The Fusarium oxysporum f. sp. ciceris/Cicer arietinum pathosystem: a case study of the evolution of plant-pathogenic fungi into races and pathotypes

Summary. The use of resistant cultivars is one of the most practical and cost-efficient strategies for managing plant diseases. However, the efficiency of resistant cultivars in disease management is limited by pathogenic variability in pathogen populations. Knowledge of the evolutionary history and potential of the pathogen population may help to optimize the management of disease-resistance genes, irrespective of the breeding strategy used for their development. In this review, we examine the diversity in virulence phenotypes of Fusarium oxysporum f. sp. ciceris, the causal agent of Fusarium wilt of chickpeas, analyze the genetic variability existing within and among those phenotypes, and infer a phylogenetic relationship among the eight known pathogenic races of this fungus. The inferred intraspecific phylogeny shows that each of those races forms a monophyletic lineage. Moreover, virulence of races to resistant chickpea cultivars has been acquired in a simple stepwise pattern, with few parallel gains or losses. Although chickpea cultivars resistant to Fusarium wilt are available, they have not yet been extensively deployed, so that the stepwise acquisition of virulence is still clearly evident. [Int Microbiol 2004; 7(2):95-104]

Key words: Cicer arietinum · Fusarium wilt · pathogenic variability · DNA fingerprinting · gene genealogy · transposons

Introduction

Plant diseases continue to threaten crop production in modern agriculture, thus challenging plant pathologists for their efficient management. A recent study indicates an average 12.4% loss of attainable harvest for eight crops of significance for food and industry, which together account for more than half of the cultivated area in the world (i.e. barley, cotton, coffee, maize, potato, rice, soybean and wheat) [30]. The monetary value of that loss amounts to 13% of the potential harvest value, to which not less than 10% post-harvest loss should be added.

The use of resistant cultivars is one of the most practical and cost-efficient strategies for plant disease management. Possibly, this control strategy will become increasingly widespread in the coming years in order to satisfy consumers' demand for healthier foods and better environmental quality through new forms of agricultural production, i.e. sustainable agriculture, organic agriculture, and ecological agriculture. However, the efficiency of resistant cultivars in disease management can be seriously limited by pathogenic variability occurring in pathogen populations, including the existence of pathogenic races (defined as biotypes of a pathogen able to cause disease on host cultivars carrying specific resistance genes) and pathotypes (defined as biotypes of a pathogen able to cause a particular disease syndrome in the host plant).

For almost 100 years, since the discovery of pathogenic races in plant pathogens and 'major gene'-regulated complete disease resistance in plants (defined as the ability to prevent the establishment of infection), plant pathologists have witnessed the breakdown of valuable disease resistance as a result of adaptations in the pathogen population to the presence of major disease-resistance genes (R). The most dramatic and best documented examples of resistance breakdown concern biotrophic pathogens, in which there is a gene-for-gene interaction [6] between virulence in the pathogen (defined as the ability to cause disease on plants with race-specific resistance) and resistance in the host (e.g. downy mildews, powdery mildews, rusts). In this interaction, avirulence genes in the pathogen show a functional correspondence with their matching genes in the host so that molecules encoded by avirulence genes are recognized directly or indirectly by the matching R gene product.

The breakdown in resistance was a consequence of directional selection for better-adapted mutants, recombinants, or immigrants, posed by the widespread and intense deployment of R genes favored by monoculture-type practices and genetic uniformity that prevails in modern agriculture [2]. The vulnerability of major-gene resistance has emphasized the advantages of partial, quantitatively inherited resistance based on the efficient expression of plant defense mechanisms against pathogens (i.e. phytoalexins, pathogenesis-related proteins, hydrolytic enzymes, phenolic compounds, cell wall reinforcement, etc.) that do not follow (or the it is difficult to demonstrate) the gene-for-gene relationship. Quantitative resistance tends to be effective against all pathogen variants. However, there have been examples in this type of resistance of an 'erosion' in efficiency rather than a breakdown, as a result of evolution in pathogen populations [5,18]. Therefore, knowledge of the evolutionary history and potential of the pathogen population may help to optimize the management of disease-resistance genes, irrespective of the breeding strategy used for their development, ie. conventional plant breeding or genetic engineering [27]. In this review, we examine the genetic variability existing within and among pathogenic races in a fungal plant pathogen, and infer a phylogenetic relationship among them, using Fusarium oxysporum f. sp. ciceris/chickpea (Cicer arietinum) as a model pathosystem for the analysis.

Virulence phenotypes in Fusarium oxysporum f. sp. ciceris

In addition to major symptom types, eight pathogenic races (races 0, 1A, 1B/C, 2, 3, 4, 5 and 6) can be identified in F. oxysporum f. sp. ciceris based on disease reactions on a set of differential chickpea cultivars [9,13]. Races 0 and 1B/C induce the yellowing symptoms (yellowing pathotype), whereas the remaining races cause wilting (wilting pathotype) [3,15,20]. The eight races also have a distinct geographic distribution. Races 2, 3 and 4 have only been reported in India [9], whereas races 0, 1B/C, 5 and 6 are found mainly in the Mediterranean region and the United States (California) [8,13,15]. Unlike the other races, race 1A is more widespread and has been reported in India, California, and the Mediterranean region [9,15].

Besides the differences in virulence phenotype, pathogenic races of F. oxysporum f. sp. ciceris differ also in aggressiveness (defined as the amount of disease caused by a pathogen genotype on a given genotype of the host plant), as indicated by differences in the amount of inoculum needed to cause severe disease in a chickpea cultivar. At 25ºC, the optimum temperature for Fusarium wilt development, the maximum amount of disease in 'P-2245' chickpea was attained despite a very low initial inoculum density (i.e. 6-50 chlamydospores/g soil) of the wilt-inducing race 5 (Fig. 1A) [28]. Conversely, an inoculum density of 20,000 chlamydospores/g soil of the yellowing-inducing race 0 was required to cause similar amount of disease in that same cultivar (Fig. 1B) [28]. Also, when the aggressiveness of race 1B/C (yellowing pathotype) was compared with that of races 1A and 5 (wilting races) on race 0-resistant 'PV-61' chickpea, maximum intensity of disease occurred with an inoculum density of 1000 chlamydospores/g soil of race 5, which was not significantly different from that reached with 5000 chlamydospores/g soil. However, 5000 chlamydospores/g soil of race 1A were required for the same intensity of disease to develop, and the maximum intensity of disease attained with 5000 chlamydospores/g soil of race 1B/C was similar to that attained with 1000 chlamydospores/g soil of race 1A (Fig. 2).

Note that races 0 and 1B/C share symptom type but differ in virulence to just one of the 10 race-differential cultivars used in this study, whereas races 1B/C and 1A differ in symptom type as well as in partial or full virulence, respectively, to race-differential cv. C-104 (Table 1). Therefore, it appears that, as a whole, the yellowing F. oxysporum f. sp. ciceris pathotype is less aggressive than the wilting pathotype, but differences in aggressiveness to chickpea cultivars can also occur among races within a pathotype. Differences in the stages of plant-pathogen interaction might account for the difference in aggressiveness between F. oxysporum f. sp. ciceris races observed in our studies. One of these would be, for example, the faster and more extensive colonization of xylem vessels in 'P-2245' chickpea roots and stem by race 5 than by race 0, as revealed in histological studies [12].

Fusarium oxysporum is a cosmopolitan soil-borne species complex composed of pathogenically specialized formae speciales and nonpathogenic root-inhabiting parasitic populations, for which no sexual means of reproduction is yet known. The forced asexual reproduction in F. oxysporum precludes regular recombination, and genetic variation must result mainly from the accumulation of mutations. As a consequence, in these fungi the whole genome is linked, transmitted as a unit from one generation to the next, and their natural populations should be clonal in nature.

Clonality in F. oxysporum has been associated with vegetative compatibility [7,21], i.e. the ability of fungal isolates to establish stable heterokaryons by continued anastomosis between adjacent homokaryotic hyphae [4,7,23]. Compatible isolates belong to the same vegetative compatibility group (VCG) and are in effect genetically isolated from isolates in other VCGs within the species by incompatibility mechanisms. In ascomycetes, to which teleomorphic stages of Fusarium belong, vegetative compatibility is homogenic in a set of incompatibility loci (designated vic or v-c), so that strains in the same VCG carry the same alleles at all vic loci [23]. Results of studies in VCG diversity and molecular genetics in F. oxysporum indicate that VCGs and molecular markers are usually not independently associated, which has been interpreted to mean that VCGs in F. oxysporum represent clones or lineages [23]. Furthermore, studies also indicate that the formae speciales in pathogenic F. oxysporum may have different evolutionary patterns. Thus, some formae speciales (i.e. albedinis, canariensis, conglutinans) must have originated once and are therefore monophyletic [1,22], whereas others (i.e. cubense, lycopersici, melonis) must have originated multiple and independent times and are therefore polyphyletic [1,29].

Whether F. oxysporum f. sp. ciceris has monophyletic or polyphyletic origin has a direct consequence on the prospects for disease management through genetic resistance. The high phenotypic and geographic diversity observed within the fungus raises the possibility that F. oxysporum f. sp. ciceris might be polyphyletic. However, despite this high degree of variability, a series of studies showed that all isolates of the pathogen studied-regardless of pathotype, race and geographic origin-are genetically identical for VCG [Nogales-Moncada AM (1997) Ph. D. Thesis, University of Córdoba, Spain], mitochondrial DNA restriction fragment length polymorphism [31], and the sequences of five genes highly conserved in fungi: the translation elongation factor 1α (EF1α), β-tubulin, histone 3, actin, and calmodulin [16]. Moreover, when sequences of the EF1α gene from 17 F. oxysporum f. sp. ciceris isolates representative of all races, pathotypes and geographic range of the pathogen, and three F. oxysporum isolates nonpathogenic to chickpea were compared with 24 EF1α sequences from isolates of 11 different formae speciales of F. oxysporum by parsimony analysis, F. oxysporum f. sp. ciceris isolates formed a clade distinct from other formae speciales and nonpathogenic isolates (Fig. 3) [16]. Altogether, the above information indicates that F. oxysporum f. sp. ciceris has a monophyletic origin.

The simplest interpretation of monophyly and lack of multilocus sequence variation in F. oxysporum f. sp. ciceris is that the pathogen derives from a small founder population that became pathogenic to Cicer spp.; as a consequence, variation in virulence phenotypes must have resulted from the accumulation of relatively recent genetic changes [7]. Based on the monophyly of F. oxysporum f. sp. ciceris, we hypothesized that such a variation must have arisen after the pathogen diverged from other taxa in the F. oxysporum species complex; and in particular that variation in pathotypes and races on chickpea arose in geographically isolated subpopulations which postdated that divergence [15].

Molecular variability and markers in Fusarium oxysporum f. sp. ciceris

The few, hypothesized genetic changes in F. oxysporum f. sp. ciceris are not reflected in the sequences of the genes studied, but they were detected by association with random amplified polymorphic DNA (RAPD) markers. Some 63 [20] and 57 [15] F. oxysporum f. sp. ciceris isolates, representative of all pathotypes, races and geographic origin of the pathogen, were independently assayed using different sets of three and seven RAPD primers, respectively. UPGMA (unweighted paired group method with arithmetic averages) cluster analyses of both RAPD data sets were consistent in grouping F. oxysporum f. sp. ciceris isolates into two main clusters that correlated with the yellowing and wilting pathotypes, respectively [15,20]. Results in the two studies were similar and are illustrated in Fig. 4 [15]. Furthermore, the yellowing isolates were divided into two distinct sub-clusters, which corresponded to races 0 and 1B/C. Results from cluster analysis also were consistent with the analysis of molecular variance; genetic differences between isolates of the yellowing and wilting pathotypes were highly significant, as were differences between races 0 and 1B/C [15].

The DNA banding patterns in the RAPD assays allowed the identification of markers that unambiguously differentiate the yellowing and wilting pathotypes [20], as well as pathogenic races 0, 1B/C, 5 and 6 [15]. The reliability of those markers for diagnostic purposes was confirmed by biological and molecular blind trials and found to be reproducible in repeated experiments [15,20]. Moreover, the race-specific RAPD markers identified were cloned and sequenced, and primers for specific PCR were developed that unambiguously differentiate races 0, 1A, 5, and 6, which are prevalent in the Mediterranean Basin [14].

Genetic variability within and between races of Fusarium oxysporum f. sp. ciceris

The association between RAPD markers, pathotypes and races in F. oxysporum f. sp. ciceris allowed the development of diagnostic markers for them; however, RAPDs lack the resolution needed to assess the genetic variability existing within races of the pathogen as well as for inferring an evolutionary relationship among them [15,20]. Nonetheless, discerning the evolutionary pathways for pathogenic variation within F. oxysporum f. sp. ciceris was possible by means of DNA fingerprinting using three repetitive DNA sequences from the pathogen genome that share similarities with fungal transposons as probes [14,17]. A total of 36 F. oxysporum f. sp. ciceris isolates and three nonpathogenic F. oxysporum isolates representative of all pathotypes, races and geographic origins of the pathogen were assayed with the three probes, resulting in 88 fragments being scored in the hybridization patterns [17].

Phenetic UPGMA analysis of the combined fingerprinting data clearly delineated pathogenic isolates from nonpathogenic ones, and further divided F. oxysporum f. sp. ciceris isolates into two distinct clusters highly correlated with the yellowing and wilting pathotypes (data not shown). Furthermore, isolates within each pathotype were distributed in sub-clusters that corresponded to their race. The amount of fingerprinting similarity within races, calculated using Jaccard's coefficient [35] in NYSYSpc2.0 (Exeter Software, Setauket, NY), varied because of differences in sample size and geographic distribution of races. Nonetheless, with samples roughly comparable in size and geographic variation, the most extreme difference was between the average similarities within races 0 and 5, which were approximately 64% and 93%, respectively [17].

Inferred phylogeny of Fusarium oxysporum f. sp. ciceris races based on DNA fingerprints

The phylogeny of Fusarium oxysporum f. sp. ciceris isolates was inferred from fingerprinting haplotypes in PAUP* 4.0b4a (Sinauer Associated, Suderland, MA, USA) by neighbor-joining [32] analysis. This cladistic analysis produced trees with branching pattern similar to that obtained with the phenetic UPGMA analysis; pathogenic isolates grouped together in a clade with high bootstrap support and they were clearly delineated from isolates nonpathogenic to chickpea. Within pathogenic isolates, two well-supported clades clearly correlated with the yellowing and wilting pathotypes, and fingerprint lineages correlated to races, with each race corresponding to a unique lineage with moderate to high bootstrap support for most races (Fig. 5). Neither UPGMA nor neighbor-joining trees showed any association between fingerprint lineage and geographic origin of isolates, except for grouping of races 2, 3, and 4, which have been reported only from India.

Several lines of evidence indicate that the yellowing race 0 is probably ancestral to the wilting races. First, race 0 is virulent on the fewest race-differential cultivars of all pathogen races [13] and is the most widespread race in the Mediterranean region, although it has not been reported in the Indian subcontinent [9,11,13]. Second, F. oxysporum f. sp. ciceris has a monophyletic origin [16,31], which argues for this pathogen to be derived from parasitic, nonpathogenic populations or from single individuals that acquired pathogenicity on chickpea [3,25]. Therefore, we would expect an ancestral race to be virulent on fewer resistant cultivars than the more recently derived races. Moreover, in this asexually reproducing fungus, genetic variation results mainly from accumulation of mutations over time, which should give rise to greater diversity in older lineages. The relatively low average similarity of fingerprinting [17] and RAPD [15] haplotypes (higher diversity) within race 0 is consistent with the hypothesis of it being an old race. Conversely, race 5 shows the lowest diversity in fingerprinting and RAPD haplotypes and is the most virulent of all races reported in the Mediterranean region [13,15,17]. According to this hypothesis, race 1B/C is the closest to race 0 in evolutionary terms. These two races share the ability to induce the yellowing syndrome in chickpea and show very similar patterns in genetic diversity and virulence [13,15,17].

The origin of pathogenic variation has intrigued plant pathologists since the discovery of pathogenic races in the early twentieth century. For asexually reproducing fungal pathogens such as formae speciales of F. oxysporum, races capable of overcoming resistance genes are thought to evolve by successive accumulation of mutations. In particular, races with close genetic relatedness in monophyletic formae speciales were hypothesized to derive in a stepwise process rather than independently [7]. Although this hypothesis is reasonable, the lack of phylogenetic resolution within monophyletic formae speciales of F. oxysporum made it difficult to analyze the evolutionary pathways of pathogenic variation. Mapping of virulence to each race-differential chickpea cultivar (see Table 1) onto the inferred phylogeny of clonal lineages (Fig. 6) showed for the first time that races of F. oxysporum f. sp. ciceris had evolved in a stepwise process [17]. Mapping was done manually to minimize the number of evolutionary events and to infer the two simplest scenarios: (i) considering only the gain, but not the loss of virulence (Fig. 6A); and (ii) allowing virulence to be either gained or lost (Fig. 6B).

The first scenario postulates at least three parallel gains of virulence and the existence of three ancestral races that have never been observed (Fig. 6A), the first acquisition of virulence from the oldest unknown race giving rise to race 0. Thereafter, yellowing race 1B/C evolved from race 0, and wilting races evolved in a stepwise manner from a second ancestral, an unknown race. This scenario is consistent with the phylogeny of races inferred above, with yellowing race 0 being ancestral to wilting races.

The second scenario inferred in the study [17] seems to be simpler. It suggests that wilting race 1A could be the common ancestor of all races (Fig. 6B), instead of a race that is not known to be extant, as proposed in the first scenario. As expected for an ancient lineage, race 1A is the most widespread geographically; it was first described in India [9] and was later found in the Mediterranean Basin and California [13].

Final remarks

The inferred phylogeny and stepwise evolution of F. oxysporum f. sp. ciceris provides some insights into long-distance migrations that have occurred for this pathogen. It can be speculated that there were at least two migration events between the Mediterranean region and India, most probably as a consequence of the spread of infected seeds [10]. If F. oxysporum f. sp. ciceris evolved in the Mediterranean region, then race 1A and the common ancestor of races 2, 3, and 4 must have been introduced into India. Alternatively, if the pathogen evolved in India, then race 1A, which may be the progenitor of all races (Fig. 6B), should have migrated in the opposite direction. In addition, a common ancestor of races 2, 3, 4 and 5 would have migrated from India to the Mediterranean region. In either case, at least two different races have migrated in or out of India. It can be speculated that the Mediterranean region, or the Fertile Crescent, which is the diversity center for Cicer spp. [26], is the most probable origin of F. oxysporum f. sp. ciceris, at some time after the domestication of chickpea, 8000-9000 years ago [26,34]. Regardless of the origin, the similarities in race composition between the Mediterranean region and California indicate that repeated migration has occurred between these two areas rather than the independent evolution of races in different regions.

It is not probable that the stepwise evolution of races in F. oxysporum f. sp. ciceris resulted from selection by specific resistance in chickpea populations. Some pathogen races are widely distributed geographically even where resistant cultivars have not been deployed. For example, resistant cultivars generally have not been used in the Mediterranean region; this region, however, has a high diversity of races [8,13-15]. Conversely, widespread use of race1A-resistant cultivars H and I in India has not yet led to reports on development of race 6, which specifically overcomes that resistance and derives from race1A (see Table 1) [13,15]; rather, races 2, 3, and 4, which are virulent to cultivars H and I, were reported in India [9] long before the release of these cultivars [24]. Thus, unlike other pathosystems (e.g. downy mildews, rusts, rice blast), there may have been little or no selection for resistance-breaking races of F. oxysporum f. sp. ciceris, which minimizes the probabilities of obtaining parallel changes in virulence, such as those frequently observed in other systems. Limited selection by resistance deployment might help to explain why a relatively simple relationship between clonal lineages and virulence acquisition, with few parallel events, can still be detected.

Acknowledgements. Research on Fusarium wilt of chickpeas in RMJD laboratory was supported by grants AGF97-1479 and AGF98-0878 from Comisión Interministerial de Ciencia y Tecnología (CICYT) of Spain.

1. Baayen RP, O´ Donnell K, Bonants PJM, Cigelnik E, Kroon LPNM, Roebroeck EJA, Waalwijk C (2000) Gene genealogies and AFLP analyses in the Fusarium oxysporum complex identify monophyletic and nonmonophyletic formae speciales causing wilt and rot disease. Phytopathology 90:891-900        [ Links ]

2. Brown JKM (1995) Pathogen´s responses to the management of disease resistance genes. In: Andrews JH, Tommerup IC (eds) Advances in plant pathology, vol. 11. Academic Press, NY, pp 75-102        [ Links ]

3. Correll JC (1991) The relationship between formae speciales, races and vegetative compatibility groups in Fusarium oxysporum. Phytopathology 81:1061-1064        [ Links ]

4. Di Primo P, Cartia G, Katan T (2001) Vegetative compatibility and heterokaryon stability in Fusarium oxysporum f. sp. radicis-lycopersici from Italy. Plant Pathol 50:371-382        [ Links ]

5. Flier WG, van den Bosch GBM, Turkensteen, LJ (2003) Stability of partial resistance in potato cultivars exposed to aggressive strains of Phytophthora infestans. Plant Pathol 52: 326-337        [ Links ]

6. Flor HH (1971) Current status of the gene-for-gene concept. Annu Rev Phytopathol 9:275-296        [ Links ]

7. Gordon TR, Martyn RD (1997) The evolutionary biology of Fusarium oxysporum. Annu Rev Phytopathol 35:111-128        [ Links ]

8. Halila HM, Strange RN (1996) Identification of the causal agent of wilt of chickpea in Tunisia as Fusarium oxysporum f. sp. ciceris race 0. Phytopathol Medit 35:67-74        [ Links ]

9. Haware MP, Nene YL (1982) Races of Fusarium oxysporum f. sp. ciceri. Plant Dis 66:809-810        [ Links ]

10. Haware MP, Nene YL, Rajeshwary R (1978) Eradication of Fusarium oxysporum f. sp. ciceris transmitted in chickpea seed. Phytopathology 68:1364-1367        [ Links ]

11. Jalali BL, Chand H (1992) Chickpea wilt. In: Singh US, Mukhopadhayay AN, Kumar J, Chaube HS (eds) Plant diseases of international importance. Vol. 1. Diseases of cereals and pulses. Prentice Hall, Englewood Cliffs, NJ, pp 420-444         [ Links ]

12. Jiménez-Díaz RM, Basallote-Ureba MJ, Rapoport H (1989) Colonization and pathogenesis in chickpea infected by races of Fusarium oxysporum f. sp. ciceris. In: Tjamos EC, Beckman CH (eds) Vascular wilt diseases of plants. NATO ASI Series Vol. H 28. Springer-Verlag, Berlin, pp 515-520        [ Links ]

13. Jiménez-Díaz RM, Alcalá-Jiménez AR, Hervás A, Trapero-Casas, JL (1993) Pathogenic variability and hosts resistance in the Fusarium oxysporum f. sp. ciceris / Cicer arietinum pathosystem. In: Proc. 3rd Eur. Semin. Fusarium Mycotoxins, Taxonomy, Pathogenicity and Host Resistance. Hodowsla Ròslin Aklimatyazacja i Nasiennictwo. Plant Breeding and Acclimatization Institute, Radzikóv, Poland, pp 87-94         [ Links ]

14. Jiménez-Gasco MM, Jiménez-Díaz RM (2003) Development of a specific polymerase chain reaction-based assay for the identification of Fusarium oxysporum f. sp. ciceris and its pathogenic races 0, 1A, 5, and 6. Phytopathology 93:200-209        [ Links ]

15. Jiménez-Gasco MM, Pérez-Artés E, Jiménez-Díaz RM (2001) Identification of pathogenic races 0, 1B/C, 5, and 6 of Fusarium oxysporum f. sp. ciceris with random amplified polymorphic DNA (RAPD). Eur J Plant Pathol 107:237-248        [ Links ]

16. Jiménez-Gasco MM, Milgroom MG, Jiménez-Díaz RM (2002) Gene genealogies support Fusarium oxysporum f. sp. ciceris as a monophyletic group. Plant Pathol 51:72-77        [ Links ]

17. Jiménez-Gasco MM, Milgroom MG, Jiménez-Díaz RM (2004) Stepwise evolution of races in Fusarium oxysporum f. sp. ciceris inferred from fingerprinting with repetitive DNA sequences. Phytopathology 94:302-307        [ Links ]

18. Johnson R (1983) Genetic background of durable resistance. In: Lamberti F, Waller JM, van der Graaf NA (eds) Durable resistance in crops. NATO ASI Series A., Life Sciences, Plenum Press, NY, pp 5-26        [ Links ]

19. Kaiser WJ, Alcalá-Jiménez AR, Hervás-Vargas A, Trapero-Casas JL, Jiménez-Díaz RM (1994) Screening of wild Cicer species for resistance to races 0 and 5 of Fusarium oxysporum f. sp. ciceris. Plant Dis 78:962-967        [ Links ]

20. Kelly AG, Alcalá-Jiménez AR, Bainbridge BW, Heale JB, Pérez-Artés E, Jiménez-Díaz RM (1994) Use of genetic fingerprinting and random amplified polymorphic DNA to characterize pathotypes of Fusarium oxysporum f. sp. ciceris infecting chickpea. Phytopathology 84:1293-1298        [ Links ]

21. Kistler HC (1997) Genetic diversity in the plant-pathogenic fungus Fusarium oxysporum. Phytopathology 87:474-479        [ Links ]

22. Kistler HC (2001) Evolution of host specificity in Fusarium oxysporum. In: Summerell BA, Leslie JF, Beckhouse D, Bryden WL, Burgess LW (eds) Fusarium. Paul E. Nelson Memorial Symposium. APS Press, St Paul, MN, pp 70-82        [ Links ]

23. Klein KK, Correll JC (2001) Vegetative compatibility group diversity. In: Summerell BA, Leslie JF, Beckhouse D, Bryden WL, Burgess LW (eds) Fusarium. Paul E. Nelson Memorial Symposium. APS Press, St Paul, MN, pp 83-96        [ Links ]

24. Kumar J, Haware MP, Smithson JB (1985) Registration of four short duration Fusarium wilt-resistant kabuli chickpea germplasm. Crop Sci 25:576-577        [ Links ]

25. Leslie JF (1993) Fungal vegetative compatibility. Annu Rev Phytopathol 31:127-150        [ Links ]

26. Lev-Yadun S, Gopher A, Abbo S (2000) The cradle of agriculture. Science 288:1602-1603        [ Links ]

27. McDonald BA, Linde C (2002) Pathogen population genetics, evolutionary potential and durable resistance. Annu Rev Phytopathol 40:349-379        [ Links ]

28. Navas-Cortés JA, Alcalá-Jiménez AR, Hau B, Jiménez-Díaz RM (2000) Influence of inoculum density of races 0 and 5 of Fusarium oxysporum f. sp. ciceris on development of Fusarium wilt in chickpea cultivars. Eur J Plant Pathol 106:135-146        [ Links ]

29. O´ Donnell K, Kistler HC, Cigelnik E, Ploetz RC (1998) Multiple evolutionary origins of the fungus causing Panama disease of banana: concordant evidence from nuclear and mitochondrial gene genealogies. Proc Natl Acad Sci USA 95:2044-2049        [ Links ]

30. Oerke E-C, Weber A, Dehne H-W, Shönbeck F (1994) Conclusions and perspectives. In: Oerke E-C, Dehne H-W, Shönbeck F, Weber A (eds) Crop production and crop protection. Elsevier, Amsterdam, pp 742-770        [ Links ]

31. Pérez-Artés E, Roncero MIG, Jiménez-Díaz RM (1995) Restriction fragment length polymorphism analysis of the mitochondrial DNA of Fusarium oxysporum f. sp. ciceris. J Phytopathol 143:105-109        [ Links ]

32. Saitou N, Nei M (1987) The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol Biol Evol 4:406-425        [ Links ]

33. Singh KB (1987) Chickpea breeding. In: Saxena MC, Singh KB (eds) The chickpea. CAB International, Wallingford, Oxon, pp 127-162        [ Links ]

34. Singh KB (1997) Chickpea (Cicer arietinum L.). Field Crop Res 53:161-170        [ Links ]

35. Sneath P, Sokal R (1973) Numerical taxonomy: the principles and practice of numerical classification. W. H. Freeman, San Francisco, CA        [ Links ]

36. Trapero-Casas A, Jiménez-Díaz RM (1985) Fungal wilt and root rot diseases of chickpea in southern Spain. Phytopathology 75:1146-1151        [ Links ] ]]> 1 2000 90 891-900 2 1995 11 75-102 3 1991 81 1061-1064 4 2001 50 371-382 5 2003 52 326-337 6 1971 9 275-296 7 1997 35 111-128 8 1996 35 67-74 9 1982 66 809-810 10 1978 68 1364-1367 11 1992 1 420-444 12 1989 H 28 Springer-Verlag 515-520 13 1993 87-94 14 2003 93 200-209 15 2001 107 237-248 16 2002 51 72-77 17 2004 94 302-307 18 1983 5-26 19 1994 78 962-967 20 1994 84 1293-1298 21 1997 87 474-479 22 2001 70-82 23 2001 83-96 24 1985 25 576-577 25 1993 31 127-150 26 2000 288 1602-1603 27 2002 40 349-379 28 2000 106 135-146 29 1998 95 2044-2049 30 1994 742-770 31 1995 143 105-109 32 1987 4 406-425 33 1987 127-162 34 1997 53 161-170 35 1973 36 1985 75 1146-1151