HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Annicchiarico, P. Articles by Iannucci, A. Search for Related Content PubMed Articles by Annicchiarico, P. Articles by Iannucci, A. Agricola Articles by Annicchiarico, P. Articles by Iannucci, A. Related Collections Other Legumes Germplasm Enhancement Plant and Environment Interactions

Breeding Strategy for Faba Bean in Southern Europe based on Cultivar Responses across Climatically Contrasting Environments

^a CRA-Istituto Sperimentale per le Colture Foraggere, 29 viale Piacenza, 26900 Lodi, Italy
^b CRA-Istituto Sperimentale per le Colture Foraggere, via Napoli 52, 71100 Foggia, Italy. The work was carried out within the Project "Increase of protein feed production" funded by the Ministry of Agricultural and Forestry Policies of Italy

^* Corresponding author (bred{at}iscf.it).

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Expanding grain legume cropping is desirable but is hindered by low yields. Seventeen faba bean (Vicia faba L.) cultivars belonging to four germplasm types—Mediterranean (from Sicily or Syria), semi-Mediterranean (from continental Italy or Spain), winter, and spring (from France or Germany)—were grown in two climatically contrasting sites (Lodi, subcontinental; Foggia, Mediterranean), two years per site and two sowing times per year, to support breeding strategies by assessing genotype x environment (GE) interactions and their relationship with spatial and temporal factors, germplasm type, and morphophysiological traits. Crossover GE interaction was large and mainly due to the geoclimatic area and the germplasm type. Additive main effects and multiplicative interaction modeling showed (i) the superiority of Mediterranean material across Foggia's environments (all autumn sown), (ii) a trend toward better performance of winter germplasm in Lodi under autumn sowing, and (iii) the similar performance of winter, spring, and semi-Mediterranean types in Lodi under late-winter sowing. Adaptation to each site was related to different and partly incompatible traits, owing mainly to site-specific optima of earliness of cycle and stress tolerance. The results support the specific breeding for each geoclimatic area based on distinct genetic bases and selection environments.

Abbreviations: AMMI, additive main effects and multiplicative interaction • GE, genotype x environment • ICARDA, International Center for Agricultural Research in the Dry Areas • MS, mean square • PC, principal component • SS, sum of squares • T, germplasm type

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
GRAIN LEGUMES MAY PLAY a crucial role in decreasing the marked deficit of high-protein feedstuff and increasing in various respects (safeguard of soil fertility, diversity of cultivated crops, symbiotic nitrogen fixation, reduction of greenhouse gas emissions) the sustainability of European crop–livestock systems (Carrouée et al., 2003). Increasing their yields, however, is essential for expanding their cultivation (Ranalli, 1995). Among them, faba bean (Vicia faba L.) is a rainfed crop whose potential growing area encompasses continental, oceanic, and Mediterranean climates by exploiting different germplasm types and sowing seasons (Lawes et al., 1983; Carrouée et al., 2003). In particular, three germplasm pools can be distinguished in relation to the target climatic zone and sowing time: (i) the Mediterranean, adapted to autumn sowing in Mediterranean regions; (ii) the spring type, adapted to early-spring sowing in cold-prone areas of northern or continental Europe; and (iii) the winter type, including a relatively small number of winter-hardy varieties adapted to autumn sowing in areas with oceanic or fairly mild-winter continental climate (Lawes et al., 1983; Stoddard et al., 2006). Winter material may increase grain yields by extending the cropping season or reducing, via earlier maturity, the exposure to terminal drought of the crop. Sometime, spring-type germplasm is further classified into small-seeded and large-seeded material (Link et al., 1996a). These subgroups have similar adaptation pattern but tend to differ according to molecular markers (Link et al., 1995). A few studies (Polignano and Spagnoletti-Zeuli, 1985; Link et al., 1996a) suggested the presence of subgroups also within the Mediterranean germplasm, depending on the latitude of its area of origin.

Mediterranean and spring germplasm types have exhibited large genotype x environment (GE) interaction between continental Europe (Germany) and Mediterranean areas (southern Spain, southern Italy or Syria) as a consequence of specific adaptation to their optimal climatic and sowing conditions (Kittlitz et al., 1993; Link et al., 1996a,b). The different adaptation pattern may largely be explained by the greater earliness of cycle, drought tolerance, and lodging susceptibility of the Mediterranean germplasm relative to the spring type (Kittlitz et al., 1993; Link et al., 1996a, 1999). Floral adaptation to a specific pollination environment featured by the overwhelming presence of a solitary bee (Eucera numida Lep.) may contribute to specific adaptation to southern Spain (Suso, 2004) but is unlikely to play a major role in Mediterranean regions, such as Italy (Foti, 1982) or Australia (Marcellos and Perryman, 1990), where bumblebees (Bombus terrestris L.) or honeybees (Apis mellifera L.) are the main pollinators.

There is direct or indirect evidence for the heritability of faba bean adaptive responses to geographical areas (Link et al., 1996b) or drought stress levels (Abdelmula et al., 1999). These responses have a major impact on the definition of breeding strategies. For instance, the much larger heterotic effects of crosses between Mediterranean and spring germplasm pools relative to intrapool crosses may be exploited for hybrid breeding, but the bearing of specific-adaptation effects inherited from parent material suggested to rather rely on intrapool crosses of specifically adapted material when breeding for Mediterranean regions (Link et al., 1996a; Schill et al., 1998). This may also apply to breeding of synthetic varieties that exploit the heterotic effects by selecting few parents with enhanced outcrossing rate (Suso et al., 2005). In general, the extent of GE interactions as determined by geographical areas and cropping years on the one hand and the characteristics of the potential genetic base on the other are important for defining adaptation strategies, yield stability targets, selection environments, parent material, and plant ideotypes for breeding (Kang, 1998; Annicchiarico, 2002). Genotype x environment interactions mainly due to geoclimatic areas may conveniently be addressed by breeding specific cultivars for each area.

The occurrence of large GE interaction between Mediterranean and continental Europe (Link et al., 1996a,b) is not surprising and justifies the presence of distinct selection programs for these regions. Little information is available, however, on the extent and the pattern of GE interaction that may occur at a smaller geographical scale, such as southern Europe. Cropping environments in this region are featured by large climatic variation between sites, between years, and along the cropping year. Terminal drought stress may be severe, especially in typical Mediterranean environments (such as those of southern Italy). Winter plant mortality may be substantial in subcontinental-climate areas (such as northern Italy) and, occasionally, also in Mediterranean areas, where plants not completely hardened or partly dehardened may be susceptible to frost events of limited duration and severity (Annicchiarico and Iannucci, 2007). The diversity in extent and combination of climatic stresses may produce large GE interaction, but the impact on it of germplasm types, climatically contrasting areas, and different test years has not been investigated.

In this study, a number of geographically diversified faba bean cultivars underwent multienvironment evaluation in two climatically contrasting Italian sites with the objective to define various elements of a breeding strategy for south-European conditions. In particular, the study aimed (i) to assess the impact of GE interaction effects due to spatial and temporal factors on the consistency of top-yielding cultivars across environments, (ii) to evaluate the similarity between environments for GE effects and its implications for adaptation strategies, (iii) to evaluate the relationship of adaptive responses with various yield components and putative adaptive traits; and (iv) to assess the adaptation pattern and the combination of adaptive traits featuring spring, Mediterranean, and winter germplasm types.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Experimental Data
The evaluation sites were Lodi (Lombardy), representative of the subcontinental climate with extended frosts in winter and modest terminal drought which is typical of northern Italy, and Foggia (Apulia), representative of the Mediterranean climate with occasional frost spells within mild winters and severe terminal drought, which is widespread in southern Italy and coastal areas of central Italy (Perini et al., 2004). In each site, the evaluation included two test years and two sowing times per year, for a total of eight test environments (Table 1 ). The first year in Lodi and both years in Foggia included two autumn sowings, which were about 2 wk apart and represented an early and late sowing, respectively, according to the climatic characteristics of each site. Adopting more sowing dates was justified by the fact that winter plant mortality in grain legumes may be more severe in early or late sowings depending on the temperature pattern of the specific year (Etévé, 1985). The second year in Lodi included a late autumn sowing and a late winter sowing (Table 1), consistent with the occasional adoption of the latter sowing in northern Italy when unfavorable climatic conditions prevent the autumn sowing.

For each site-year combination, the experiment was designed as a split-plot with three replications, assigning sowing times to main plots and cultivars to subplots. Each plot was 9 m² and included eight rows 3 m long, 36 cm apart. The seedling density targeted by seed rates varied depending on seed size according to Stringi and Giambalvo (2001) and implied 45 germinating seeds m^–2 for minor, 36 for equina, and 27 for major material. The sowing depth was 6 cm. Seeds were treated with Wakil (50 g Fludioxynil kg^–1 + 100 g Cymoxanil kg^–1 + 175 g Metalaxil kg^–1) at the rate of 2 g kg^–1 of seed. The fertilization was site-specific and included 20 kg ha^–1 of N, 48 kg ha^–1 of P₂O₅ and 112 kg ha^–1 of K₂O in Lodi, and 54 kg ha^–1 of N and 138 kg ha^–1 of P₂O₅ in Foggia.

The traits, all recorded on a plot basis, were (i) grain yield expressed at 13% seed moisture, straw dry matter (assessed from the combine cutting height, collecting all the residues from threshing), aerial biomass dry matter, and harvest index, all observed on a 4.5 m² harvest area; (ii) plant mortality in percentage during winter and over the crop cycle, based on plant counts along 1 linear m in each of two rows performed at the onset and at the end of winter and at harvesting; (iii) susceptibility to Botrytis fabae, assessed by the 9-level visual scale reported by Bernier et al. (1984); (iv) onset of flowering (as days from 1 March to when 50% of plants had the first open flower) and duration of flowering (in days); (v) canopy height at the onset of flowering and at maturity; (vi) susceptibility to lodging, as percentage of lodged plants at harvesting; (vii) seed weight (on 250 random, dried grains); and (viii) the following traits averaged across a random sample of six ripe plants: height of the first pod (until its bearing node), plant seed dry weight, plant dry biomass, number of fertile stems per plant, number of fertile nodes (main stem), number of pods per fertile node (main stem), and number of seeds per pod (main stem). Damage from B. fabae or lodging was recorded in a subset of environments in which it was sizable. Winter mortality was not observed in the second year in Foggia (where it was negligible) and Lodi (where its reliable assessment was prevented by the slow plant emergence caused by low autumn temperatures).

Statistical Analysis
Plot data of each trait were submitted to an analysis of variance (ANOVA) with genotype (i.e., cultivar) and environment as fixed factors. For traits observed in all test environments, the ANOVA assessed the genotypic, environmental, and GE interaction effects relative to germplasm type (T), cultivar within T, location, time of autumn sowing, year of both autumn sowings (Foggia) or late autumn sowing (Lodi), and autumn vs. late-winter sowing in the second year in Lodi (Table 2 ). Main effect and GE interaction variation for T and cultivar within T were multiple-df contrasts (Gomez and Gomez, 1984). Their expected mean square (MS) values for main effect variation were equal to _e² + e b (m_i – m.)²/(t – 1), where _e² is the pooled experiment error variance, e and b are numbers of environments and experiment blocks, respectively, and the remaining terms stand for the sum of squares (SS) [(m_i – m.)²] and the df values relative to the t true treatment means considered by each contrast (Dagnelie, 1975). As the expected MS of these terms only differed for the variance of the relevant effects, their observed MS (concisely expressed as the MS ratio of T to cultivar within T) indicated which of the two genotypic effects tended to be larger. Likewise, the MS comparison of T x environment vs. cultivar within T x environment interaction (or other MS comparisons relative to multiple-df contrasts that partition the GE interaction variation) indicated which of the two effects tended to be larger, as the expected MS of these multiple-df contrasts had _e² in common and differed for the variance of the relevant GE effects (Dagnelie, 1975). Percentage data were submitted to angular transformation before ANOVA.

Statistical Analysis System (SAS Institute, Cary, NC) software was used for all analyses except AMMI and joint regression, which were performed by IRRISTAT (released by the International Rice Research Institute, Los Baños, Laguna, Philippines).

	RESULTS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Grain Yield Responses
Test year and long-term values of winter and spring temperatures and spring rainfall are given for the two sites in Table 1. Lodi was characterized by sizable winter cold stress (mean winterkilled plants = 14%) and an unusually dry spring in the first year, and by severe winter cold stress and limited drought stress in the second year. In Foggia the first year was characterized by fairly high winter plant mortality (20%) due to late frosts on incompletely hardened plants, followed by the ordinary, severe level of terminal drought stress. The second year was climatically more favorable than the average for the site. The ANOVA indicated the higher potential for faba bean yield of the subcontinental site over the Mediterranean site (2.85 vs. 1.89 t ha^–1) (Table 2).

Genotypic and GE interaction effects were significantly affected by T and cultivar within T (P < 0.001), but the impact of T tended to be greater when comparing its MS with that of cultivar within T (Table 2). Germplasm type interacted with location, test year in Foggia, and sowing season in Lodi, whereas cultivar within T interacted with location (P < 0.001; Table 2). The much larger MS of T x location interaction relative to other GE interaction MS (Table 2) suggested that this source of variation was the main determinant of GE interaction. There was significant crossover interaction of germplasm types across locations, as the Mediterranean germplasm was the best yielding in Foggia and the worst yielding in the subcontinental-climate location (Table 3 ). The other germplasm types performed similarly in Lodi, whereas the semi-Mediterranean material outperformed the winter and spring types in Foggia (Table 3).

View larger version (15K): [in this window] [in a new window]   Figure 1. (a) Average nominal grain yield of four faba bean germplasm types, and (b) nominal yield of nine best-performing cultivars, as a function of the environment score on the first genotype x environment interaction principal component (PC). Line identifier for germplasm type used in (a) also applies to individual cultivars in (b). (See Table 1 for acronym of environments.)

Genotypic and GE interaction effects were significant (P < 0.05) for all traits except for GE effects relative to susceptibility to B. fabae. Range values of cultivars over sites are given in Table 4 for each trait. With a few exceptions (susceptibility to B. fabae; number of seeds per pod), T contributed to entry variation more than cultivar within T when comparing the respective MS values (Table 4).

Owing to GE interaction effects and the different traits associated with them, higher yield in the two locations was mostly related to different characteristics. In the Mediterranean-climate site it was related (P < 0.05) to earlier and less prolonged flowering, taller stature at the onset of flowering, higher harvest index, greater aerial biomass, heavier and more productive plants, and lower plant mortality (mainly due to spring mortality) (Table 4). These features characterized the locally best-adapted germplasm type, that is, the Mediterranean one, in the germplasm comparison reported in Table 3 (for sake of synthesis, comparisons across locations were preferred to site-specific ones whenever the ANOVA MS ratio of T to T x location interaction exceeded 2.0, as this ratio level implies a similar rank of germplasm types across locations).

Higher grain yield in Lodi was associated (P < 0.05) with later flowering, taller stature at the onset of flowering and at maturity, greater standing ability, higher first pod, greater straw and aerial biomass, lower plant mortality in winter and overall, more fertile nodes on the main stem, less fertile stems per plant, and smaller grains (Table 4). The winter material represented well this combination of traits, whereas the Mediterranean germplasm tended to the opposite features and the spring germplasm displayed high plant mortality (Table 3). The inverse relationships of number of fertile stems and seed weight with grain yield in this site (Table 4) probably arose from the inverse correlation (r < –0.66, P < 0.01) of these yield components with the component most closely related to yield, namely, the number of fertile nodes. Higher yield in Lodi was also associated with more harvested plants per square meter (r = 0.67, P < 0.01), but information on this yield component was not reported because of the different seedling density of entries.

Some traits associated with higher yield in one or both sites exhibited large T x location interaction, which tended to emphasize, in the specifically adapted germplasm, the locally favorable characteristic. This was the case for (i) crop aerial biomass, canopy height at the onset of flowering and overall plant mortality (highest for the Mediterranean germplasm in Foggia and the winter germplasm in Lodi); (ii) winter plant mortality (lowest for the winter material in Lodi, and similar for all germplasm types in Foggia); and (iii) number of fertile nodes on the main stem (among the highest in Lodi and sharply reduced in Foggia for the winter material) (Table 3).

	DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
This study indicates that large crossover GE interaction may occur not only between continental and Mediterranean Europe (Link et al., 1996a,b) but also between contrasting geoclimatic areas of southern Europe, highlighting the impact on it of the germplasm type and its morphophysiological characteristics. Actually, the inclusion of winter germplasm besides spring and Mediterranean ones and the widespread adoption of autumn sowing also for the cold-prone area boosted the contribution to GE interaction of entry variation for winter hardiness in comparison to Link et al. (1996a,b). The distinction of Mediterranean and semi-Mediterranean types largely accounted for the wide variation in adaptive response among entries selected in Mediterranean areas that emerged already in Link et al. (1996a). It is noteworthy that the ordinary levels of variation in drought and cold stress across the target agricultural environments gave rise to much larger crossover GE interaction than that reported between managed environments with sharply contrasting levels of drought stress (Amede et al., 1999; Link et al., 1999). The superiority of AMMI over joint regression for modeling GE effects agrees with results by Link et al. (1996b) for faba bean and is common when adaptive responses are affected by two or more environmental constraints (Annicchiarico, 2002).

Earliness of crop cycle of the cultivars was not assessed directly but may be inferred either from the combination of early and less prolonged flowering or from low number of fertile nodes on the main stem under relatively favorable spring growing conditions (Dumoulin et al., 1994), such as those in Lodi. The Mediterranean germplasm, which exhibited these characteristics, was specifically adapted to Mediterranean environments, in agreement with the known importance of earlier maturity for escaping terminal drought in faba bean (Lawes et al., 1983; Link et al., 1996a,b) and other grain legumes (Stoddard et al., 2006). These characteristics contributed also to its poor adaptation to autumn-sown subcontinental-climate environments, where delayed floral initiation could provide a mechanism for escaping winter frosts (Herzog, 1988) and lateness of cycle allowed to fully exploit the favorable growing conditions during spring. Greater drought tolerance of Mediterranean entries (as inferred by lower spring mortality in Foggia) and greater winter hardiness of winter material (as indicated by winter survival in Lodi) contributed to the contrasting adaptation pattern of these germplasms.

In addition, the number of fertile stems, besides that of fertile nodes, may contribute to site-specific adaptation on the basis of correlation results. More stems, exhibited by Mediterranean germplasm in both sites, were useful in Foggia to counterbalance the severe stand reduction. The usefulness of greater branching emerged in another Mediterranean region, northern Syria (Saxena et al., 1981). More seeds per pod, advocated by Lawes et al. (1983) as a major trait for yield improvement, showed no relationship to yield.

Correlation results also suggest the positive value for Lodi of entry ability to limit seed losses at harvesting by means of higher first pod and greater standing ability. The short stature of plants at the onset of flowering in Lodi (arising from limited autumn and winter growth) implied an average height of the first pod lower than in Foggia (16.3 vs. 22.3 cm) and well below the value of 20 cm indicated by Lawes et al. (1983) to exclude seed losses. The modest extent of lodging in Lodi suggests that lower standing ability of the Mediterranean germplasm, which emerged earlier (Kittlitz et al., 1993), was a minor reason for the poor adaptation of this material in Lodi.

Higher crop aerial biomass, taller canopy at the onset of flowering, and lower overall plant mortality were the only traits associated with higher yield in both geoclimatic areas. However, they exhibited large genotype x location interaction effects, which paralleled those for yield and the specific-adaptation responses of cultivars, suggesting that we should consider the entry values of each trait in the two sites as two different characters (under partly different genetic control) rather than the same character (Falconer, 1989). This reinforces the conclusion that the two geoclimatic areas require distinct combinations of traits for cultivar adaptation. Greater aerial biomass was the main characteristic associated with higher yield in south Australian environments (Agung and McDonald, 1998), whereas a positive relationship between GE effects for yield and plant stature emerged earlier between favorable and drought-stress environments (Link et al., 1999).

Germplasm type x location interaction effects were also observed for winter plant mortality and number of fertile nodes. The former agree with the sizable GE interaction for freezing tolerance reported by Herzog (1988) between prehardening temperatures sufficient (as in Lodi) or insufficient (as in Foggia) for complete plant hardening and the positive effect on freezing tolerance of large seedling (typical of large-seeded, Mediterranean material) under incomplete hardening. The great susceptibility to drought stress at early podding stage of faba bean (Mwanamwenge et al., 1999) may account for the greater decrease in fertile nodes shown in Foggia by late-flowering types, especially winter material, and the consequent T x location interaction.

The extent of crossover GE interaction between geoclimatic areas and the difficulty to combine into a unique genotype the different and partly incompatible adaptive traits that are useful in each area support a specific-adaptation strategy based on distinct genetic bases and selection environments even for a fairly small region such as Italy. The presence of a Mediterranean and a subcontinental climatic area also for other south European regions (Spain, Portugal, the Balkan Peninsula) in a tentative definition of adaptive zones for faba bean varieties (Metayer, 2004) suggests, however, that specific breeding may be applied to transnational south European climatic areas to widen its commercial scope. Our results may also have implications for other climatically heterogeneous regions (e.g., Australia, where faba bean is grown along a wide latitudinal gradient; Matthews and Marcellos, 2003). Schill et al.'s (1998) conclusion to rely on crosses within Mediterranean material for Mediterranean areas while exploiting the heterotic effects of Mediterranean x Central European interpool crosses for continental-climate areas may also apply to the current, less-contrasting areas, given the definite superiority of Mediterranean material in Foggia and the only moderate yield difference between winter, spring, and semi-Mediterranean types in Lodi. Some traits of specific interest for either area that are not affected markedly by GE interaction, such as earliness and duration of flowering, number of fertile stems per plant, and height of the first pod, may be used for early identification of specifically adapted material in early selection stages independently of the test environment. Emerging physiological traits related to drought tolerance (Khan et al., 2007) may contribute to selection for Mediterranean environments.

	ACKNOWLEDGMENTS

 
We thank S. Proietti and V. Miullo for excellent technical assistance.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
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.

Received for publication January 3, 2008.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

• Abdelmula, A.A., W. Link, E. von Kittlitz, and D. Stelling. 1999. Heterosis and inheritance of drought tolerance in faba bean, Vicia faba L. Plant Breed. 118:485–490.[CrossRef]
• Agung, S., and G.K. McDonald. 1998. Effects of seed size and maturity on the growth and yield of faba bean (Vicia faba L.). Aust. J. Agric. Res. 49:79–88.[CrossRef][Web of Science]
• Amede, T., E. von Kittlitz, and W. Link. 1999. Differential drought responses of faba bean (Vicia faba L.) inbred lines. Plant Breed. 183:35–45.
• Annicchiarico, P. 2002. Genotype x environment interactions: Challenges and opportunities for plant breeding and cultivar recommendations. FAO Plant Production and Protection Paper 174. FAO, Rome.
• Annicchiarico, P., F. Bellah, and T. Chiari. 2006. Repeatable genotype x location interaction and its exploitation by conventional and GIS-based cultivar recommendation for durum wheat in Algeria. Eur. J. Agron. 24:70–81.[CrossRef]
• Annicchiarico, P., and A. Iannucci. 2007. Winter survival of pea, faba bean and white lupin cultivars across contrasting Italian locations and sowing times, and implications for selection. J. Agric. Sci. 145:611–622.
• Bernier, C.C., S.B. Hanounik, M.M. Hussein, and H.A. Mohamed. 1984. Field manual of common faba bean diseases in the Nile Valley. ICARDA, Aleppo, Syria.
• Carrouée, B., K. Crépon, and C. Peyronnet. 2003. High-protein crops: Their advantages in French and European forage production systems. (In French, with English abstract.). Fourrages 174:163–182.
• Dagnelie, P. 1975. Théorie et méthodes statistiques. Vol. 2. (In French.) Les presses agronomiques, Gembloux, Belgium.
• Dumoulin, V., B. Ney, and G. Etévé. 1994. Variability of seed and plant development in pea. Crop Sci. 34:992–998.[Abstract/Free Full Text]
• Etévé, G. 1985. Breeding for cold tolerance and winter hardiness in pea. p. 131–136. In P.D. Hebblethwaite, M.C. Heath and T.C.K. Dawkins (ed.) The pea crop: A basis for improvement. Butterworths, London.
• Falconer, D.S. 1989. Introduction to quantitative genetics. 3rd ed. Longman, Harlow, UK.
• Finlay, K.W., and G.N. Wilkinson. 1963. The analysis of adaptation in a plant-breeding programme. Aust. J. Agric. Res. 14:742–754.[CrossRef][Web of Science]
• Foti, S. 1982. Fava (Vicia faba L. o Faba vulgaris Moench.). In R. Baldoni and L. Giardini (ed.) Coltivazioni erbacee. Pàtron, Bologna, Italy.
• Gauch, H.G. 1992. Statistical analysis of regional yield trials: AMMI analysis of factorial designs. Elsevier, Amsterdam.
• Gauch, H.G., and R.W. Zobel. 1997. Identifying mega-environments and targeting genotypes. Crop Sci. 37:311–326.[Abstract/Free Full Text]
• Gomez, K.A., and A.A. Gomez. 1984. Statistical procedures for agricultural research. 2nd ed. John Wiley & Sons, New York.
• Herzog, H. 1988. Winter hardiness in faba beans: Varietal differences and interrelations among selection criteria. Plant Breed. 101:269–276.[CrossRef]
• Kang, M.S. 1998. Using genotype-by-environment interaction for crop cultivar development. Adv. Agron. 62:199–252.
• Khan, H.R., W. Link, T.J. Hocking, and F.L. Stoddard. 2007. Evaluation of physiological traits for improving drought tolerance in faba bean (Vicia faba L.). Plant Soil 292:205–217.[CrossRef][Web of Science]
• Kittlitz, E.v., K.I.M. Ibrahim, P. Ruckenbauer, and L.D. Robertson. 1993. Analysis and use of interpool-crosses (Mediterranean x central European) in faba beans (Vicia faba L.): I. Performance of Mediterranean and central European faba beans in Syria and Germany. Plant Breed. 110:307–314.[CrossRef]
• Lawes, D.A., D.A. Bond, and M.H. Poulsen. 1983. Classification, origin, breeding methods and objectives. p. 23–76. In P.D. Hebblethwaite (ed.) The faba bean (Vicia faba L.): A basis for improvement. Butterworths, London.
• Link, W., A.A. Abdelmula, E. von Kittlitz, S. Bruns, H. Riemer, and D. Stelling. 1999. Genotypic variation for drought tolerance in Vicia faba. Plant Breed. 118:477–483.[CrossRef]
• Link, W., C. Dixkens, M. Singh, M. Schwall, A.E. Melchinger, and P. Jaccard. 1995. Genetic diversity in European and Mediterranean faba bean germplasm revealed by RAPD markers. Theor. Appl. Genet. 90:27–32.[Web of Science]
• Link, W., B. Schill, A.C. Barbera, J.I. Cubero, A. Filippetti, L. Stringi, and E. von Kittlitz. 1996a. Comparison of intra- and inter-pool crosses in faba beans (Vicia faba L.): I. Hybrid performance and heterosis in Mediterranean and German environments. Plant Breed. 115:352–360.[CrossRef]
• Link, W., B. Schill, and E. von Kittlitz. 1996b. Breeding for wide adaptation in faba bean. Euphytica 92:185–190.[CrossRef][Web of Science]
• Marcellos, H., and T. Perryman. 1990. Control over pollination and ovule fertilization in Vicia faba. Euphytica 49:5–13.[CrossRef][Web of Science]
• Matthews, P., and H. Marcellos. 2003. Faba bean. 2nd ed. Agfact P4.2.7. NSW Dep. Primary Industries, Orange, NSW, Australia.
• Metayer, N. 2004. Vicia faba breeding for sustainable agriculture in Europe: Identification of regional priorities and definition of target genotypes. GIE Févérole, Paris.
• Mwanamwenge, J., S.P. Loss, K.H.M. Siddique, and P.S. Cocks. 1999. Effect of water stress during floral initiation, flowering and podding on the growth and yield of faba bean (Vicia faba L.). Eur. J. Agron. 11:1–11.[CrossRef]
• Perini, L., M.C. Beltrano, G. Dal Monte, S. Esposito, T. Caruso, A. Motisi, and F.P. Marra. 2004. Atlante agroclimatico: Agroclimatologia, pedologia, fenologia del territorio italiano. (In Italian.) UCEA, Rome.
• Piepho, H.-P. 1995. Robustness of statistical tests for multiplicative terms in the additive main effects and multiplicative interaction model for cultivar trials. Theor. Appl. Genet. 90:438–443.[Web of Science]
• Polignano, G.B., and P.L. Spagnoletti-Zeuli. 1985. Variation and covariation in Vicia faba L. populations of Mediterranean origin. Euphytica 34:659–668.[CrossRef][Web of Science]
• Ranalli, P. 1995. Improvement of pulse crops in Europe. Eur. J. Agron. 4:151–166.
• Saxena, M.C., G.C. Hawtin, and H. El-Ibrahim. 1981. Aspects of faba bean ideotypes for drier conditions. p. 210–231. In R. Thomson (ed.) Vicia faba: Physiology and breeding. M. Nijhoff, The Hague, The Netherlands.
• Schill, B., A.E. Melchinger, R.K. Gumber, and W. Link. 1998. Comparison of intra- and inter-pool crosses in faba beans (Vicia faba L.): II. Genetic effects estimated from generation means in Mediterranean and German environments. Plant Breed. 117:351–359.[CrossRef]
• Stoddard, F.L., C. Balko, W. Erskine, H.R. Khan, W. Link, and A. Sarker. 2006. Screening techniques and sources of resistance to abiotic stresses in cool-season food legumes. Euphytica 147:167–186.[CrossRef][Web of Science]
• Stringi, L., and D. Giambalvo. 2001. Fava (Vicia faba L.). p. 403–445. In P. Ranalli (ed.) Leguminose e agricoltura sostenibile. (In Italian.) Calderini edagricole, Bologna, Italy.
• Suso, M.J. 2004. Pollinators: The neglected side of faba bean improvement. Grain Legumes 40:8–9.
• Suso, M.J., L. Harder, M.T. Moreno, and F. Maalouf. 2005. New strategies for increasing heterozigosity in crops: Vicia faba mating system as a study case. Euphytica 143:51–65.[CrossRef][Web of Science]

This article has been cited by other articles:

				P. Annicchiarico Adaptation of Cool-Season Grain Legume Species across Climatically-Contrasting Environments of Southern Europe Agron. J., October 21, 2008; 100(6): 1647 - 1654. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Annicchiarico, P. Articles by Iannucci, A. Search for Related Content PubMed Articles by Annicchiarico, P. Articles by Iannucci, A. Agricola Articles by Annicchiarico, P. Articles by Iannucci, A. Related Collections Other Legumes Germplasm Enhancement Plant and Environment Interactions