'ICA Pijao x Montcalm' and the 'Jamapa x Calima' populations; we would like to use the data to determine the magnitude of the genetic contribution to the expression of traits in the crop model.

Impacts
Dr. Eduardo Vallejos has used a portion of the data collected for the CROPGRO-Dry Bean crop model for molecular marker and genetic mapping research at the University of Florida. Marilis Esteves Nieves, a graduate student supported by this project, scheduled for completing her of master's requirements by November 2006.

Publications

  • Saliceti, M., E. Roman-Paoli and J.S. Beaver. Repeatability of morphological & phenological traits used by the Cropgro-Dry Bean crop model. Paper presented at the Annual Meeting of the Bean Improvement Cooperative Program held in Newark, Delaware. 30 October to 3 November 2005
  • Esteves-Nieves, M., E. Roman-Paoli, and J. Beaver. Utilizacion de la simulacion como herramienta para el manejo a gran escala de habichuela y maiz. Poster presented at 52th PCCMCA Annual Meeting. April 24-28, 2006. Montelimar, Nicaragua
  • Saliceti Rivera, M. 2005. Evaluacion de dos poblaciones de habichuela (Phaseolus vulgaris L.) para ser utilizadas en el modelo de simulacion CROPGRO. Recinto Universitario de Mayaguez, Universidad de Puerto Rico. Master Thesis. http://grad.uprm.edu/tesis/salicetirivera.pdf.