Yield Trial Data Research Articles

Genotype/Genotype × Environment Biplot Analysis for Cultivar Evaluation and Mega-environment Investigation in Primocane-fruiting Red Raspberry

Primocane-fruiting (PF) red raspberry (Rubus idaeus L.) cultivars are being grown in many regions as their popularity increases. However, testing of this perennial fruit crop is expensive and requires many years. Large genotype (G) × environment (E) interactions can make identification of superior genotypes difficult. The G/G × E (GGE) biplot can be used to measure cultivar performance and group locations into mega-environments. The GGE biplot was applied to yield trial data of three PF red raspberry cultivars Autumn Bliss, Heritage, and Redwing grown in 17 environments (year-location combinations). The 17 environments encompassed six locations in Ontario and Quebec, Canada between 1989 and 1996. `Autumn Bliss' produced the highest yields in 11 of 17 environments. `Heritage' was usually the lowest yielding cultivar. Two mega-environments were identified based on the performance of `Autumn Bliss' and `Redwing'. Some environmental variables were likely to be responsible for the discriminating ability of the test environments as they were correlated with the primary effects. The GGE biplot was an effective analysis to determine mega-environments and the cultivars best adapted to each.

Developing Genetic Coefficients for Crop Simulation Models with Data from Crop Performance Trials

Successful uses of crop models in technology transfer and decision support tools require that coefficients describing new cultivars be available as soon as the cultivars are marketed. The objectives of this study were (i) to develop an approach to estimate cultivar coefficients for the CROPGRO–Soybean model from typical information provided by crop performance tests, (ii) to evaluate the suitability of yield trial data for deriving genetic coefficients and site‐specific soil traits for use in crop models, and (iii) to explore the extent to which our approach allowed the crop model to reproduce observed genotype × environment (GE) interactions, cultivar ranking, and year‐to‐year yield variability. Crop performance tests typically record harvest maturity date, seed yield, seed size, height, and lodging. A stepwise procedure using data on 11 cultivars grown at five sites in Georgia over 4 to 10 yr efficiently decreased the root mean square error (RMSE) between observed and predicted data. For ‘Stonewall’, a maturity group VII cultivar, the RMSE of 769 kg ha−1 between the actual and modeled seed yield, estimated initially by means of the existing general maturity group coefficients, was reduced to 404 kg ha−1 For the same cultivar, the initial RMSE of 5.3 and 9.3 d between the actual and simulated anthesis and harvest maturity dates, respectively, estimated by means of the existing general maturity group coefficients, were reduced to 2.9 and 5.8 d. In addition to deriving useful information on site characteristics and cultivar traits, our approach has enabled CROPGRO to satisfactorily mimic the genotypic yield ranking and much of observed genotype × environment interactions. Across all environments, the difference in genotype ranking based on yield between measured and predicted values was one or less for 61% of the environments.

Econometric estimation of a global spillover matrix for wheat varietal technology

Abstract An econometric approach using international and national yield trial data is employed to estimate a spillover matrix for wheat varietal technology. The global spillover matrix is estimated based on international yield trial data from 1979–1980 to 1987–1988, that include 195 international trial locations and 209 wheat varieties. The locations were classified across countries using the CIMMYT's wheat megaenvironment system and varieties were classified by both their environmental and institutional origin. The model gave good explanatory power and confirmed the location specificity hypothesis, at least, for the varieties developed by national programs (NARS). The spillover matrix shows that NARS varieties developed in the ‘home’ environment generally perform better on average than varieties developed in other megaenvironments. Also, the matrix is not symmetric. CIMMYT varieties perform better on average in irrigated and high rainfall environments than NARS varieties developed for these environments. The yield advantage of CIMMYT varieties in many test megaenvironments indicates the potential of CIMMYT varieties to spill-over to these test megaenvironments. Results also indicate that national programs are efficient in selecting from among imported technologies. Analysis of international data is complemented by the analysis of country-level data for Pakistan and Kenya that confirms the above results. The country-level analysis, however, indicates that CIMMYT germplasm does not do so well in some sub-environments, such as the irrigated short-duration environment. The results of the spillover matrix have implications for the design of crop breeding programs both at the national and international levels. Information provided by the spillover matrix can be utilized by national programs to deploy their resources more efficiently by following a mixed strategy of direct importation of technology in some environments and local development of technologies in other environments which are unique to the country.

Detecting and handling heteroscedasticity in yield trial data

Data from yield trials conducted in different environments are frequently analysed by ordinary analysis of variance. One of the preconditions for such analyses is that error variances be homogeneous across treatments and replicates. If a mixed model with fixed genotypes and random environments is assumed, genotype-environmental interaction is a random effect. For the ordinary ANOVA to be valid in this case, it is then also required that interaction variances be homogeneous for different genotypes. This paper suggests Box's correction of the ANOVA degrees of freedom for the case of heterogeneous interaction and error variances Different estimators of the correction factors are compared via Monte carlo Simulation.

Genetic Models for Predicting Maize Single‐Cross Performance in Unbalanced Yield Trial Data

Methods for predicting single‐cross performance would facilitate maize (Zea mays L.) hybrid development. The objective of this study was to evaluate different genetic models for best linear unbiased prediction (BLUP) of single‐cross performance in unbalanced yield data. Sixty‐seven single crosses between nine Iowa Stiff Stalk Synthetic (SSS) and 16 non‐SSS inbreds were evaluated separately in 31 different sets of multiplication yield trials from 1989 to 1993. Sets of p single crosses were chosen randomly as predictor hybrids. Yields of the remaining m = (67 — p) missing single crosses were predicted as YM = CMP C‐PP1 yp, where ym = m ✕ 1 vector of predicted yields; CMP = m × p matrix of genetic covariances between the missing and predictor hybrids; CPP = p × p phenotypic covariance matrix among the predictor hybrids; and yP = p × 1 vector of average yields of the predictor hybrids, corrected for yield trial effects. Correlations between predicted and observed single‐cross yields in the unbalanced data set ranged from 0.583 to 0.749. The correlations ranged from 0.388 to 0.493 when the missing and predictor hybrids had no parental inbreds in common, suggesting that BLUP may be useful for preliminary screening of single crosses between inbreds that have not been tested in any hybrid combination. Prediction of specific combining ability (SCA) was not effective because SCA variance was small. Genetic models that included additive × additive epistasis did not lead to better predictions of single‐cross performance compared with the model that included only testcross additive and SCA effects. THE IDENTIFICATION of pairs of inbreds with superior.

Gains from Selection under Drought versus Multilocation Testing in Related Tropical Maize Populations

Ideal maize (Zea mays L.) cuitivars for tropical areas should yield well both in the presence and absence of drought, but optimal selection strategies for accomplishing this goal are not dear. This study evaluated progress from selection of two related tropical populations across a broad range of environmental conditions. ‘Tuxpeflo Sequia’ (TS) had undergone fuli‐sib recurrent selection for eight cycles at one location under managed levels of drought stress, while ‘Tuxpefio 1’ (T1) was selected for six cycles in a modified full‐sib selection scheme that relied heavily on muitilocation yield trial data. Combined over 12 environments (with mean yields ranging from 0.30−7.93 Mg ha−1), regression analysis revealed significantly different rates of change per cycle for TS and T1, respectively, for grain yield (1.68 and 1.06%, P < 0.10), anthesis‐siiking interval (ASI) ( −8.59 and 0%, P < 0.10), ears per plant (1.26 and 0%, P < 0.05), and plant height (−0.83 and 1.29%, P < 0.01). Days to anthesis decreased in both TS and T1 (−0.36 and −0.15% per cycle, respectively), but the difference between populations was not significant at P < 0.10. The interaction of environments with the linear rate of gain in grain yield was not significant in either population, indicating similar progress across the range of environmental conditions sampled. Stability analysis indicated that TS Cycles 6 and 8 and the check variety ‘La Posta Sequia Best’ were the most stable and high yielding entries in the trial. Better yield gain in TS is likely related to its selection for reduced ASI under controlled stress at a single site. Selection under managed levels of drought stress at one location together with muitiiocation testing may be desirable components of maize breeding programs for drought‐prone tropical areas.

Application of a generalized Grubbs' model in the analysis of genotype-environment interaction

The presence of genotype-environment interaction is a common feature of yield trial data. Estimating variance components attributable to the interactions of single cultivars is a useful means of assessing phenotypic stability. The stability variance is one such measure. In this paper it is argued that the stability variance is appropriate only if the error variances are homogeneous across genotypes and the number of replicates is the same in each environment. An alternative is suggested for the case that either of these two assumptions is violated.

Clustering Cultivars into Groups without Rank‐Change Interactions

Crossover (i.e., rank‐change) interactions (COIs) are of particular interest in the interpretation of cultivar yield trial data. The shifted multiplicative model (SHMM) was used to search for subsets of cultivars in which cultivar COIs could be regarded as insignificant or small in a 1985 Kentucky wheat (Triticum aestivum L.) trial with 41 cultivars and seven locations. The first step was to obtain a dendrogram by complete linkage (farthest neighbor) cluster analysis with the “distance” between a pair of cultivars defined as the residual sum of squares when SHMM with one multiplicative term (SHMM1) is fitted subject to a constraint that the fitted model be free of COIs. Statistical tests were computed on subsets of cultivars suggested by branching of the dendrogram to identify subsets in which a constrained SHMM, would account for the variation. Five clusters of five cultivars each and four smaller clusters were found satisfactory. Six cultivars did not cluster with any other cultivar. Analysis of all 2 × 2 interactions indicated that the method effectively assigned cultivars involved in large COIs to different clusters. Only 6% of COIs within clusters were significant as compared to 29% among clusters. Clustering would allow breeders and growers to restrict selection of cultivars to the better ones in each cluster.

Agricultural technology: International dimensions (with emphasis on rice)

It is well understood that the biological performance of agricultural crops is affected by soil, temperature, rainfall, and day length. As a consequence, genetic improvements in plants are location specific to a considerable extent. This, in turn, has implications for the design of agricultural research systems and for agricultural extension activities as well. An index for inter-location crop performance differences based on crop yield trial data is developed in this paper. This index measures the geo-climate “distance” or differential between two regions. An application to rice production, utilizing international yield trial data, is reported. The study of rice productivity in India indicates that a state or region in India benefits relatively little from rice research undertaken in another state or region when there is a significant geo-climate distance between the regions. Most technology transfer between regions is thus indirect in that it depends on the adaptive research capacity in the receiving regions.

Interpretive Analysis for Forage Yield Trial Data

AbstractForage cultivar evaluation is often done in small plots with multiple harvests throughout the growing season. Data is often summarized by presenting a yearly total yield for each cultivar in addition to the mean for each harvest date. Data summarization often becomes burdensome and difficult to interpret. Regressing yield against a growth index associated with harvest dates can be utilized to describe forage perfomrmance in a concise and easily interpreted format. Subsets of data from tall fescue (Festuca arundinacea Schreb.) yield trials conducted in Alabama and Kentucky were used to demonstrate the technique. The analysis involves regressing yield of a cultivar against an index calculated as the mean of all entries at a harvest date minus the grand mean. The resulting regression coefficient (b) describes cultivar yield response over several harvest and is indicative of performance under variable growth conditions.

Imputing missing yield trial data

The Additive Main effects and Multiplicative Interaction (AMMI) statistical model has been demonstrated effective for understanding genotype-environment interactions in yields, estimating yields more accurately, selecting superior genotypes more reliably, and allowing more flexible and efficient experimental designs. However, AMMI had required data for every genotype and environment combination or treatment; i.e., missing data were inadmissible. The present paper addresses the problem. The Expectation-Maximization (EM) algorithm is implemented for fitting AMMI depite missing data. This missing-data version of AMMI is here termed "EM-AMMI". EM-AMMI is used to quantify the direct and indirect information in a yield trial, providing theoretical insight into the gain in accuracy observed and into the process of imputing missing data. For a given treatment, the direct yield data are the replicates of that treatment, and the indirect data are all the other yield data in the trial. EM-AMMI is used to inpute missing data for a New York soybean yield trial. Important applications arise from both unintentional and intentional missing data. Empirical measurements demonstrate good predictive success, and statistical theory attributes this success to the Stein effect.

Model Selection and Validation for Yield Trials with Interaction

SUMMARY The additive main effects and multiplicative interaction (AMMI) model first applies the additive analysis of variance (ANOVA) model to two-way data, and then applies the multiplicative principal components analysis (PCA) model to the residual from the additive model, that is, to the interaction. AMMI analysis of yield trial data is a useful extension of the more familiar ANOVA, PCA, and linear regression procedures, particularly given a large genotype-by-environment interaction. Model selection and validation are considered from both predictive and postdictive perspectives, using data splitting and F-tests, respectively. A New York soybean yield trial serves as an example. Yield trials generate observations of yield, ordinarily replicated, for a number of genotypes grown in a number of environments (site-year combinations). Often the data are rather noisy, with a standard deviation for plot yields in excess of 25 % of the mean. An additional challenging feature of these data is the frequent presence of important and complex genotype-by-environment (GE) interactions. Plant breeders use yield trials to identify promising genotypes, and agronomists use them to make recommendations for farmers. The level of success in meeting these goals depends critically on two factors: (i) the accuracy of yield estimates, and (ii) the magnitudes of genotype-by-site, genotype-by-year, and genotype-by-site-by-year interactions (Talbot, 1984). In essence, these two factors reflect within-trial accuracy and between-trial predictability. This paper addresses only the first concern. Nothing is said regarding between-trial predictability, or agrotechnology transfer, other than to observe that success with withintrial accuracy is a necessary prelude to success with between-trial predictability. The within-trial accuracy of a statistical model may be assessed by two fundamentally different criteria: Postdictive success concerns a model's fit to its own data, whereas predictive success concerns the fit between a model constructed using part of the data and validation data not used in modelling. In either case, the statistical setting is that of an incompletely specified model, where empirical considerations enter into the decision to include a given' potential source in the model, or alternatively to relegate it to the model's residual (Bancroft, 1964). Whenever the data are noisy, postdiction and prediction are different tasks, and in general the model chosen by predictive criteria will be different and simpler than the model chosen by postdictive criteria. Because the several replicates of a yield trial constitute a noisy sample, a model chosen by predictive criteria may be expected to provide yield estimates that are closer to the true means (and to validation observations) than will a model chosen by postdictive criteria. This claim is readily subjected to experimental test. In what follows, the relative performances of statistical models chosen

Statistical Analysis of a Yield Trial

AbstractYield trials frequently have both significant main effects and a significant genotype X environment (GE) interaction. Traditional statistical analyses are not always effective with this data structure: the usual analysis of variance (ANOVA), having a merely additive model, identifies the GE interaction as a source but does not analyze it; principal components analysis (PCA), on the other hand is a multiplicative model and hence contains no sources for additive genotype or environment main effects; and linear regression (LR) analysis is able to effectively analyze interaction terms only where the pattern fits a specific regression model. The consequence of fitting inappropriate statistical models to yield trial data is that the interaction may be declared nonsignificant, although a more appropriate analysis would find agronomically important and statistically significant patterns in the interaction. Therefore, agronomists and plant breeders may fail to perceive important interaction effects. This paper compares the above three traditional models with the additive main effects and multiplicative interaction (AMMI) Model, in an analysis of a soybean [Glycine max (L.) Merr.] yield trial. ANOVA fails to detect a significant interaction component, PCA fails to identify and separate the significant genotype and environment main effects, and LR accounts for only a small portion of the interaction sum of squares. On the other hand, AMMI analysis reveals a highly significant interaction component that has clear agronomic meaning. Since ANOVA, PCA, and LR are sub‐cases of the more complete AMMI model, AMMI offers a more appropriate first statistical analysis of yield trials that may have a genotype X environment interaction. AMMI analysis can then be used to diagnose whether or not a specific sub‐case provides a more appropriate analysis. AMMI has no specific experimental design requirements, except for a two‐way data structure.

An economic selection index for sugar cane breeding

For decision-making about varietal release, plant breeders use yield-trial data supplemented by intuitive selection indices based on secondary characters. It is suggested that they could be replaced by explicit economic indices. Profitability is taken as the function to be maximised. An example drawn from sugar cane in Barbados, West Indies shows that secondary characters can indeed reasonably be given weights which turn out to be economically substantial. In principle, similar economic indices could be developed for any crop. All such equations are essentially local in application.

The nature of genotype × environment interactions for wheat yield in Western Australia

The object of this study was to examine genotype × environment interaction (GE) for wheat yield in Western Australia and to consider its breeding implications. One hundred lines were grown in 28 environments in each of 2 years, 1978 and 1979. Linear regression techniques explained only a small fraction of the GE variance. Analyses of variance indicated that years interacted strongly with lines and also with lines × locations. The importance of the year interaction was further highlighted by the multidimensional technique of pattern analysis which revealed a distinct separation of the 1978 and 1979 environments. Environmental groupings with years showed no regional basis other than that associated with successive sowing dates at a location. Groupings of locations were not repeatable between years. Additional Western Australia wheat yield trial data were examined. The major of variation was environmentally determined and interactions of lines with locations, with years, and with location-years were generally highly significant. All results indicate a selection problem for wheat breeders in Western Australia and the necessity for wide testing over several seasons.

Lattice Designs for Unreplicated Yield Trials of Maize Varieties at Several Plant Densities1

The analysis of lattice designs with varieties showing a linear response to levels of some factor confounded with replications is described and illustrated with a hypothetical example. The theory is applied to a 7✕7 quadruple lattice maize (Zea mays L.) variety trial with replications of the lattice planted at different plant densities. A linear regression model is used which assumes that the yield of each variety expressed as log (yield/plant) responds linearly to plant density. In a manner analogous to ordinary lattice designs, corrections for block differences to be applied Jo variety means and regression coefficients are obtained. Two methods of adjusting for variable stands, viz., covariance analysis with deviation from average stand at) a given plant density level as the covariable (Method I) and adjustment of the raw data using preliminary estimates of the variety regression coefficients (Method II), are suggested and compared with a generalized least squares solution (Method III). Method III is the theoretically correct procedure but we consider it impractical for large experiments. Both Methods I and II gave good approximations to the results for Metlfod III.The experiment showed significant differences among variety means and among variety regression coefficients. Information on varietal response to plant density is obtained at the cost of very little loss of precision on variety means. It is suggested that for equivalent varietal replication more comprehensive yield trial data could be obtained by evaluation at more plant density levels with less (or possibly no) replication ot each density.