Browsing by Subject "BLUP"

Now showing 1 - 3 of 3

Biometrical approaches for analysing gene bank evaluation data on barley (Hordeum spec.)
(2007) Hartung, Karin; Piepho, Hans-Peter
This thesis explored methods to statistically analyse phenotypic data of gene banks. Traits of the barley data (Hordeum spp.) of the gene bank of the IPK-Gatersleben were evaluated. The data of years 1948-2002 were available. Within this period the ordinal scale changed from a 0-5 to a 1-9 scale after 1993. At most gene banks reproduction of accessions is currently done without any experimental design. With data of a single year only rarely do accessions have replications and there are only few replications of a single check for winter and summer barley. The data of 2002 were analysed separately for winter and summer barley using geostatistical methods. For the traits analysed four types of variogram model (linear, spherical, exponential and Gaussian) were fitted to the empirical variogram using non-linear regression. The spatial parameters obtained by non-linear regression for every variogram model then were implemented in a mixed model analysis and the four model fits compared using Akaike's Information Criterion (AIC). The approach to estimate the genetical parameter by Kriging can not be recommended. The first points of the empirical variogram should be explained well by the fitted theoretical variogram, as these represent most of the pairwise distances between plots and are most crucial for neighbour adjustments. The most common well-fitting geostatistical models were the spherical and the exponential model. A nugget effect was needed for nearly all traits. The small number of check plots for the available data made it difficult to accurately dissect the genetical effect from environmental effects. The threshold model allows for joint analysis of multi-year data from different rating scales, assuming a common latent scale for the different rating systems. The analysis suggests that a mixed model analysis which treats ordinal scores as metric data will yield meaningful results, but that the gain in efficiency is higher when using a threshold model. The threshold model may also be used when there is a metric scale underlying the observed ratings. The Laplace approximation as a numerical method to integrate the log-likelihood for random effects worked well, but it is recommended to increase the number of quadrature points until the change in parameter estimates becomes negligible. Three rating methods (1%, 5%, 9-point rating) were assessed by persons untrained (A) and experienced (B) in rating. Every person had to rate several pictograms of diseased leaves. The highest accuracy was found with Group B using the 1%-scale and with Group A using the 5%-scale. With a percentage scale Group A tended to use values that are multiples of 5%. For the time needed per leaf assessment the Group B was fastest when using the 5% rating scale. From a statistical point of view both percent ratings performed better than the ordinal rating scale and the possible error made by the rater is calculable and usually smaller than with ratings by rougher methods. So directly rating percentages whenever possible leads to smaller overall estimation errors, and with proper training accuracy and precision can be further improved. For gene banks augmented designs as proposed by Federer and by Lin et al. offer themselves, so an overview is given. The augmented designs proposed by Federer have the advantage of an unbiased error estimate. But the random allocation of checks is a problem. The augmented design by Lin et al. always places checks in the centre plot of every whole plot. But none of the methods is based on an explicit statistical model, so there is no well-founded decision criterion to select between them. Spatial analysis can be used to find an optimal field layout for an augmented design, i.e. a layout that yields small least significant differences. The average variance of a difference and the average squared LSD were used to compare competing designs, using a theoretical approach based on variations of two anisotropic models and different rotations of anisotropy axes towards field reference axes. Based on theoretical calculations, up to five checks per block are recommended. The nearly isotropic combinations led to designs with large quadratic blocks. With strongly anisotropic combinations the optimal design depends on degree of anisotropy and rotation of anisotropy axes: without rotation small elongated blocks are preferred; the closer the rotation is to 45° the more squarish blocks and the more checks are appropriate. The results presented in this thesis may be summarised as follows: Cultivation for regeneration of accessions should be based on a meaningful and statistically analysable experimental field design. The design needs to include checks and a random sample of accessions from the gene pool held at the gene bank. It is advisable to utilise metric or percentage rating scales. It can be expected that using a threshold model increases the quality of multivariate analysis and association mapping studies based on phenotypic gene bank data.
Response to modified recurrent full-sib selection in two European F2 maize populations analyzed with quantitative genetic methods
(2006) Flachenecker, Christian; Melchinger, Albrecht E.
Many plant breeding strategies lead to a reduction in the genetic variance of the source population. However, a sufficient genetic variance is essential for the long-term selection response. Hence, the aim of recurrent selection (RS) is a continuous increase in the frequencies of favorable alleles while maintaining genetic variability in the population. Several intrapopulation RS methods have been proposed in maize: e.g., mass selection, half-sib selection, full-sib (FS) selection, S1 selection. Among them, recurrent FS selection is characterized by a short cycle length, complete parental control, and a high selection response. The goal of this thesis was to investigate the changes in the population structure over several cycles of a modified recurrent FS selection program in two European F2 maize populations. In detail, the objectives were to (i) monitor trends across selection cycles in the estimates of population mean, inbreeding coefficients, and variance components, (ii) determine selection response for per se and testcross performance, (iii) compare predicted with realized selection response, (iv) extend the population diallel analysis under full consideration of inbreeding depression due to random genetic drift, (v) separate genetic effects due to selection from those due to random genetic drift, and (vi) investigate the usefulness of best linear unbiased prediction (BLUP) estimates of parents for predicting progeny performance under the recurrent FS selection scheme applied. Four early maturing European flint inbreds were used as parents to produce two F2 populations (A×B and C×D). Both populations were three times intermated by chain crossing to reduce the gametic phase disequilibrium. Starting from the F2Syn3 population obtained in this manner, a modified recurrent FS selection program was conducted over four cycles in population A×B and over seven cycles in population C×D. In each cycle, 144 FS families were tested in field trials and, in parallel, six plants from each FS family were selfed. The selfed ears of the 36 families with the highest selection index (SI = 2 × dry matter content + grain yield) were selected and intermated according to a pseudo-factorial mating scheme. In this mating scheme, the gametic contribution of the best selected FS families is doubled compared with the gametic contribution of the remaining selected FS families. Afterwards, all cycles of both populations were tested in two population diallel analyses in six environments. Based on the known pedigree records, the inbreeding coefficient of each FS family and the coancestry coefficients among them were calculated. Variance components and BLUP values were obtained using phenotypic means and coancestry coefficients. For grain yield, the selection response per cycle, which could be expected after correcting for the effects of random genetic drift, was higher than reported in the literature (14.1% and 8.3% in populations A×B and C×D, respectively). We ascribe the comparatively high selection response mainly to the pseudo-factorial mating scheme. This mating scheme is expected to increase the selection response compared with commonly applied random mating schemes, without a major reduction in the effective population size (Ne). In this study, the expected Ne was 32, suggesting a minor influence of random genetic drift compared with that of selection. This assumption was verified by an extended population diallel analysis, showing that random genetic drift reduced the selection response only by about 1-2% in both populations. In contrast to an estimation of variance components with moment estimators, the REML procedure has no special requirements on the mating scheme and accounts for any relationship among families in a breeding population. As expected from the high Ne applied in our study, we observed only a moderate decrease in additive variance for grain yield and grain moisture in both populations. Nevertheless, the variance components were still associated with high standard errors, which prevented the revealing of trends across cycles. A larger number of test locations and larger population size would reduce the standard errors of variance components at the cost of oversized and expensive field trials. Methods for predicting the performance of progenies are important to optimize RS programs. Due to simplifying assumptions, a prediction with phenotypic means is often inaccurate. An alternative method is BLUP, which was suggested for predicting the performance of untested single-cross hybrids but has not been applied in RS programs. In our study, the prediction of progeny performance based on BLUP was only marginally better than prediction based on the phenotypic mean. However, higher degree of relationship between the entries and lower heritabilities would increase the advantage of BLUP compared with phenotypic means.
The development of phenotypic protocols and adjustment of experimental designs in Pelargonium zonale breeding
(2018) Molenaar, Heike; Piepho, Hans-Peter
Ornamental plant variety improvement is limited by current phenotyping approaches and the lack of use of experimental designs. Robust phenotypic data obtained from experiments laid out to best control local variation by blocking allow adequate statistical analysis and are crucial for any breeding purpose, including MAS. Often experiments consist of multiple phases like in P. zonale breeding, where in the first phase stock plants are cultivated to obtain the stem cutting count and in the second phase the stem cuttings are further assess for root formation. The first analyses of rooting experiments raised questions regarding options for improving the two-phase experimental layout, for example whether there is a disadvantage to using exactly the same design in both phases. The other question was, whether a design can be optimized across both phases, such that the MVD can be decreased. Instead of generating a separate layout for each phase. Moreover, optimal selection methods that maximize selection gain in P. zonale breeding based on available data collected from unreplicated trials and containing pedigree information were sought. This thesis was conducted to evaluate the benefits of using two-phase experimental designs and corresponding analysis in P. zonale for production related traits, for which it was necessary to establish phenotyping protocols. To optimize the rooting experiments with their two-phase nature, alternative approaches were explored involving two-phase design generation either in phase wise order or across phases. Furthermore, selection methods considering pedigreeinformation (family-index selection) or not (individual selection), were evaluated to enhance selection efficiency in P. zonale breeding. The benefits of using experimental designs in P. zonale breeding was shown by the simulated response to selection. Alternative designs were evaluated by the MVD obtained by the intrablock analysis and the joint inter-block-intra-block analysis. The efficiency of individual and family-index selection was evaluated in terms of heritability obtained from linear mixed models implementing the selection methods. Simulated response to selection varied greatly, depending on the genotypic variances of the breeding population and traits. However, by using efficient designs allowing adequate analysis, a varietal improvement of over 20% of stock plant reduction is possible for stem cutting count, root formation, branch count and flower count. The smallest MVD for alternative designs was most frequently obtained for designs generated across phases rather than for each phase separately, in particular when both phases of the design were separated with a single pseudolevel. Family-index selection was superior to individual selection in P. zonale indicating that the pedigree-based BLUP procedure can further enhance selection efficiency in productionrelated traits in P. zonale. The quantification of genotypic variation by phenotypic protocols and the optimized two-phase designs for estimating genotypic values were necessary and successful steps in laying the foundation for effective MAS. Phenotypic protocols effectively characterized the genetic material on an observational unit level, while the two-phase experimental designs enabled effective characterization on a genotype level by adjusting entry means using linear mixed models. The resulting adjusted entry means are the basis for future genotype phenotype association for MAS.