Browsing by Subject "Plant breeding"

Now showing 1 - 10 of 10

Application of Near-Infrared Spectroscopy in Plant Breeding Programs
(2006) Montes, Juan Manuel; Melchinger, Albrecht E.
The success of plant breeding programs depends on the availability of genetic variation and efficient data collection processes that allow large-scale screenings of genotypes. When genetic variation is present, the goal is to identify those genotypes that are closest to the breeding objectives. In this context, the evaluation of a large number of genotypes requires optimization of the data collection process in order to provide reliable information for making selection decisions. The process of data collection must yield an accurate and precise assessment of genotypes timely because the information is needed to plan the next generation for breeding and cultivar development. Laboratory NIRS is routinely used in the data collection process of many breeding programs, but it requires the withdrawal of field plot samples and involves manual work. Applications of the near-infrared spectroscopy on choppers (NOC) and near-infrared spectroscopy on combine harvester (NOCH) are a step forward to the automation of data collection processes, by which sampling, labor, and sources of error in the data can be reduced. The objective of this thesis research was to assess the potential of NOC and NOCH for application in breeding programs of grain maize, rapeseed, and silage maize. Plot combine harvesters and choppers were equipped with diode-array spectrometers for collection of near-infrared plot spectra, and used to harvest experimental varieties of breeding programs in Central Europe. Two alternative sample presentation designs (conveyor belt and spout) were used for the NOC systems. The NOCH systems used the conveyor belt as sample presentation design. NOCH showed a high potential for determination of dry matter (DM), crude protein (CP), and starch (ST) contents of maize grain. NOCH calibration models yielded standard errors of prediction (SEP) and coefficients of determination of validation (R2V) of 1.2% and 0.95 for DM, 0.3% and 0.88 for CP, and 1.0% and 0.79 for ST, respectively. The potential of NOCH for determination of DM, CP, oil and glucosinolate contents of rapeseed was also high. NOCH calibration models yielded standard errors of cross validation (SECV) and coefficients of determination of cross validation (R2CV) of 0.3% and 0.96 for DM, 0.6% and 0.69 for CP, 0.9% and 0.71 for oil, and 2.2 μmol/g and 0.40 for glucosinolate, respectively. The NOC systems showed high potential for the determination of DM, ST, and soluble sugars (SS) content of silage maize hybrids. The NOC system equipped with a conveyor belt design yielded calibration models with SEP and R2V of 0.9% and 0.93 for DM, and 2.1% and 0.78 for ST, respectively. For the NOC system equipped with the spout design, the SEP and R2V amounted to 1.4% and 0.84 for DM, 2.3% and 0.75 for ST, and 0.9% and 0.81 for SS. The potential of both NOC systems for determination of fiber contents (CF, ADF, and NDF), digestibility and energy-related traits was lower than for DM, ST, and SS. The precision of NOCH for the determination of DM content in maize grain was higher than by traditional drying-oven method. A higher precision of NOCH is also expected for other traits and may also be extended to the NOC systems because the sampling error associated with traditional processes of data collection is reduced drastically by NOC and NOCH. The investigation of the effects caused by the calibration technique, mathematical transformation of the near-infrared spectra, and scatter correction on the development of NOCH calibration models for the prediction of DM, CP, and ST content in maize grain revealed that calibration technique was the most important factor affecting the prediction ability, whereas the importance of mathematical transformation and scatter correction depended on the particular constituent considered. Presently, there exists high uncertainty about the optimal NOC and NOCH sample presentation designs for agricultural harvesters. The dynamic signal range, i.e., the range of spectral values on which predictions are based, and the amount of plot material measured were identified as guide parameters for optimization of sample presentation designs. In addition, calibration transferability between NOC systems with different sample presentation designs proved to be feasible after merging spectra from both NOC systems in the calibration set. In conclusion, NOC and NOCH show high potential for replacing laboratory NIRS analysis of several traits in a plant breeding context and yield a more accurate and precise evaluation of field plot characteristics. Therefore, technological applications of the electromagnetic radiation is predicted to have a high impact in plant breeding, precision farming, and agriculture.
A comprehensive characterization of agronomic and end-use quality phenotypes across a quinoa world core collection
(2023) Craine, Evan B.; Davies, Alathea; Packer, Daniel; Miller, Nathan D.; Schmöckel, Sandra M.; Spalding, Edgar P.; Tester, Mark; Murphy, Kevin M.
Quinoa (Chenopodium quinoa Willd.), a pseudocereal with high protein quality originating from the Andean region of South America, has broad genetic variation and adaptability to diverse agroecological conditions, contributing to the potential to serve as a global keystone protein crop in a changing climate. However, the germplasm resources currently available to facilitate quinoa expansion worldwide are restricted to a small portion of quinoa’s total genetic diversity, in part because of day-length sensitivity and issues related to seed sovereignty. This study aimed to characterize phenotypic relationships and variation within a quinoa world core collection. The 360 accessions were planted in a randomized complete block design with four replicates in each of two greenhouses in Pullman, WA during the summer of 2018. Phenological stages, plant height, and inflorescence characteristics were recorded. Seed yield, composition, thousand seed weight, nutritional composition, shape, size, and color were measured using a high-throughput phenotyping pipeline. Considerable variation existed among the germplasm. Crude protein content ranged from 11.24% to 17.81% (fixed at 14% moisture). We found that protein content was negatively correlated with yield and positively correlated with total amino acid content and days to harvest. Mean essential amino acids values met adult daily requirements but not leucine and lysine infant requirements. Yield was positively correlated with thousand seed weight and seed area, and negatively correlated with ash content and days to harvest. The accessions clustered into four groups, with one-group representing useful accessions for long-day breeding programs. The results of this study establish a practical resource for plant breeders to leverage as they strategically develop germplasm in support of the global expansion of quinoa.
Development and assessment of a multi-sensor platform for precision phenotyping of small grain cereals under field conditions
(2014) Busemeyer, Lucas; Würschum, Tobias
The growing world population, changing food habits especially to increased meat consumption in newly industrialized countries, the growing demand for energy and the climate change pose major challenges for tomorrows agriculture. The agricultural output has to be increased by 70% by 2050 to achieve food and energy security for the future and 90% of this increase must be achieved by increasing yields on existing agricultural land. Achieving this increase in yield is one of the biggest challenges for the global agriculture and requires, among other things, an efficient breeding of new, higher-yielding varieties adapted to the predicted climate change. To achieve this goal, new methods need to be established in plant breeding which include efficient genotyping and phenotyping approaches of crops. Enormous progress has been achieved in the field of genotyping which enables to gain a better understanding of the molecular basis of complex traits. However, phenotyping must be considered as equally important as genomic approaches rely on high quality phenotypic data and as efficient phenotyping enables the identification of superior lines in breeding programs. In contrast to the rapid development of genotyping approaches, phenotyping methods in plant breeding have changed only little in recent decades which is also referred to as phenotyping bottleneck. Due to this discrepancy between available phenotypic and genotypic information a significant potential for crop improvement remains unexploited. The aim of this work was the development and evaluation of a precision phenotyping platform for the non-invasive measurement of crops under field conditions. The developed platform is assembled of a tractor with 80 cm ground clearance, a carrier trailer and a sensor module attached to the carrier trailer. The innovative sensors for plant phenotyping, consisting of several 3D Time-of-Flight cameras, laser distance sensors, light curtains and a spectral imaging camera in the near infrared reflectance (NIR) range, and the entire system technology for data acquisition were fully integrated into the sensor module. To operate the system, software with a graphical user interface has been developed that enables recording of sensor raw data with time- and location information which is the basis of a subsequent sensor and data fusion for trait determination. Data analysis software with a graphical user interface was developed under Matlab. This software applies all created sensor models and algorithms on sensor raw data for parameter extraction, enables the flexible integration of new algorithms into the data analysis pipeline, offers the opportunity to generate and calibrate new sensor fusion models and allows for trait determination. The developed platform facilitates the simultaneous measurement of several plant parameters with a throughput of over 2,000 plots per day. Based on data of the years 2011 and 2012, extensive calibrations were developed for the traits plant height, dry matter content and biomass yield employing triticale as a model species. For this purpose, 600 plots were grown each year and recorded twice with the platform followed by subsequent phenotyping with state-of-the-art methods for reference value generation. The experiments of each year were subdivided into three measurements at different time points to incorporate information of three different developmental stages of the plants into the calibrations. To validate the raw data quality and robustness of the data collection and reduction process, the technical repeatability for all developed data analysis algorithms was determined. In addition to these analyses, the accuracy of the generated calibrations was assessed as the correlations between determined and observed phenotypic values. The calibration of plant height based on light curtain data achieved a technical repeatability of 0.99 and a correlation coefficient of 0.97, the calibration of dry matter content based on spectral imaging data a of 0.98 and a of 0.97. The generation and analysis of dry biomass calibrations revealed that a significant improvement of measurement accuracy can be achieved by a fusion of different sensors and data evaluations. The calibration of dry biomass based on data of the light curtains, laser distance sensors, 3D Time-of-Flight cameras and spectral imaging achieved a of 0.99 and a of 0.92. The achieved excellent results illustrate the suitability of the developed platform, the integrated sensors and the data analysis software to non-invasively measure small grain cereals under field conditions. The high utility of the platform for plant breeding as well as for genomic studies was illustrated by the measurement of a large population with a total of 647 doubled haploid triticale lines derived from four families that were grown in four environments. The phenotypic data was determined based on platform measurements and showed a very high heritability for dry biomass yield. The combination of these phenotypic data with a genomic approach enabled the identification of quantitative trait loci (QTL), i.e., chromosomal regions affecting this trait. Furthermore, the repeated measurements revealed that the accumulation of biomass is controlled by temporal genetic regulation. Taken together, the very high robustness of the system, the excellent calibration results and the high heritability of the phenotypic data determined based on platform measurements demonstrate the utility of the precision phenotyping platform for plant breeding and its enormous potential to widen the phenotyping bottleneck.
Estimating heritability in plant breeding programs
(2019) Schmidt, Paul; Piepho, Hans-Peter
Heritability is an important notion in, e.g., human genetics, animal breeding and plant breeding, since the focus of these fields lies on the relationship between phenotypes and genotypes. A phenotype is the composite of an organism’s observable traits, which is determined by its underlying genotype, by environmental factors and by genotype-environment interactions. For a set of genotypes, the notion of heritability expresses the proportion of the phenotypic variance that is attributable to the genotypic variance. Furthermore, as it is an intraclass correlation, heritability can also be interpreted as, e.g., the squared correlation between phenotypic and genotypic values. It is important to note that heritability was originally proposed in the context of animal breeding where it is the individual animal that represents the basic unit of observation. This stands in contrast to plant breeding, where multiple observations for the same genotype are obtained in replicated trials. Furthermore, trials are usually conducted as multi-environment trials (MET), where an environment denotes a year × location combination and represents a random sample from a target population of environments. Hence, the observations for each genotype first need to be aggregated in order to obtain a single phenotypic value, which is usually done by obtaining some sort of mean value across trials and replicates. As a consequence, heritability in the context of plant breeding is referred to as heritability on an entry-mean basis and its standard estimation method is a linear combination of variances and trial dimensions. Ultimately, I find that there are two main uses for heritability in plant breeding: The first is to predict the response to selection and the second is as a descriptive measure for the usefulness and precision of cultivar trials. Heritability on an entry-mean basis is suited for both purposes as long as three main assumptions hold: (i) the trial design is completely balanced/orthogonal, (ii) genotypic effects are independent and (iii) variances and covariances are constant. In the last decades, however, many advancements in the methodology of experimental design for and statistical analysis of plant breeding trials took place. As a consequence it is seldom the case that all three of above mentioned assumptions are met. Instead, the application of linear mixed models enables the breeder to straightforwardly analyze unbalanced data with complex variance structures. Chapter 2 exemplarily demonstrates some of the flexibility and benefit of the mixed model framework for typically unbalanced MET by using a bivariate mixed model analyses to jointly analyze two MET for cultivar evaluation, which differ in multiple crucial aspects such as plot size, trial design and general purpose. Such an approach can lead to higher accuracy and precision of the analysis and thus more efficient and successful breeding programs. It is not clear, however, how to define and estimate a generalized heritability on an entry-mean basis for such settings. Therefore, multiple alternative methods for the estimation of heritability on an entry-mean basis have been proposed. In Chapter 3, six alternative methods are applied to four typically unbalanced MET for cultivar evaluation and compared to the standard method. The outcome suggests that the standard method over-estimates heritability, while all of the alternative methods show similar, lower estimates and thus seem able to handle this kind of unbalanced data. Finally, it is argued in Chapter 4 that heritability in plant breeding is not actually based on or aiming at entry-means, but on the differences between them. Moreover, an estimation method for this new proposal of heritability on an entry-difference basis (H_Delta^2/h_Delta^2) is derived and discussed, as well as exemplified and compared to other methods via analyzing four different datasets for cultivar evaluation which differ in their complexity. I argue that regarding the use of heritability as a descriptive measure, H_Delta^2/h_Delta^2, can on the one hand give a more detailed and meaningful insight than all other heritability methods and on the other hand reduces to other methods under certain circumstances. When it comes to the use of heritability as a means to predict the response to selection, the outcome of this work discourages this as a whole. Instead, response to selection should be simulated directly and thus without using any ad hoc heritability measure.
Extensions of genomic prediction methods and approaches for plant breeding
(2013) Technow, Frank; Melchinger, Albrecht E.
Marker assisted selection (MAS) was a first attempt to exploit molecular marker information for selection purposes in plant breeding. The MAS approach rested on the identification of quantitative trait loci (QTL). Because of inherent shortcomings of this approach, MAS failed as a tool for improving polygenic traits, in most instances. By shifting focus from QTL identification to prediction of genetic values, a novel approach called 'genomic selection', originally suggested for breeding of dairy cattle, presents a solution to the shortcomings of MAS. In genomic selection, a training population of phenotyped and genotyped individuals is used for building the prediction model. This model uses the whole marker information simultaneously, without a preceding QTL identification step. Genetic values of selection candidates, which are only genotyped, are then predicted based on that model. Finally, the candidates are selected according their predicted genetic values. Because of its success, genomic selection completely revolutionized dairy cattle breeding. It is now on the verge of revolutionizing plant breeding, too. However, several features set apart plant breeding programs from dairy cattle breeding. Thus, the methodology has to be extended to cover typical scenarios in plant breeding. Providing such extensions to important aspects of plant breeding are the main objectives of this thesis. Single-cross hybrids are the predominant type of cultivar in maize and many other crops. Prediction of hybrid performance is of tremendous importance for identification of superior hybrids. Using genomic prediction approaches for this purpose is therefore of great interest to breeders. The conventional genomic prediction models estimate a single additive effect per marker. This was not appropriate for prediction of hybrid performance because of two reasons. (1) The parental inbred lines of single-cross hybrids are usually taken from genetically very distant germplasm groups. For example, in hybrid maize breeding in Central Europe, these are the Dent and Flint heterotic groups, separated for more than 500 years. Because of the strong divergence between the heterotic groups, it seemed necessary to estimate heterotic group specific marker effects. (2) Dominance effects are an important component of hybrid performance. They had to be included into the prediction models to capture the genetic variance between hybrids maximally. The use of different heterotic groups in hybrid breeding requires parallel breeding programs for inbred line development in each heterotic group. Increasing the training population size with lines from the opposite heterotic group was not attempted previously. Thus, a further objective of this thesis was to investigate whether an increase in the accuracy of genomic prediction can be achieved by using combined training sets. Important traits in plant breeding are characterized by binomially distributed phenotypes. Examples are germination rate, fertility rates, haploid induction rate and spontaneous chromosome doubling rate. No genomic prediction methods for such traits were available. Therefore, another objective was to provide methodological extensions for such traits. We found that incorporation of dominance effects for genomic prediction of maize hybrid performance led to considerable gains in prediction accuracy when the variance attributable to dominance effects was substantial compared to additive genetic variance. Estimation of marker effects specific to the Dent and Flint heterotic group was of less importance, at least not under the high marker densities available today. The main reason for this was the surprisingly high linkage phase consistency between Dent and Flint heterotic groups. Furthermore, combining individuals from different heterotic groups (Flint and Dent) into a single training population can result in considerable increases in prediction accuracy. Our extensions of the prediction methods to binomially distributed data yielded considerably higher prediction accuracies than approximate Gaussian methods. In conclusion, the developed extensions of prediction methods (to hybrid prediction and binomially distributed data) and approaches (training populations combining heterotic groups) can lead to considerable, cost free gains in prediction accuracy. They are therefore valuable tools for exploiting the full potential of genomic selection in plant breeding.
Factors influencing the accuracy of genomic prediction in plant breeding
(2017) Schopp, Pascal; Melchinger, Albrecht E.
Genomic prediction (GP) is a novel statistical tool to estimate breeding values of selection candidates without the necessity to evaluate them phenotypically. The method calibrates a prediction model based on data of phenotyped individuals that were also genotyped with genome-wide molecular markers. The renunciation of an explicit identification of causal polymorphisms in the DNA sequence allows GP to explain significantly larger amounts of the genetic variance of complex traits than previous mapping-based approaches employed for marker-assisted selection. For these reasons, GP rapidly revolutionized dairy cattle breeding, where the method was originally developed and first implemented. By comparison, plant breeding is characterized by often intensively structured populations and more restricted resources routinely available for model calibration. This thesis addresses important issues related to these peculiarities to further promote an efficient integration of GP into plant breeding.
Optimizing the prediction of genotypic values accounting for spatial trend andpopulation structure
(2010) Müller, Bettina Ulrike; Piepho, Hans-Peter
Different effects, like the design of the field trial, agricultural practice, competition between neighboured plots, climate as well as the spatial trend, have an influence on the non-genotypic variation of the genotype. This effects influence the prediction of the genotypic value by the non-genotypic variation. The error, which results from the influence of the non-genotypic variation, can be separated from the phenotypic value by field design and statistical models. The integration of different information, like spatial trend or marker, can lead to an improved prediction of genotypic values. The present work consists of four studies from the area of plant breeding and crop science, in which the prediction of the genotypic values was optimized with inclusion of the above mentioned aspects. Goals of the work were: (1) to compare the different spatial models and to find one model, which is applicable as routine in plant breeding analysis, (2) to optimize the analysis of unreplicated trials of plant breeding experiments by improving the allocation of replicated check genotypes, (3) to improve the analysis of intercropping experiments by using spatial models and to detect the neighbour effect between the different cultivars, and (4) to optimize the calculation of the genome-wide error rate in association mapping experiments by using an approach which regards the population structure. Different spatial models and a baseline model, which reflects the randomization of the field trial, were compared in three of the four studies. In one study the models were compared on basis of different efficiency criteria with the goal to find a model, which is applicable as routine in plant breeding experiments. In the second study the different spatial models and the baseline model were compared on unreplicated trials, which are used in the early generation of the plant breeding process. Adjacent to the comparison of the models in this study different designs were compared with the goal to see if a non-systematic allocation of check genotypes is more preferable than a systematic allocation of check genotypes. In the third study these different models were tested for intercropping experiments. In this study it should be tested, if an improvement is expectable for these non randomized or restricted randomized trials by using a spatial analysis. The results of the three studies are that no spatial model could be found, which is preferable over all other spatial models. In a lot of cases the baseline model, which regards only the randomization, but no spatial trend, was better than the spatial models, also for the restricted or non-randomized intercropping trials. In all three studies the basic principle was followed to start first with the baseline model, which is based on the randomization theory, and then to extend it by spatial trend, if the model fit can be improved. In the second study the systematic and non-systematic allocation of check plots in unreplicated trials were compared to solve the question if a non-systematic allocation leads to more efficient estimates of genotypes as the systematic allocation. The non-systematic allocation of check plots led to an unbiased estimation in three of four uniformity trials. As well as in the third study an analysis was done, if the border plots of the different cultivars are influenced by the neighboured cultivar and if there are significant differences to the inner plot. The position of the cultivars, border plot or inner plot, had a significant influence of the yield. If maize was cultivated adjacent to pea, the yield of the border plot of maize was much higher than the inner plot of maize. When wheat was cultivated behind maize, there were no significant differences in the yield, if the plot was a border plot or inner plot. In addition to optimizing the field design for unreplicated trials and the extension of the models by spatial trend the marker information was integrated in a fourth study. An approach was proposed in this study, which calculates the genome wide error for association mapping experiments and accounts for the population structure. Advantages of this approach in contrast to previously published approaches are that the approach on the one hand is not too conservative and on the other hand accounts the population structure. The adherence of the genome wide error rate was tested on three datasets, which were provided by different plant breeding companies. The results of these studies, which were obtained in this thesis, show that by the different extensions, like integration of spatial trend and marker information, and modifications of the field design, an improved prediction of the genotypic values can be achieved.
Phenotypic and genotypic assessment of traits with relevance for hybrid breeding in European winter wheat
(2015) Langer, Simon Martin; Würschum, Tobias
Hybrid breeding in wheat has recently received increased interest, especially in Europe, and large public and private projects investigating hybrid wheat breeding have been launched. Hybrid breeding has been a great success story for allogamous crops and is seen as a promising approach to increase the yield potential in wheat. Wheat covers more of the world’s surface than any other food crop and is the second main staple crop for human consumption. It can be produced under widely varying conditions and is grown all around the globe, yet, yield gain has declined and is lagging behind the needs of the constantly growing human population. Future challenges in wheat breeding such as the establishment of hybrid varieties and the adaptation of breeding germplasm to increasing stresses caused by climatic changes also in Europe require knowledge-based improvements of relevant traits and phenotyping approaches suited for applied high-throughput plant breeding. A major limitation for the establishment and the production of hybrid wheat is the lack of a cost-efficient hybrid seed production system. This requires the generation of parental ideotypes which maximize the cross-fertilization capability. Male parents should have an extended time of flowering, extrude anthers and widely shed large amounts of viable pollen. Females need increased receptivity for male pollen by opening the glumes and extruding stigmatic hair. Furthermore, male plants should be taller than females and a synchronized timing of flowering between the two parents is also of utmost importance. Employing a set of European elite winter wheat lines, we developed and evaluated phenotyping methods for important floral and flowering traits with relevance for improved cross-pollination (Publication I). We observed high heritabilities for important traits such as ‘pollen mass’ (h2=0.72) and ‘anther extrusion’ (h2=0.91). In addition, genotypic variances were significant which warrants further breeding success. Positive correlations were found among important flowering and floral characteristics which enables the improvement of outcrossing by indirect selection. ‘Pollen mass’ for example, was associated with ‘anther extrusion’, ‘anther length’ and ‘plant height’. Our findings suggest the utility of the developed phenotyping approaches for applied plant breeding and the potential of the traits to assist in the design of the male ideotype for increased cross-fertilization. We investigated the genetic architecture of flowering time and plant height (Publication II and III). A panel of 410 European winter wheat varieties was genotyped by a genotyping-by-sequencing approach and in addition, analyzed for the effects of specific candidate genes. The major factor affecting flowering time was the photoperiod regulator Ppd-D1 (58.2% of explained genotypic variance) followed by Ppd-B1 copy number variation (3.2%). For plant height, the two candidate loci Rht-D1 (37.0%) and Rht-B1 (14.0%) had the largest effects on the trait but contrary to reports in the literature did not contribute to flowering time control. In addition, we identified several small effect QTL and epistatic QTL responsible for fine-adjustments of these two traits. Population structure and genetic relatedness in European elite wheat lines was assessed using different types of markers (Publication IV). Results for relatedness differed for the marker types but consistently showed the absence of a major population structure. Regarding the large wheat genome our results revealed that a high number of markers is necessary as there are regions with only low coverage. Concordantly, we were not able to identify the major flowering locus Ppd-D1 without targeted candidate gene analysis. Observations on the findings on population structure could be confirmed in Publication II and III and in addition, the geographical distribution of important flowering time and plant height genes displayed the historical development of wheat breeding in Europe. This information on genetic relatedness among lines can also be employed to assist the establishment of hybrid wheat.
Potential impacts of iron biofortification in India
(2006) Bhutta, Zulfiqar A.; Stein, Alexander J.; Qaim, Matin; Meenakshi, J. V.; Nestel, Penelope; Sachdev, H. P. S.
Iron deficiency is a widespread nutritional problem in developing countries, causing impaired physical activity and cognitive development, as well as maternal mortality. Although food fortification and supplementation programmes have been effective in some countries, their overall success remains limited. Biofortification, that is, breeding crops for higher micronutrient content, is a relatively new approach. We propose a methodology for ex-ante impact assessment of iron biofortification, which builds on disability-adjusted life years (DALYs) and a large household data set. Our analysis of iron-rich rice and wheat in India indicates sizeable potential health benefits. The cost-effectiveness of iron biofortification compares favourably with other interventions.
Statistical methods for analysis of multienvironment trials in plant breeding : accuracy and precision
(2021) Buntaran, Harimurti; Piepho, Hans-Peter
Multienvironment trials (MET) are carried out every year in different environmental conditions to evaluate a vast number of cultivars, i.e., yield, because different cultivars perform differently in various environmental conditions, known as genotype×environment interactions. MET aim to provide accurate information on cultivar performance so that a recommendation of which cultivar performs the best in a growers’ field condition can be available. MET data is often analysed via mixed models, which allow the cultivar effect to be random. The random effect of cultivar enables genetic correlation to be exploited across zones and considering the trials’ heterogeneity. A zone can be viewed as a larger target of population environments. The accuracy and precision of the cultivar predictions are crucial to be evaluated. The prediction accuracy can be evaluated via a cross-validation (CV) study, and the model selection can be done based on the lowest mean squared error prediction (MSEP). Also, since the trials’ locations hardly coincide with growers’ field, the precision of predictions needs to be evaluated via standard errors of predictions of cultivar values (SEPV) and standard errors of the predictions of pairwise differences of cultivar values (SEPD). The central objective of this thesis is to assess the model performance and conduct model selection via a CV study for zone-based cultivar predictions. Chapter 2 assessed the performance between empirical best linear unbiased estimations (EBLUE) and empirical best linear unbiased predictions (EBLUP) for zone-based prediction. Different CV schemes were done for the single-year and multi-year datasets to mimic the practice. A complex covariance structure such as factor-analytic (FA) was imposed to account for the heterogeneity of cultivar×zone (CZ) effect. The MSEP showed that the EBLUP models outperformed the EBLUE models. The zonation was necessary since it improved the accuracy and was preferable to make cultivar recommendations. The FA structure did not improve the accuracy compared to the simpler covariance structure, and so the EBLUP model with a simple covariance structure is sufficient for the single and multi-year datasets. Chapter 3 assessed the single-stage and stagewise analyses. The three weighting methods were compared in the stagewise analysis: two diagonal approximation methods and the fully efficient method with the unweighted analysis. The assessment was based on the MSEP instead of Pearson’s and Spearman’s correlation coefficients since the correlation coefficients are often very close between the compared models. The MSEP showed that the single-stage EBLUP and the stagewise weighting EBLUP strategy were very similar. Thus, the loss of information due to diagonal approximation is minor. In fact, the MSEP showed a more apparent distinction between the single-stage and the stagewise weighting analyses with the unweighted EBLUE compared to the correlation coefficients. The simple compound-symmetric covariance structure was sufficient for the CZ effect than the more complex structures. The choice between the single-stage and stagewise weighting analysis, thus, depends on the computational resources and the practicality of data handling. Chapter 4 assessed the accuracy and precision of the predictions for the new locations. The environmental covariates were combined with the EBLUP in the random coefficient (RC) models since the covariates provide more information for the new locations. The MSEP showed that the RC models were not the model with the smallest MSEP, but the RC models had the lowest SEPV and SEPD. Thus, the model selection can be done by joint consideration of the MSEP, SEPV, and SEPD. The models with EBLUE and covariate interaction effects performed poorly regarding the MSEP. The EBLUP models without RC performed best, but the SEPV and SEPD were large, considered unreliable. The covariate scale and selection are essential to obtain a positive definite covariance matrix. Employing unstructured covariance int the RC is crucial to maintaining the RC models’ invariance feature. The RC framework is suitable to be implemented with GIS data to provide an accurate and precise projection of cultivar performance for the new locations or environments. To conclude, the EBLUP model for zoned-based predictions should be preferred to obtain the predictions and rankings closer to the true values and rankings. The stagewise weighting analysis can be recommended due to its practicality and its computational efficiency. Furthermore, projecting cultivar performances to the new locations should be done to provide more targeted information for growers. The available environmental covariates can be utilised to improve the predictions’ accuracy and precision in the new locations in the RC model framework. Such information is certainly more valuable for growers and breeders than just providing means across a whole target population of environments.