Search

The maturity group (MG) system is widely used to group soybean [Glycine max (L.) Merr.] varieties based on their growth periods and photothermal responses. However, there is still no universal standard or quantifiable methodology for MG classification. In this study, phenological traits of 107 Chinese, 4 Far East Russian representative soybean varieties, and 113 North American reference varieties covering 13 MGs were evaluated at eight locations (ranging from 30 to 50° N) in four ecoregions of China for two consecutive years (2014 and 2015). Relative maturity groups (RMGs) were attributed to all the varieties based on the linear regression models. To decimalize the RMG values of the early-maturing varieties belonging to MGs below 0, negative values were defined for MGs 00, 000, and 0000. The additive main effects and multiplicative interaction (AMMI) model was used to screen 185 standard candidate varieties for MGs 0000 to VIII. This study provided a systematic and quantifiable methodology for RMG identification in soybeans. The methodology is expected to be widely adopted by soybean regionalization and germplasm exchanges throughout the world and will be helpful for characterizing the photothermal sensitivity and adaptability of the given soybean varieties.

The identification of genotypes presenting wide adaptability and stability is pivotal in breeding programs. To identify such genotypes, it is necessary to use sophisticated analytical tools to establish the genotypes × environments interaction (GEI) pattern across multi-environment trials and select for genotypic stability and adaptability. The aim of the present study was to estimate GEI using Bayesian analysis of Additive Main Effects and Multiplicative Interaction (AMMI) models for both balanced and unbalanced data sets and estimate the predictive ability of model. Two studies were assessed to showcase this approach; in the first, 10 commercial maize (Zea mays) single-cross hybrids and 45 double-cross hybrids were evaluated at 15 different locations. In the second study, 28 hybrids were evaluated in 35 different environments distributed over two different harvest seasons (first and second harvests) with unbalanced data sets within and between harvests. The Bayesian analysis of the AMMI models was robust in dealing with the unbalanced data. This approach is promising for the identification of interaction patterns and the estimation of GEI. The genotypes and environments could be grouped according to their interaction patterns even using the unbalanced data sets, showing that Bayesian analysis of AMMI models could be applied effectively for multi-environment trials. The prediction for missing hybrids was satisfactory in a simulated unbalanced design and captured the GEI and patterns in the data. This allowed the direct comparison of genotypes from the first and second harvests and the estimation of selection gain.

Grain yield data of winter wheat (Triticum aestivum L.) trials in Poland had a four-way factorial design of 24 genotypes by 20 locations by two managements by 3 yr. The experimental design had genotype–management strip plots with two replications for genotypes, with somewhat more genotypes than the 24 having no missing data. The research objectives were to extend additive main effects and multiplicative interactions (AMMI) analysis from two-way to higher-way datasets to reduce spurious complexity originating from noise, delineate wheat mega-environments in Poland, and make genotype recommendations within each mega-environment. Statistical analysis began with adjusting the yield estimates using the strip-plot experimental design and then combining the results in a genotype × location × management × year (GLMY) table. This table was analyzed by a four-way ANOVA mixed model. Next the GLMY dataset was reorganized into a two-way classification, namely a genotype × environment (G × E) dataset, where the 120 environments were defined as combinations of location, management, and year. This two-way dataset was analyzed by AMMI, with practical limitations of working with only a few mega-environments focusing interest on the AMMI1 member of this model family. The first principal component had an evident geographical interpretation, contrasting northeast Poland (colder climate) and southwest Poland (warmer climate). Suitable genotypes were recommended within each of these two mega-environments. The methodological significance of this paper is the extension of AMMI analysis from the customary two-way G × E datasets to higher-way datasets, such as the present four-way GLMY dataset.

Development of dedicated bioenergy crop production systems will require accurate yield estimates, which will be important for determining many of the associated environmental and economic impacts of their production. Shrub willow (Salix spp) is being promoted in areas of the USA and Canada due to its adaption to cool climates and wide genetic diversity available for breeding improvement. Willow breeding in North America is in an early stage, and selection of elite genotypes for commercialization will require testing across broad geographic regions to gain an understanding of how shrub willow interacts with the environment. We analyzed a dataset of first-rotation shrub willow yields of 16 genotypes across 10 trial environments in the USA and Canada for genotype-by-environment interactions using the additive main effects and multiplicative interactions (AMMI) model. Mean genotype yields ranged from 5.22 to 8.58 oven-dry Mg ha−1 yr−1. Analysis of the main effect of genotype showed that one round of breeding improved yields by as much as 20% over check cultivars and that triploid hybrids, most notably Salix viminalis × S. miyabeana, exhibited superior yields. We also found important variability in genotypic response to environments, which suggests specific adaptability could be exploited among 16 genotypes for yield gains. Strong positive correlations were found between environment main effects and AMMI parameters and growing environment temperatures. These findings demonstrate yield improvements are possible in one generation and will be important for developing cultivar recommendations and for future breeding efforts.

The additive main effects and multiplicative interaction (AMMI) model is frequently applied in plant breeding for studying the genotype × environment (G × E) interaction. One of the main difficulties related to this method of analysis is the incorporation of inference to the bilinear terms that compose the biplot representation. This study aimed to incorporate credible intervals for the genotypic and environmental scores in the AMMI model by using an informative prior for the genotype effect. This approach differs from the Bayesian methods that have been presented so far, which assume the same restrictions as the fixed effects model. The method was exemplified by using data from a study with 55 maize hybrids in nine different environments for which variable being studied was the yield of unhusked ears. Our results demonstrated that the credible intervals allowed for the identification of genotypes and environments that did not contribute to the G × E interaction. In addition, it facilitated recognition of homogeneous subgroups of genotypes and environments (with respect to the effect of the interaction) and the adaptability of genotypes to specific environments of great interest to breeders. The posterior distributions of singular vectors were bimodal but with the same density peaks in absolute value. This reflects the arbitrary choice of signs of the main component that was used in different mathematical algorithms. Although our data set was based on unrelated single cross hybrids, the choice of genotypes as random effects enabled the Bayesian AMMI to accommodate the additive and nonadditive relationship matrices. Additionally, the flexibility of the analysis facilitated working with unbalanced data and the incorporation of heterogeneity of variances.

Genotype-by-environment (G × E) interaction (GEI) and quantitative trait locus (QTL)-by-environment interaction (QEI) are common phenomena in multiple-environment trials and represent a major challenge to breeders. The additive main effects and multiplicative interaction (AMMI) model is a widely used tool for the analysis of multiple-environment trials, where the data are represented by a two-way table of G × E means. For complete tables, least squares estimation for the AMMI model is equivalent to fitting an additive two-way ANOVA model for the main effects and applying a singular value decomposition to the interaction residuals, thereby implicitly assuming equal weights for all G × E means. However, multiple-environment data with strong GEI are often also characterized by strong heterogeneous error variation. To improve the performance of the AMMI model in the latter situation, we introduce a generalized estimation scheme, the weighted AMMI or W-AMMI algorithm. This algorithm is useful for studying GEI and QEI. For QEI, the W-AMMI algorithm can be used to create predicted values per environment that are subjected to QTL analysis. We compare the performance of this combined W-AMMI and QTL mapping strategy to direct QTL mapping on G × E means and to QTL mapping on AMMI-predicted values, again with QTL analyses for individual environments. Finally, we compare the W-AMMI QTL mapping strategy, with a multi-environment mixed model QTL mapping approach. Two data sets are used: (i) data from a simulated pepper (Capsicum annuum L.) back cross population using a crop growth model to relate genotypes to phenotypes in a nonlinear way, and (ii) the doubled-haploid Steptoe × Morex barley (Hordeum vulgare L.) population. The QTL analyses on the W-AMMI-predicted values outperformed the QTL analyses on the G × E means and on the AMMI-predicted values, and were very similar to the mixed model QTL mapping approach with regard to the number and location of the true positive QTLs detected, especially for QTLs associated with the interaction and for environments with higher error variance. W-AMMI analysis for GEI and QEI provides an easy-to-use and robust tool with wide applicability.

The objectives of this study were to investigate the pattern of genotype environment interaction and identify megaenvironments (ME), essential test locations, and suitable genotypes for each ME. The genotype plus genotype by environment interaction (GGE) biplot and additive main effects and multiplicative interaction analysis (AMMI) were used to demonstrate the potential of using a graphical biplot to analyze the genotype by environment interaction (GEI) in data of multienvironment trials (MET). These trials of the Central Cooperative Agricultural Research (COODETEC) evaluated 36 advanced breeding lines and 27 check cultivars for 3 years (2008– 2010) at 12 test locations. Yield data were analyzed using the genotype plus GGE and AMMI biplot methods. The test environments were classified into two ME (i.e. ME1: Castro, Gua rapuava, Nao-Me-Toque, Abelardo Luz, and Cachoeira do Sul; and ME2: Campo Mourao, Dourados, Palmital, Palotina, Ponta Pora ,and Rolândia. In ME1, the locations Guarapuava, Cachoeira do Sul, and Abelardo Luz were classified as ideal, while in ME2, Ponta Pora, Dou rados and Palotina were close to ideal. Guarapuava was the essential test location in ME1; however the identification of an essential test location in ME2 was unclear. Superior cultivars and advanced lines were identified in both ME that could be valuable for spring wheat (Triticum aestivum L.) improvement or cultivars released.

The genotype main effects and genotype-by-environment interaction effects (GGE) model and the additive main effects and multiplicative interaction (AMMI) model are two common models for analysis of genotype-by-environment data. These models are frequently used by agronomists, plant breeders, geneticists and statisticians for analysis of multi-environment trials. In such trials, a set of genotypes, for example, crop cultivars, are compared across a range of environments, for example, locations. The GGE and AMMI models use singular value decomposition to partition genotype-by-environment interaction into an ordered sum of multiplicative terms. This article deals with the problem of testing the significance of these multiplicative terms in order to decide how many terms to retain in the final model. We propose parametric bootstrap methods for this problem. Models with fixed main effects, fixed multiplicative terms and random normally distributed errors are considered. Two methods are derived: a full and a simple parametric bootstrap method. These are compared with the alternatives of using approximate F-tests and cross-validation. In a simulation study based on four multi-environment trials, both bootstrap methods performed well with regard to Type I error rate and power. The simple parametric bootstrap method is particularly easy to use, since it only involves repeated sampling of standard normally distributed values. This method is recommended for selecting the number of multiplicative terms in GGE and AMMI models. The proposed methods can also be used for testing components in principal component analysis.

potato (Solanum tuberosum L.) genotypes exhibit significant variation for chip color across different storage regimes. Lines that maintain light chip color after long postharvest storage durations and low storage temperatures are desirable to the potato industry. Since storage regimes vary among growers and processors, lines that exhibit stable chip performance across various storage regimes have a high probability of commercial success. The objective was to test if treating storage regimes as “environments,” analyzing genotype × storage environment interactions, and applying stability statistics can help identify desirable chip processing lines. To examine this, chip color of 47 breeding clones and six standard varieties was evaluated across eight storage environments. Chip color data was analyzed using stability metrics as well as stability-adjusted selection indices. The effects of genotype, storage environment, and genotype × environment on chip color were significant, explaining 47, 24, and 17% of total variance, respectively. Types I, II, and III stable lines were identified through stability analyses. Type I stability was significantly correlated with mean chip color. The most desirable lines were identified under long and cold storage environments. Using Type I stability and stability-adjusted indices, this study identified breeding clones for advancement, including W5840-4, W64845, and W6929-1, which outperformed standard chipping varieties for chip quality and stability. Dep. of Horticulture, Univ. of Wisconsin-Madison, 1575 Linden Drive, Madison, WI 53706. Received 6 Feb. 2013. *Corresponding author (krak@wisc.edu). Abbreviations: AMMI, additive main effect and multiplicative interaction; bi, Eberhart and Russell’s regression coefficient; G×E, genotype × environment; HARS, Hancock Agricultural Research Station; PC, principal component; Pi, superiority of the ith line; RARS, Rhinelander Agricultural Research Station; sdi, line i’s deviation from regression; si, line i’s mean sum of squares of departure from regressions; si, Shukla’s stability variance; YSi, Kang’s selection index. Published in Crop Sci. 53:1944–1952 (2013). doi: 10.2135/cropsci2013.02.0089 © Crop Science Society of America | 5585 Guilford Rd., Madison, WI 53711 USA All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

There has been extensive literature on modeling gene-gene interaction (GGI) and gene-environment interaction (GEI) in case-control studies with limited literature on statistical methods for GGI and GEI in longitudinal cohort studies. We borrow ideas from the classical two-way analysis of variance literature to address the issue of robust modeling of interactions in repeated-measures studies. While classical interaction models proposed by Tukey and Mandel have interaction structures as a function of main effects, a newer class of models, additive main effects and multiplicative interaction (AMMI) models, do not have similar restrictive assumptions on the interaction structure. AMMI entails a singular value decomposition of the cell residual matrix after fitting the additive main effects and has been shown to perform well across various interaction structures. We consider these models for testing GGI and GEI from two perspectives: likelihood ratio test based on cell means and a regression-based approach using individual observations. Simulation results indicate that both approaches for AMMI models lead to valid tests in terms of maintaining the type I error rate, with the regression approach having better power properties. The performance of these models was evaluated across different interaction structures and 12 common epistasis patterns. In summary, AMMI model is robust with respect to misspecified interaction structure and is a useful screening tool for interaction even in the absence of main effects. We use the proposed methods to examine the interplay between the hemochromatosis gene and cumulative lead exposure on pulse pressure in the Normative Aging Study.

The new stem rust race Ug99 is an alarming threat to the wheat (Triticum aestivum L.) production of the world for the very fact that it overcomes many of the known Sr genes including the most commonly used genes such as Sr31, Sr24, and Sr36. The wheat breeding program at the International Center for Agricultural Research in the Dry Areas (ICARDA) has distributed the first stem rust resistant spring wheat yield trials composed of 24 wheat genotypes for the Central and West Asia and North Africa (CWANA) region to identify high yielding wheat genotypes, which can be recommended for direct release and/or parentage purposes by the National Agricultural Research Systems (NARSs). The additive main effects and multiplicative interaction (AMMI) analysis of variance for grain yield (kg ha-1) of 24 wheat genotypes tested in 15 environments showed that wheat grain yield was significantly (p < 0.01) affected by environments (E), genotypes (G), and genotype × environment interaction (GEI). From the total sum of squares due to treatments (G + E + GEI), E contributed 90.1% and GEI accounted for 8.3% while G effect represented only 1.6%. According to AMMI2 estimate, Baasha-14 and Amir-2 with mean grain yield levels of 5168 and 5127 kg ha-1, respectively, are identified as the best genotypes with high yield potential across a wide range of environments. Based on correlation and clustering analysis among environments, the Ghab site in Syria (E7) is identified as the highest yielding and best correlated site and therefore it is recommended as a key location for evaluation of wheat germplasm yield potential in ICARDA's wheat breeding program. We recommend the high yielding and Ug99 resistant genotypes identified in this study for direct release as replacement of the susceptible varieties by the respective NARS and/or to be used as parents at ICARDA and by NARS collaborators.

Summary The human oral microbiome is comprised of approximately 800 different bacterial species many of which are as yet uncultivated. Their dynamics and variability in relation to health and disease are still poorly understood. Here we tested the hypothesis that the emergence of stress-induced periodontal diseases is predictable based on the composition of the initial microbiota. As a model, we analysed 58 individuals performing a challenging expedition (exposure to various stress-factors due to changes in diet, hygiene, temperature, physical and mental stress) in remote regions of the Himalayans (Annapurna Himal). Plaque samples were taken at start (Bhulbule) and destination (3000 meter difference in altitude) seven days later (Manang). Twenty-eight individuals remained symptom-free (Group I) while 30 participants developed periodontal problems, mostly gingivitis (Group II). The microbiota was monitored via T-RFLP-analysis of amplified 16S rRNA genes directly from the plaque samples. Based on the Additive-Main-Effects-Multiplicative-Interactions-model (AMMI) using the T-Rex software variation from T-RF main effects was at least 95%, indicating that most variation was due to inherent differences in microbial communities among individuals. However, an interaction signal up to 3% was consistently observed between groups I and II but not between the two time points of sampling regardless of selected analytical parameters. The data, supported by heterogeneity, diversity and similarity indices indicated marked differences between groups I and II already prior the onset of clinical symptoms. These differences may provide the basis for using ecological parameters of oral microbial communities as early diagnostic marker for the onset of oral disorders and infections.

Genetic improvement of crop performance is essential to increase yield. The aims of this paper are to present combined additive main effects and multiplicative interaction (AMMI) analysis and cluster analysis to winter wheat (Triticum aestivum L.) grain-yield data collected in preregistration trials performed over many years in an incomplete three-way genotype x location x year (GLY) classification and to demonstrate the usefulness of this procedure for studying the patterns of cultivar adaptation. The 21 genotypes were divided into three groups that were homogeneous in their pattern of genotype x location (GL) interactions. These groups of genotypes were further divided into subgroups with homogeneous genotype means. In this study 11 genotypes that were grouped in three subgroups had wide adaptability. Among them the cultivar TONACJA was identified as a good check cultivar for preregistration trials. One cultivar had narrow adaptation and nine were relatively poorly adapted. The combined AMMI and cluster analysis method proved effective for describing diverse patterns of yield response to agro-ecological conditions in locations spread across most of the Polish winter wheat growing areas.

Development of cultivars resistant to sheath blight (SB) disease caused by Rhizoctonia solani Kuhn is an important breeding objective for the southern U.S. rice (Oryza sativa L.) industry. The aim of this research was to explore performance, genotype (G) and genotype x environment (GE) interactions, and stability for SB resistance of 322 doubled-haploid (DH) lines of the SB2 mapping population produced by the RiceCAP project. Analysis was performed by heritability-adjusted genotype plus genotype x environment interaction (HA-GGE) biplot analysis, Huhn's nonparametric stability statistics, median polish, and additive main effects and multiplicative interaction (AMMI). Genetic material was evaluated in replicated plots for 2 yr (2006 and 2007) in Louisiana and Arkansas. A single mega-environment was identified consisting of the four year-location combinations. Heritability-adjusted genotype plus genotype x environment interaction biplot analyses identified 11 DH lines with high and stable levels of resistance at both locations. Five susceptible DH lines were also identified with greater stability than the susceptible parent used to develop the SB2 population. Results from Huhn's nonparametric approach and median polish complemented those of HA-GGE biplot analysis. Material identified in this study represents a potential source of high and stable SB resistance for cultivar development that warrants additional field-plot evaluation.

Previous studies using turfgrass quality (TQ) data from the National Turfgrass Evaluation Program (NTEP) used cross validation with additive main effects and multiplicative interactions (AMMI) models to demonstrate accuracy gain within a single trial. The objective of this study is to test AMMI predictive accuracy for new years or new years × locations. Five winners recommended by the raw data AMMIF and five winners recommended by the more accurate AMMI5 model were selected, resulting in a 10-entry roster at six validation sites for evaluation over 5 yr. Average TQ for AMMI5 selections was 5.57 and was significantly greater than AMMIF selections averaging only 5.31. In five of six validation locations, higher TQ was observed by planting AMMI5 selections over AMMIF. The correlation between the 60 validation observations and AMMI5 predictions was 0.628 but for AMMIF was 0.479. Better predictive success indicated by a significantly lower mean squared deviation (MSD) was observed with AMMI5 (0.453) than AMMIF (0.608). The AMMI5 model predicted winners were five times as effective as AMMIF in predicting observed winners from the top group (rank one and two). There was decisive evidence of a parsimonious AMMI model increasing predictive success across years despite a meager 1.24 statistical efficiency with AMMI5.

Two new strategies are proposed to improve the detection and understanding of quantitative trait loci (QTL), especially those exhibiting QTL × environment interactions (QEI), in the context of experiments conducted in multiple environments. First, a parsimonious Additive Main effects and Multiplicative Interaction (AMMI) model is applied to the phenotypic data to gain accuracy and thereby to increase the logarithm of odds (LOD) scores for QTL detections. Second, the environments are ordered by AMMI parameters that summarize genotype × environment interaction information to reveal consistent patterns and systematic trends that often have an evident ecological or biological interpretation. The combination of greater accuracy for the phenotypic data and systematic trends for the environments provides for more consistent and understandable QTL results. These new strategies are illustrated with two examples: preharvest sprouting scores of a biparental wheat (Triticum aestivum L.) population from 14 environments spread over 5 yr, and yield for a doubled-haploid barley (Hordeum vulgare L.) population tested in 16 environments. AMMI parameters can also provide successful predictions of entire QTL scans for new environments. The statistical methods developed here are of great generality, applicable across microbial and plant populations grown in multiple environments, and may be adapted to animal and human genetic studies.

Bacterial wilt of forage grasses, caused by the pathogen Xanthomonas translucens pv. graminis (Xtg), is a major disease of forage grasses such as Italian ryegrass (Lolium multiflorum). The plant genotype-bacterial isolate interaction was analysed to elucidate the existence of race-specific responses and to assist the identification of plant disease resistance genes. In a greenhouse experiment, 62 selected plant genotypes were artificially inoculated with six different bacterial isolates. Significant differences in resistance were observed among L. multiflorum genotypes (P 0·05) could be observed using linear regression modelling. However, additive main effects and multiplicative interaction effects (ammi) analysis revealed five genotypes which did not cluster close to the origin of the biplot, indicating specific interactions between these genotypes and some bacterial isolates. Simple sequence repeat (SSR) markers were used to identify marker-resistance associations using the same plant genotypes and bacterial isolates. The SSR marker NFA027 located on linkage group (LG) 5 was significantly associated with bacterial wilt resistance across all six bacterial isolates and explained up to 37·4% of the total variance of AUDPC values. Neither the inoculation experiment nor the SSR analyses revealed major host genotype-pathogen isolate interactions, thus suggesting that Xtg resistance, observed so far, is effective across a broad range of different bacterial isolates and plant genotypes.

Rapeseed ( Brassica napus L.) is regarded as one of the most important oilseed crops worldwide. An experiment was conducted to quantitatively examine the genetic parameters of oil yield, seed yield, and yield components for rapeseed by using a half diallel of nine rapeseed cultivars. The 36 F1 hybrids, their parents, and four other commercial varieties were planted in breeding nurseries in 2008 and 2009. During the growing season, the number of pods per plant (NPP), number of seeds per pod (NSP), pod length (PL), 1000 seed weight (SW), harvest index (HI), seed yield (SY), oil percent (OP), and oil yield (OY) were measured. The data were analyzed with the additive main effects and multiplicative interaction (AMMI) model and its biplots. The first two principal components (IPC1 and IPC2) were used to create a two-dimensional biplot. IPC1 accounted for 28.7% (SW in 2008) to 48.9% (OP in 2009) of the sum of squares of the male by female (M × F) interaction, while IPC2 accounted for 18.3% (PL in 2009) to 31.6% (number of seeds per pod in 2008) of the sum of squares of the M × F interaction. Two heterotic groups were identified for OY, and six between-group crosses, that is, [Talaye, Modena, Opera] × [Fornax, Orient], would provide maximum heterosis for this trait. The AMMI analysis with the half diallel method identified not only the general combining ability (GCA) and specific combining ability (SCA) effects but also the F1 crosses whose offspring may display heterosis in their offspring. The results can be used to design efficient breeding strategies in rapeseed.

Plant breeders aim to select genotypes with stable phenotypes across environments. Multienvironment trials (METs) including 3 yr and six locations for 17 genotypes of chickpea (Cicerarietinum L.) were performed in Iran. Bartlett's test of homogeneity of variances was not significant and so the combined analysis was done. This analysis indicated that the environment (E), genotype (G), and the genotype by environment (GE) interaction were significant. Thus, stability analysis was necessary, and so additive main effects and multiplicative interaction (AMMI) stability parameters and some regression methods were used for a stability analysis. According to regression method, the high-yielding genotype, G8, was more responsive (b i > 1) to improved environmental conditions than the other genotypes. Stability parameters of Tai indicated that genotypes G3 and G17 and stability statistics of AMMI analyses showed that G2 and G3 were the most stable genotypes. The first and the second factors (F1 and F2) of factor analysis explained 73.84% total variance of stability parameters, and F1 separated the D i 2 and α i from the other parameters. The F2 separated static and dynamic concepts of stability, in which the mean yield, regression coefficient, EVF, SIPCF, and AMGEF parameters had a dynamic concept of stability, and the other remaining parameters had static concept of stability. Finally, according to all stability statistics and regarding the high mean yield, genotype G8 (Flip 93-93) was recommended for national release in Iran, as it adequately demonstrated wide adaptation across environments.

Genotype x environment (GE) interactions are important sources of variation in crop breeding programs. The objectives of this study were (i) to analyze GE interactions on grain yield of one bread wheat (Triticum aestivum L.) and 19 durum wheat [T. turgidum L. subsp. durum (Desf.) Husn.] genotypes by the additive main effects and multiplicative interaction (AMMI) model, and (ii) to evaluate genotype (G), environment (E), and GE interactions with the statistical parameters of AMMI statistic coefficient (D), AMMI stability value (ASV), regression coefficient (b) and ecovalence (W 2 ). Main effects due to E, G, and GE interaction as well as first three interaction principal component axes (IPCA 1 to 3) were found to be significant (P < 0.01). The AMMI biplots distinguished genotypes with wide and specific adaptation and environments with high and low genotype discrimination ability. Ranking the statistical parameters indicated that they are similar in general, especially for the genotypes with similar responses and also for the environments with similar discrimination ability. The results show that the statistical parameters as well as the AMMI model are useful for analyzing GE interaction. Based on the AMMI model, the recommended genotypes had more grain yield improvement in favorable than unfavorable environments. The genotypes selected based on stability of grain yield combined high stability and high-yielding performance.

Alternaria leaf petiole and stem blight is an economically important disease of sweet potato (Ipomoea batatus L.) in tropical and sub-tropical environments. Published research on cultivar resistance to the sweet potato disease is limited. To evaluate cultivar reaction and stability to the disease, multi-location and replicated experiments were established in 12 environments in Uganda. Disease severity (area under disease progress curves - AUDPC), and cultivar root yield were also assessed. Significant differences (P < 0.001) in AUDPC were detected among cultivars. Mean AUDPC ranged from 46.3 (Araka Red) to 78.4 (New Kawogo) across locations and seasons and the genotypes Araka Red and Tanzania had the lowest disease values. The location and season effects accounted for 67.1% and 7.5% of the total variance of AUDPC recorded among cultivars. The ranking of cultivars based on predicted AUDPC from Additive Main Effect and Multiplicative Interactive model (AMMI) showed that the NASPOT 1, the susceptible check, and New Kawogo were most susceptible to the disease in 11 of the 12 environments. Low and stable disease was consistently recorded and predicted on NASPOT 3 and the landrace cultivars Tanzania, Dimbuca, and Araka Red across environments. These results suggest that landrace cultivars had relative stability to the disease and wide adaptation across environments. These results suggest that AMMI statistical model and other multivariate techniques can be utilized for prediction of Alternaria disease stability in these locations.

Information on most-adapted species is required to increase the economic sustainability and the utilization of cool-season feed legumes in southern Europe. This study aimed to compare pea (Pisum sativum L.), faba bean (Vicia faba L.), narrow-leafed lupin (Lupinus angustifolius L.), and white lupin (L. albus L.) across Italian environments which contrasted for geoclimatic area (subcontinental or Mediterranean) and sowing season (autumn or late-winter). Each species was represented by two cultivars tending to different geoclimatic adaptation. Data Set 1 included eight cultivars and five environments; Data Set 2 comprised two pea and two faba bean cultivars, and 10 environments. On average, pea exhibited higher grain yield than any species in subcontinental environments and lupin species in Mediterranean environments. Crossover genotype x environment interaction (GEI) occurred between species in Data Set 1 and between cultivars within species in both data sets (P < 0.05). Additive main effects and multiplicative interaction (AMMI) analysis of grain yield data indicated that species and cultivar adaptation patterns were more influenced by the climatic area than by location, sowing time, or test year within climatic area, highlighting the interest of comparing species on the basis of locally-adapted cultivars. Entry grain yield was associated with higher aerial biomass across subcontinental environments, and earlier phenology and higher harvest index across Mediterranean environments. In the species comparison based on locally top-performing cultivars, white lupin outyielded other species for crude protein in both climatic areas, whereas pea outperformed others across subcontinental environments and did not differ from faba bean or white lupin across Mediterranean environments for production of milk feed units (P < 0.05).

A population of 164 recombinant inbred lines (RILs) of rice (Oryza sativa L.) derived from a cross between Milyang23 and Gihobyeo was evaluated for nine phenotypic characters over three years and two regions in Korea. The population had been previously mapped using 414 molecular markers. Genotype x environment (G x E) interaction was analyzed for six grain yield-related traits and three agronomic traits across years and locations using the AMMI model. The quantitative trait loci (QTLs) were detected by interval mapping and composite interval mapping. A total of 75 QTLs were identified for the nine traits across five environments and they were categorized as (i) 29 QTLs with main effect, (ii) 18 QTLs with minor effect, (iii) 13 QTLs with G x E interaction effect, (iv) six QTLs with both main effect and G x E interaction effect, and (v) nine potential QTLs with low log of the odds (LOD) scores. The AMMI model explained from 68.6% to 84.7% of the interaction effect and 19 QTLs were significantly associated with G x E interaction. Culm length had the least G x E, while the maximum G x E interaction was exhibited for spikelets per panicle (39.7%) and percent ripened grain (35.3%). Markers closely linked to main effect QTLs will be most useful for marker-assisted breeding.

The use of genotype main effect (G) plus genotype-by-environment (GE) interaction (G+GE) biplot analysis by plant breeders and other agricultural researchers has increased dramatically during the past 5 yr for analyzing multi-environment trial (MET) data. Recently, however, its legitimacy was questioned by a proponent of Additive Main Effect and Multiplicative Interaction (AMMI) analysis. The objectives of this review are: (i) to compare GGE biplot analysis and AMMI analysis on three aspects of genotype-by-environment data (GED) analysis, namely mega-environment analysis, genotype evaluation, and test-environment evaluation; (ii) to discuss whether G and GE should be combined or separated in these three aspects of GED analysis; and (iii) to discuss the role and importance of model diagnosis in biplot analysis of GED. Our main conclusions are: (i) both GGE biplot analysis and AMMI analysis combine rather than separate G and GE in mega-environment analysis and genotype evaluation, (ii) the GGE biplot is superior to the AMMI1 graph in mega-environment analysis and genotype evaluation because it explains more G+GE and has the inner-product property of the biplot, (iii) the discriminating power vs. representativeness view of the GGE biplot is effective in evaluating test environments, which is not possible in AMMI analysis, and (iv) model diagnosis for each dataset is useful, but accuracy gain from model diagnosis should not be overstated.

Nineteen white yam (Dioscorea rotundata Poir.) genotypes were grown in four agroecological zones in Nigeria for 2 years to study their responses to yam virus diseases, and to investigate the influence of genotype x environment (G x E) interactions on their reactions to virus infection, using the Additive Main Effects and Multiplicative Interactions (AMMI) Statistical Model. Environments, genotypes and G x E interactions were all highly significant (P < 0.0001) for the virus diseases. The G x E interactions accounted for 21.4% of the treatment sum of squares, and influenced the relative ranking of the genotypes across the eight environments. Genotypes TDr 93-32 and TDr 93-27 were identified as the most stable genotypes, while TDr 1621 and TDr 1640 were shown to have resistance to virus diseases. The study also identified Jos with high yam virus disease severity as the most appropriate site for the screening of D. rotundata genotypes.

The Additive Main effects and Multiplicative Interaction (AMMI) model, Genotype main effects and Genotype X Environment interaction (GGE) model, and Principal Components Analysis (PCA) are singular value decomposition (SVD) based statistical analyses often applied to yield-trial data. This paper presents a systematic comparison, using both statistical theory and empirical investigations, while considering both current practices and best practices. Agricultural researchers using these analyses face two inevitable choices. First is the choice of a model for visualizing data. AMMI is decidedly superior, not for statistical reasons, but rather for agricultural reasons. AMMI partitions the overall variation into genotype main effects, environment main effects, and genotype X environment interactions. These three sources of variation present agricultural researchers with different challenges and opportunities, so it is best to handle them separately, while still considering all three in an integrated manner. Second is the choice of a member of a given model family for gaining predictive accuracy. AMMI, GGE, and other SVD-based model families are essentially equivalent, but best practices require model diagnosis for each individual dataset to determine which member is most predictively accurate. Making these two choices well allows researchers to extract more usable information from their data, thereby increasing efficiency and accelerating progress.

An understanding of the characteristics of crop varieties and advanced lines could help improve their cultivation and to further enhance their potential. The objectives of this study were to estimate the genotype (G), environment (E) and genotype x environment (GE) interactions on the grain yield of Chinese spring wheat genotypes in 2000 and 2001 by the additive main effects and multiplicative interaction (AMMI) model, and to evaluate the relationships between yield and its components by correlation and path analysis. Grain yield varied from 3.9 to 5.2 t ha -1 , among which SW8188 had the highest yield performance, followed by 58769-6 and Chuannong 16. Three interaction principal components (IPC) accounted for a total of 79.99 % and 72.96 % of the interactions with 41.05 % and 52.08 % for the corresponding degrees of freedom in 2000 and 2001, respectively. When IPC3 was significant, the stability coefficient D i was more useful in the evaluation of the stability of each genotype. The estimates of D i in the 2 years indicated that the D i values varied between genotypes and years. The D i values ranged from 1.804 to 14.665 and 2.497 to 12.481 in 2000 and 2001 respectively. The suitable locations (environments) for all genotypes were characterized. These results would be useful for improving the Chinese spring wheat cultivation and improvement.

Sweetpotato [Ipomoea batatas (L.) Lam.] is cultivated across a wide range of agrogeographical conditions. The objectives of this study were to analyze genotype X environment (G x E) interactions for sweetpotato yield (i.e., storage root yield, biomass, harvest index) and nutritional traits [i.e., root dry matter (RDM), starch (STA), and β-carotene (BCR) content; and leaf carotene (BCL) and chlorophyll (CHL) content] in multienvironmental trials (MET) across ecogeographic regions. Nine clones of diverse origins were tested and compared with check clones at seven locations in Peru using two N treatments (N = 0 or 80 kg ha -1 ). The G X E analysis was conducted with regression, additive main effects and multiplicative interaction (AMMI), and cluster analyses. The G x E interactions were smaller than the genetic variation of nutritional traits. The G X E interactions were larger or nearly equal to the genetic variation of yield traits (except harvest index), and were mainly determined by subsets of genotypes and environments. The contribution of N input to G x E was often not significant. Genotypes were observed with wide adaptation and high yields (about 19 to 22 Mg ha -1 ) across all three environmental groups that were derived from the cluster analysis. However, a specifically adapted genotype was observed with considerable yield advantage over all widely-adapted genotypes in low-yielding environments (from 9 to 18 Mg ha -1 ). Locations differed in their selection ability for storage root yield. We concluded that it is possible to breed for high yield and wide adaptability in sweetpotato in Peru, and it can be ensured that low-yielding or marginal environments are not neglected in breeding efforts.

ses were not always effective in analyzing the MET data structure. The ANOVA is an additive model that The identification of the highest yielding cultivar for a specific describes main effects effectively and determines if GE environmentonthebasisofbothgenotype(G)andgenotypeenvironment (GE) interaction would be useful to breeders and producers since interaction is a significant source of variation, but it yield estimates based only on G and environment (E) effects are does not provide insight into the patterns of genotypes insufficient. The objective of this study was to demonstrate the useful- or environments that give rise to the interaction. The ness of additive main effects and multiplicative interactions (AMMI) PCA is a multiplicative model that contains no sources model analysis and G plus GE interaction (GGE) biplots, obtained of variation for additive G or E main effects and does from sites regression (SREG) model analysis in interpreting GE grain not analyze the interactions effectively. The linear reyield data. Replicated grain yield data of six rice (Oryza sativa L.) gression method uses E means, which are frequently a cultivars (Cocodrie, Cypress, Jefferson, Lemont, Saber, and Wells) poor estimate of environments, such that the fitted lines from three main cropping seasons (2000, 2001, and 2002) at four locain most cases account for a small fraction of the total tions in Texas, USA (Bay City, Eagle Lake, Ganado, and Beaumont) GE (Zobel et al., 1988). The AMMI model analysis were obtained and used for this purpose. Through AMMI model analysis, the magnitude and significance of the effects of GE interaction and combines the ANOVA (with additive parameters) and its interaction principal components relative to the effects of G and PCA (with multiplicative parameters) into a single analE were estimated. The stability and adaptability of specific cultivars ysis. The AMMI model analysis is useful in making wereassessedbyplottingtheirnominalgrainyieldsatspecificenviron- cultivar recommendations, specifically by megaenvironments in an AMMI biplot, which aided in the identification of mega- ment analysis, in which the best performing cultivar for environments (environments with the same highest yielding cultivar). each subregion of the crop’s growing region is identified Appropriate check cultivars for all locations or for specific locations (Zobel et al., 1988; Gauch and Zobel, 1997). Gauch and were identified. Through GGE biplots of SREG model analysis re- Zobel (1997) demonstrated the usefulness of AMMI sults, the relative yield performance of cultivars at a specific environ- analysis in supporting breeding program decisions, such ment were illustrated, the performance of a cultivar at different envias in the selection of environments or test site locations. ronments was compared, the performance of two cultivars at different environments were compared, the highest yielding cultivars at the dif- Although AMMI model analysis results are based only ferent megaenvironments were identified, and ideal cultivars and test on yield data (not environmental data), Ebdon and locations were identified. Gauch (2002a) reported that AMMI environmental (interaction) statistics were correlated with environmental factors, such as precipitation, mean daily maximum and

Producing specific varieties for each subregion of a target region, instead of widely adapted varieties, may exploit positive genotype × location (GL) interactions to increase crop yields. With reference to the Algerian durum wheat (Triticum durum Desf.) region, as defined in a Geographic Information System (GIS), this study aimed at (i) comparing additive main effects and multiplicative interaction (AMMI) vs. joint regression modeling of GL effects, (ii) verifying the reliability of a GIS-based definition of two subregions that extended the site classification on the basis of GL effects as a function of long-term winter mean temperature in the GIS, and (iii) comparing wide vs. specific adaptation in terms of observed and predicted yield gains (the latter for scenarios of defined and undefined selection locations). Twenty-four cultivars from international centers, Europe, and North Africa were evaluated across 3 yr in a total of 47 environments by randomized complete block designs with four replications per trial. The AMMI modeling was distinctly superior to joint regression. The AMMI + cluster analysis and pattern analysis classified test locations consistently and in good agreement with the GIS-based subregion definition. Under the hypothesis of six selection environments assigned to subregions in proportion to their size (three sites in each of two years) for late stage selection, specific adaptation provided 2 to 7% greater gains than wide adaptation over the region at similar costs. The advantage of specific adaptation was much larger (39% determined on the basis of observed gains) for the smaller, stressful inland subregion, where specific adaptation may also enhance food security. Performing earlier specific selection, or adopting distinct genetic bases for each subregion, may further increase the advantage of specific breeding.

Cassava is an important food and cash crop in Malawi. It is also becoming increasingly important for industrial use. The aim of this study was to investigate the native starch quality of different Malawi cassava genotypes. Trials were conducted at Chitedze and Makoka in Malawi in the 2000/01 season. Apart from root dry matter and starch extraction, starch quality parameters considered included protein, moisture and ash contents, pH and whiteness. Various stability measures were used to deal with the problem of genotype × environment interaction. The results showed that all the cassava genotypes produced starch with no protein and with colour as white as required by the industry. Moisture and ash contents and pH fell within the industry-recommended ranges. This suggests that native cassava starch is suitable for use in various industries. Additive main effects and multiplicative interaction (AMMI) was strongly correlated with other measured stability parameters and is therefore recommended for stability analysis of starch quality parameters. Genotype had a larger influence than environment on root dry matter. This agrees with the hypothesis that one or a few major genes control root dry matter in cassava. Copyright © 2004 Society of Chemical Industry

A shrinkage estimation method, herein named the CCC method, for multiplicative models has been shown to improve predictions of cell means in multienvironment cultivar trials; however, better estimations of shrinkage factors are needed. Objectives of this study were (i) to develop shrinkage factors for AMMI (Additive Main effect and Multiplicative Interaction) multiplicative terms based on the eigenvalue partition (EVP) method and (ii) to compare AMMI models fitted by EVP and CCC shrinkage methods, unshrunken AMMI models chosen by cross validation, and best linear unbiased prediction (BLUP) based on a fixed main effects and random genotype × environment interaction (GEI) model. The prediction methods were applied to three multienvironment cultivar trials, and also to simulated data. The four alternative prediction methods were compared by cross validation with the root mean squared predictive difference (RMSPD). According to the RMSPD cross validation criterion, the EVP shrinkage estimation method produced the most accurate predictions in almost all cases. When error variance was of a small to moderately large magnitude, results for EVP and CCC methods were in close agreement, CCC being slightly worse; EVP became better as error variance became very large, probably because of the unreliability of Gollobs degrees of freedom as a measure of error variance absorption by the principal components (PC) when data are extremely noisy. Shrunken EVP models were generally more predictively accurate than truncated least squares-fitted AMMI models and BLUPs, which was also true for CCC except when error variance was large. The EVP shrinkage method appears to be promising for obtaining improved predictions of cultivar performance in multienvironment cultivar trials.

Breeders need more information on selecting testers to identify lines for formation of synthetics and need more user-friendly methods to study general combining ability (GCA) and specific combining ability (SCA) of genotypes. Forty-three lines were crossed to two narrow-based (S3 line and S3 x S3 hybrid) and two broad-based testers, which were open-pollinated varieties (OPVs). Topcrosses (line x tester, L x T) were evaluated in eight acid soil environments. The additive main effect and multiplicative interaction (AMMI) and the site regression (SREG) models were used to study the GCA of the testers and lines and the SCA of the L x T interaction. The SREG biplot contains the effect of the lines plus the L x T interaction and displays both GCA and SCA, whereas the AMMI biplot contains the effect of the L × T interaction and displays only the SCA. One synthetic was developed by recombining six lines identified as superior by each tester. The four synthetics were evaluated in 12 acid and nine nonacid soil environments. The synthetic developed with the S3 line as tester yielded the highest and the one developed with an OPV as tester yielded the lowest. The AMMI and SREG models seem to provide an effective tool to visualize and study GCA and SCA of genotypes.

The additive main effects and multiplicative interaction (AMMI) model has been proposed for the analysis of genotype-environmental data. For plant breeding, the recovery of pattern might be considered to be the principal objective of analysis. However, some problems still remain with the analysis, notably in selecting the number of multiplicative components in the model. Methods based on distributional assumptions do not have a sound methodological basis, while existing data-based approaches do not optimize the cross-validation process. This paper first summarizes the AMMI model and outlines the available methodology for selecting the number of multiplicative components to include in it. Then two new methods are described that are based on a full leave-one-out procedure optimizing the cross-validation process. Both methods are illustrated and compared on some unstructured multivariate data. Finally, their applications to analysis of genotype x environment interaction (GEI) are demonstrated on experimental grain yield data. Conclusions of the study are that the leave-one-out procedure is preferable in practice to either distributional F-test or cross-validation randomization methods, and of the two leave-one-out procedures the Eastment-Krzanowski method exhibits the greater parsimony and stability.

Genotype X environment interaction is a commonly observed phenomenon in experiments in plant breeding and genetics. This interaction can be modeled by different statistical models which can include covariates, such as in biadditive factorial regression, or in other ways, such as in the classical additive main effect and multiplicative interaction (AMMI) model. The aim of this paper was to explain genotype x environment interaction in multienvironment trials of winter wheat (Triticum aestivum L.) by environmental variates to assess the genotype sensitivities to the environmental conditions and to compare the results from biadditive factorial regression (BIAREG) and AMMI. Data consisted of 13 lines grown in France in 14 environments (combinations of two years, four locations and two treatments). Grain yield and heading date were measured and the environments were characterized by climatic data (water deficit, radiation, temperature above 25°C) and by observations of powdery mildew infection (caused by Erysiphe graminis DC F. sp. tritici), lodging, and nitrogen status. AMMI model explained most of the genotype × environment interaction (77.4%) but did not provide a direct biological explanation. BIAREG explained a slightly smaller part (74.0%) but with fewer degrees of freedom. The contributions of the environmental variates to the synthetic variates revealed two important subsets of initial covariates-hiotic variates and nitrogen status in contrast to climatic ones-associated with the interaction. The interactive pattern of the genotypes and the environments was similar for both models. BIAREG is more powerful than AMMI because it provides a description of the sensitivities of the genotypes in regard to the observed environmental covariates. When environments can be characterized by variates, we suggest that BIAREG can complement AMMI because it provides direct biological explanation of the interaction.

Nine cassava genotypes were grown at six representative sites in Nigeria for 3 years to study their response to cassava mosaic disease (CMD), investigate the influence of genotype x environment (G x E) interactions on their reactions to the disease, and identify genotypes with stability to the disease, using the Additive Main Effects and Multiplicative Interaction statistical model. Environments, genotypes and G x E interactions were highly significant (P < 0.01) for the disease. The G x E interactions accounted for 19.5% of the treatment sums of squares for CMD and influenced the relative ranking of genotypes across environments. The magnitude of the G x E interaction effect for CMD was larger than that of genotypes. Examination of the G x E interaction structure revealed specific areas where screening of cassava genotypes for resistance to CMD could be performed best. The study identified genotypes such as TMS 30001 and 63397 with resistance to CMD and CMD-stable clone U/41044, which could be distributed to growers, and sites such as Ibadan and Ubiaja with high CMD severity for screening genotypes for reaction to CMD.

Summary Forty wildtype isolates of Sphaeropsis sapinea were grouped into the morphotypes A and B based on previously defined differences in cultural and morphological criteria as well as restriction sites for Dde I and Bst UI endonucleases in nuclear ribosomal DNA amplicons. Thirteen of 20 type A isolates and nine of 20 type B isolates contained detectable dsRNA (55%) of different molecular weight and size. dsRNA was transmitted into conidia at a frequency of 71–100%. By selecting single conidia, dsRNA-free subcultures were obtained from six of 22 isolates containing dsRNA. Pathogenicity tests on expanding buds of landscape trees of three species of Pinus showed highly significant statistical interactions between isolate virulence, Pinus species, and year. Pine species-year had a profound impact on virulence. The pattern in the interactions was revealed by principal component analysis of the interaction sums of squares of the anova (Additive Main Effects and Multiplicative Interaction; AMMI). Pinus sylvestris was highly interactive in its susceptibility to S. sapinea with seasonal effects. P. nigra and P. resinosa were more stable. The interactivity analysis was used to apportion interaction to specific isolates to improve the accuracy of the estimates of virulence. Estimates of the relative virulence of isolates were predicted over five different Pinus species-years. Isolates were ranked in virulence and interactivity using the AMMI model. This model permitted mean separation tests of the relative virulence among isolates over the combined Pinus species-years. One isolate was identified as potentially having dsRNA-mediated hypovirulence based on the significantly greater virulence of its isogenic, dsRNA-free subculture, as expressed over the three Pinus species and 2 years. Type A isolates containing dsRNA ranged from stable to highly interactive and from low to high in virulence. Type B isolates containing dsRNA were similar in interactivity but virulence ranged from avirulent to moderate, seldom exceeding the mean for S. sapinea. dsRNA-free isogenic subcultures tended not to express higher virulence than their dsRNA-containing parent strains but often changed in interactivity. Therefore, in one year a dsRNA-free subculture might be more virulent than its dsRNA-containing parent. In another year the dsRNA-free subculture might be less virulent. Resume Quarante isolats de type sauvage de Sphaeropsis sapinea ont ete groupes en morphotypes A et B d'apres les differences anterieurement definies pour les criteres culturaux et morphologiques ainsi que pour les sites de restriction des endonucleases DdeI et BstUI dans les produits d'amplification de l'ADN ribosomique nucleaire. Treize des 20 isolats de type A et neuf des 20 isolats de type B (55%), contenaient un ARN bicatenaire detectable de taille et de poids moleculaire differents. L'ARN bicatenaire etait transmis aux conidies a la frequence de 71-100%. Des sub-cultures mono-conidiennes indemnes d'ARN bicatenaire ont ete obtenues a partir de six des 22 isolats contenant de l'ARN bicatenaire. Des tests de pouvoir pathogene en nature sur des bourgeons en expansion de trois especes de Pinus, ont montre des interactions hautement significatives entre la virulence des isolats, l'espece de pin et l'annee. La parametre espece-annee avait un fort impact sur la virulence. La structure des interactions a ete mise en evidence par une analyse AMMI (Additive Main Effects and Multiplicative Interaction). La sensibilite de Pinus sylvestrisaS. sapineaetait en forte interaction avec les effets saisonniers. Celles de Pinus nigra et P. resinosaetaient plus stables. L'analyse de l'interactivite a ete utilisee pour assigner l'interaction a des isolats particuliers afin d'ameliorer la precision de l'estimation de la virulence. La virulence relative estimee des isolats a ete predite pour cinq couples espece-annee differents. Les isolats ont ete classes pour la virulence et l'interactivite en utilisant le modele AMMI. Ce modele a permi des tests de separation de moyenne de la virulence relative pour les combinaisons espece-annee. Un isolat a ete identifie comme ayant potentiellement une hypovirulence controlee par l'ARN bicatenaire, car sa sub-culture sans ARN bicatenaire avait une virulence sigificantivement plus elevee sur les trois especes de pin et pendant deux annees. Les isolats de type A contenant de l'ARN bicatenaire se repartissaient depuis la stabilite jusqu’a une forte interactivite, et de la virulence faible a forte. Les isolats de type B contenant de l'ARN bicatenaire se comportaient de meme pour l'interativite, mais ils allaient de l'avirulence a une virulence moderee, depassant rarement la moyenne de S. sapinea. Les sub-cultures sans ARN bicatenaire tendaient a ne pas exprimer une virulence plus forte que celle des cultures parentales a ARN bicatenaire, mais elles etaient souvent differentes pour l'interactivite. Ainsi, une annee donnee, une sub-culture sans ARN bicatenaire pouvait etre plus virulente que sa culture parentale, et une autre annee elle pouvait etre moins virulente. Zusammenfassung Vierzig Wildtyp-Isolate von Sphaeropsis sapinea wurden aufgrund von zuvor definierten Unterschieden hinsichtlich Morphologie, Kulturmerkmalen und unterschiedlicher Lage von Schnittstellen der Endonukleasen DdeI und BstUI in der nuklearen ribosomalen DNA den Morphotypen A und B zugeteilt. Dreizehn von 20 Typ A Isolaten und neun von 20 Type B Isolaten enthielten Doppelstrang-RNA (dsRNA) unterschiedlicher Grosse und unterschiedlichen Molekulargewichts. Die Ubertragungshaufigkeit von dsRNA in die Konidien betrug 71-100%. Durch Selektion einzelner Konidien wurden von sechs der 22 getesteten dsRNA-infizierten Isolate dsRNA-freie Subkulturen erhalten. Pathogenitatstests an austreibenden Knospen von drei Pinus-Arten zeigten statistisch signifikante Interaktionen zwischen den Faktoren ,,Pinus-Art‘‘ und ,,Untersuchungsjahr‘‘ nach Varianzanalysen (anova) der Virulenz. Die Interaktion von Pinus-Art und Jahr hatte einen starken Einfluss auf die Virulenz. Das Muster dieser Interaktion wurde mittels Hauptkomponentenanalyse der Summe der quadratischen Abweichungen (SQ) des Interaktionstherms der Varianzanalyse ermittelt. Diese sogenannte Interaktivitatsanalyse ist unter dem englishen Begriff AMMI (Additive Main Effects and Multiplicative Interaction) bekannt und hilft bei der Aufteilung der Virulenz auf die verschiedenen Isolate beim Auftreten eines signifikanten Interaktionstherms in der anova. Es wurde die relative Virulenz der Isolate fur funf Kombinationen von Pinus-Art und Jahr geschatzt, und die Isolate wurden entsprechend Virulenz und Interaktivitat mittels des AMMI Models klassifiziert. Die Anfalligkeit von Pinus sylvestris hing sehr stark von saisonalen Effekten ab. Bei P. nigra und P. resinosa war dies weniger ausgepragt. Ein Isolat wurde als von dsRNA besiedelt betrachtet, weil dsRNA-freie Subkulturen des selben Isolats in beiden Jahren eine signifikant starkere Virulenz bezuglich der drei getesteten Pinus Arten aufwiesen. Typ A Isolate mit dsRNA variierten sowohl zwischen schwacher und starker Interaktivitat als auch zwischen schwacher und starker Virulenz. Typ B Isolate mit dsRNA verhielten sich ahnlich hinsichtlich der Interaktivitat, sie waren aber entweder avirulent oder besassen nur schwache Virulenz, welche nur selten die mittlere Virulenz aller Isolate uberstieg. Im Durchschnitt war die Virulenz von isogenen, dsRNA-freien Isolaten nicht starker als diejenige der dsRNA-infizierten Mutterkulturen, aber die Interaktivitat war haufig verandert. Ein dsRNA-freies Isolat konnte deshalb in einem Jahr virulenter und in einem anderen weniger virulent sein als seine dsRNA-infizierte Mutterkultur.