Genetic diversity of Fusarium oxysporum and related species pathogenic on tomato in Algeria and other Mediterranean countries




In order to characterize the pathogen(s) responsible for the outbreak of fusarium diseases in Algeria, 48 Fusarium spp. isolates were collected from diseased tomato in Algeria and compared with 58 isolates of Fusarium oxysporum originating from seven other Mediterranean countries and 24 reference strains. Partial sequences of the translation elongation factor EF-1α gene enabled identification of 27 isolates as F. oxysporum, 18 as F. commune and three as F. redolens among the Algerian isolates. Pathogenicity tests confirmed that all isolates were pathogenic on tomato, with disease incidence greater at 28°C than at 24°C. All isolates were characterized using intergenic spacer (IGS) DNA typing, vegetative compatibility group (VCG) and PCR detection of the SIX1 (secreted in xylem 1) gene specific to F. oxysporum f. sp. lycopersici (FOL). No DNA polymorphisms were detected in the isolates of F. redolens or F. commune. In contrast, the 27 Algerian isolates of F. oxysporum were shown to comprise nine IGS types and 13 VCGs, including several potentially new VCGs. As none of the isolates was scored as SIX1+, the 27 isolates could be assigned to F. oxysporum f. sp. radicis-lycopersici (FORL). Isolates from Tunisia were also highly diverse but genetically distinct from the Algerian isolates. Several Tunisian isolates were identified as FOL by a PCR that detected the presence of SIX1. The results show that isolates from European countries were less diverse than those from Tunisia. Given the difference between Algerian populations and populations in other Mediterranean countries, newly emergent pathogenic forms could have evolved from local non-pathogenic populations in Algeria.


Fusarium diseases of tomato plants (Solanum esculentum) are known to be mainly caused by two formae speciales of the morphospecies Fusarium oxysporum, F. oxysporum f. sp. radicis-lycopersici (FORL), responsible for crown and root rot, and F. oxysporum f. sp. lycopersici (FOL), responsible for a vascular wilt disease. Members of the species complex F. solani may cause some crown and root rots of solanaceous plants such as potato and tomato under tropical conditions (Romberg & Davis, 2007; Blancard et al., 2009) and some other species may be potentially responsible for pre-emergence damping off of various crops including tomato (Palmero et al., 2009). In contrast to these species, formae speciales FORL and FOL have strict host specificity. Within FOL, three races have been identified, differentiated by their specific pathogenicity to different cultivars carrying specific resistance genes (Grattidge & O’Brien, 1982). However, no races are known within FORL. Both formae speciales display a considerable genetic diversity. Like many formae speciales in F. oxysporum, FORL and FOL are composed of several vegetative compatibility groups (VCG) and are known to be polyphyletic (Kistler, 1997; O’Donnell et al., 2009). To date, nine VCGs have been identified among FORL isolates (VCG 0090 to 0094 and 0096 to 0099) and five VCGs among FOL isolates (0030 to 0033 and 0035) (Katan, 1999; Cai et al., 2003). The two formae speciales can be identified and distinguished by time-consuming pathogenicity tests on tomato and by the disease symptoms. VCG can also help identify formae speciales but this method requires obtaining mutants and pairing them with reference strains until a heterokaryon is detected, which is also very time-consuming (Correll et al., 1987). However, VCG grouping remains the reference method to characterize pathogenic strains of F. oxysporum because it makes it possible to compare new isolates with reference strains and previously collected isolates. Recently, a rapid molecular method has been proposed to identify FOL, based on the PCR amplification of specific markers corresponding to the SIX genes (secreted in xylem) (van der Does et al., 2008; Lievens et al., 2009). Genetic diversity within formae speciales can be effectively characterized by targeting different DNA regions, the ribosomal intergenic spacer (IGS) being one of the most polymorphic (Lori et al., 2004; Abo et al., 2005; O’Donnell et al., 2009).

Fusarium diseases of tomato have caused severe damage in tomato growing areas all over the world. During the past decade, diseases caused by FOL have decreased in European countries due to the availability of resistant cultivars. However, resistance genes can be overcome, resulting in the emergence of new races. Only tolerant cultivars are available for FORL but the disease can be effectively prevented using grafted material (Blancard et al., 2009). However, this strategy is not always used and fusarium disease due to FORL is still present in many Mediterranean countries including Turkey (Can et al., 2004) and Malta (Porta-Puglia & Mifsud, 2005), where it was recently reported for the first time.

Tomato is an important and common crop in North Africa. Contrary to the decline in several other countries, fusarium diseases are still frequent in tomato-growing areas in Tunisia (Hibar et al., 2007). Similarly, an outbreak of fusarium diseases of tomato was recently discovered in Algeria (N. Hamini-Kadar, University of Oran, Algeria, personal communication). However, nothing is known about the pathogen(s) responsible for this outbreak of disease. This study was therefore initiated to determine: (i) whether the pathogens are diverse or are composed of a limited number of genotypes; and (ii) whether they are clonally related to other known strains of FOL and FORL found in neighbouring countries or are newly emergent pathogenic forms. These questions were addressed by characterizing the genetic diversity of the fungal isolates responsible for fusarium diseases in Algeria, and by comparing them with isolates originating from other Mediterranean countries.

Material and methods

Fungal isolates

From 2006 to 2008, surveys were conducted in tomato-growing areas of Algeria, mainly around Alger and Oran. Diseased plants showed discoloration of stem vascular tissue, wilting of the lower leaves, and crown and root rot. Isolates with morphological characteristics similar to those of F. oxysporum (Nelson et al., 1983) were collected from diseased plants or soil and preserved in the MIAE collection (Microorganisms of Interest for Agriculture and Environment, INRA Dijon, France).

The 48 isolates collected in Algeria were compared with 58 isolates of F. oxysporum originating from seven other Mediterranean countries: seven isolates previously identified as FOL, 34 isolates previously identified as FORL and 17 isolates of F. oxysporum originating from diseased tomato for which the forma specialis was unknown (Table 1). Single-spore isolations were conducted for the 106 isolates. In addition, 24 strains were used as reference strains in VCG tests and/or as controls in the molecular characterization and pathogenicity tests (Table 1). To characterize VCGs, reference strains of all known VCGs within the formae speciales lycopersici and radicis-lycopersici were used, with the exception of VCG 0035 and VCG 0097, which were not available from any collection.

Table 1.   Isolates of Fusarium analysed in this study
SpeciesIsolateMIAE accession numberGeographic originOriginal code (source)aHostYear of isolationIGS typebVCGcSIX1 PCRdPercentage of diseasee
CountryRegionTownAt 24°CAt 28°C
  1. *VCG testers.

  2. MIAE, Microorganisms of Interest for Agriculture and environment, INRA Dijon, France.

  3. aAB, A. Buffière, Institut National de la Recherche Agronomique (INRA), Montfavet, France; AM, A. Moretti, INRA, Montfavet, France; DV, D. Vakalounakis, National Agriculture Research Foundation (NAGREF), Plant Protection Institute, Heraklio, Crete, Greece; ILN, I. Larena Nistal, Department of Plant Protection, Madrid, Spain; KE, K. Elena, Department of Phytopathology, Athens, Greece; KH, K. Hibar, Ecole Supérieure d’Horticulture de Chott-Mariem, Sousse, Tunisia; LRG, L. Rosewich Gale, USDA-ARS, University of Minnesota, St Paul, USA; MJ, M. Jacquet, Vilmorin, Centre de Recherche de la Costière, Ledenon, France; NHK, N. Hamini-Kadar, Université d’Oran es-Sénia, Oran, Algeria; QM, Q. Migheli, Dipartimento di Protezione delle Piante, Universita degli Studi di Sassari, Sassari, Italy; SY, S. Yücel, Ministry of Agriculture, Plant Protection Institute, Adana, Turkey; TK, T. Katan, Hebrew University of Jerusalem, Jerusalem, Israel.

  4. bIntergenic spacer (IGS) types are defined in Table 3.

  5. cIsolates with the same number or the same letter belong to the same vegetative compatibility group (VCG); HSI, heterokaryon self-incompatibility; sm, single-member VCG; Nd, not determined because nit mutants could not be obtained.

  6. d+ indicates that a 990 bp amplicon was obtained with primers designed to amplify part of the secreted in xylem (SIX1) gene; − indicates that no amplicon was detected.

  7. ePercentage of diseased and dead plants. Nd, not determined.

Fusarium communeNH100101AlgeriaNorthwestOran(NHK)Tomato200678A8040
NH800108AlgeriaNorthwestSidi Bel Abbes(NHK)Tomato200778A90Nd
NH900109AlgeriaNorthwestSidi Bel Abbes(NHK)Tomato200778HSI090
NH1000110AlgeriaNorthwestSidi Bel Abbes(NHK)Tomato200778Nd0100
NH1400114AlgeriaNorthwestSidi Bel Abbes(NHK)Tomato200778A1040
NH1500115AlgeriaNorthwestSidi Bel Abbes(NHK)Tomato200778A070
NH1700117AlgeriaNorthwestSidi Bel Abbes(NHK)Tomato200778A3020
NH1800118AlgeriaNorthwestSidi Bel Abbes(NHK)Tomato200778A100Nd
F. redolensNH2900129AlgeriaNorthAlger(NHK)Tomato200880HSI40100
F. oxysporumNH1100111AlgeriaNorthwestSidi Bel Abbes(NHK)Soil200720B1050
 NH2100121AlgeriaNorthwestSidi Maarouf(NHK)Tomato200774sm3050
 NH2200122AlgeriaNorthwestSidi Maarouf(NHK)Tomato20075HSI2030
 NH2300123AlgeriaNorthwestAin Temouchent(NHK)Tomato200879C5020
 NH2400124AlgeriaNorthwestAin Temouchent(NHK)Tomato200879C1060
 NH2500125AlgeriaNorthwestAin Temouchent(NHK)Tomato200879C4080
 NH2600126AlgeriaNorthwestAin Temouchent(NHK)Tomato200879C3080
 NH2700127AlgeriaNorthwestSidi Maarouf(NHK)Tomato20085HSI3030
 NH2800128AlgeriaNorthwestSidi Maarouf(NHK)Tomato20085HSI060
 Fol1500504Tunisia   Tomato1968250030+90Nd
 KH6600506TunisiaCôte FOM-05 (KH)Tomato2005250030+2080
 KH6700514TunisiaCôteChott MeriamFo4-ch-06 (KH)Tomato200667Nd040
 KH6800505TunisiaCentre Fol-04 (KH)Tomato2004250030+70Nd
 KH6900507TunisiaEastMonastirFoB1·05 (KH)Tomato2005250030+100Nd
 KH7000508TunisiaEastMonastirFoB2-05 (KH)Tomato2005250030+80Nd
 KH7100515TunisiaCentre Fo1 B3 08 (KH)Tomato200825HSI+90Nd
 KH7200516TunisiaCentre FoB3·08 (KH)Tomato2008250030+90Nd
 KH7300509TunisiaSouth Fo2-05 (KH)Tomato2005250090100Nd
 KH7400510TunisiaSouth Fo3-05 (KH)Tomato20052500903040
 KH7500511TunisiaSouth Fo4-05 (KH)Tomato20052I3040
 KH7600512TunisiaCentre Fo1SB-05 (KH)Tomato200581J4020
 KH7700513TunisiaCentre Fo2SB-05 (KH)Tomato200581J1010
 KH7800539TunisiaSouth Fo4-07 (KH)Tomato20072I70Nd
 Forl7100459CreteLerapetraLasithiForl78 (DV)Tomato198523Nd80Nd
 Fox700480Greece  F623 (KE)Tomato200120Nd80Nd
 Forl9000481TurkeyEast Mediterranean Mersin-1 (SY)Tomato200632009480Nd
 Forl9100517TurkeyEast Mediterranean Ceyhan-1 (SY)Tomato200725009090Nd
 Forl9300518IsraelSouthArava ValleyForl (MR)Soil2008250090100Nd
 Fox400502Spain AguilasAFA 241-2 (MJ)Tomato20072500923040
 Fox500503Spain AguilasFORLaguilas (MJ)Tomato20002500918090
 Forl7200537Spain  FORL (ILN)Tomato<199025Nd80Nd
 Fol2000538SpainSouth FOL (ILN)Tomato1995250030+80Nd
 Forl7300501ItalySardinia S6 (QM) 199825009090Nd
 Forl7400487ItalySardinia S13 (QM) 199825009070100
 Forl7500488ItalySardinia S7 (QM) 1998250090100Nd
 Forl7600489ItalySardinia DP95 (QM) 1998250090100Nd
 Forl7700490ItalySardinia DP83 (QM) 1998250091100Nd
 Forl7800491ItalySardinia DP93 (QM) 199825009290Nd
 Forl7900492ItalySicily DP282 (QM) 1997250096100Nd
 Forl8000485ItalySardinia S16 (QM) 199825009070Nd
 Forl8100493ItalySardinia S3 (QM) 1995250090020
 Forl8200484ItalySardinia S9 (QM) 19982500905030
 Forl8300494ItalySardinia S10 (QM) 19982500904020
 Forl8400495ItalySardinia S11 (QM) 1998250090100Nd
 Forl8500496ItalySardinia S12 (QM) 199725009070Nd
 Forl8600486ItalyCalabria DP61 (QM) 1998250091100Nd
 Forl8700497ItalyCalabria DP63 (QM) 19982500915020
 Forl8800498ItalySicily DP238 (QM) 1996/9725009160Nd
 Forl8900499ItalySicily DP31 (QM) 1996/972500910100
 Fol2700462FranceSouthwestPerpignan Soil1980250030+100Nd
 Fol2800463FranceSouthwestPerpignan Soil1980250030+90Nd
 Fol2900464FranceSouthwest  Tomato1981250030+100Nd
 Fol3000467France Perpignan Tomato1985230031+100Nd
 Fol3200468FranceNorthwestBrest Tomato1985230031+80Nd
 Forl200469FranceSouth (AM)Tomato199125009190Nd
 Forl500470France  (AM)Tomato199325009150Nd
 Forl1000472France  (AM)Tomato199532009430Nd
 Forl1500474FranceSouth (AM)Tomato19943Nd50Nd
 Forl1700473FranceSouthwestAlenya Tomato1984250090100Nd
 Forl1800465FranceSouthAlenya Tomato1984250090100Nd
 Forl2100466FranceSouthAvignon-Berre Tomato198425009140Nd
 Forl3000479FranceSouthwest  Tomato1997320094100Nd
 Forl4500475FranceNorthwestBrest Tomato1998320094100Nd
 Forl9200476FranceSouth Forl ref 41 (AB)Tomato1995/9825Nd70Nd
 Fox00500FranceSouth Fo tomate (AB)Tomato2006250091100100
Reference strains
 F. oxysporumFo4700047FranceSouthChâteaurenard Soil197822HSI00
 F. oxysporum f. sp. lycopersiciFol800046FranceCentreVillefranche s Saône Tomato197123HSI+100100
Fol2400460FranceNorthwestNantes Tomato1977230031+100Nd
Fol2600461France Nantes Tomato1977230031+100Nd
Fol100*00521USAOhio OSU451B (LRG)  230031+NdNd
Fol101*00525   Fol R (TK)  250030+NdNd
Fol102*00526   Fol m (TK)  250030+NdNd
Fol103*00522USAArkansas Fol MM59/5 (TK)  250032+NdNd
Fol104*00523USAArkansas Fol MM66/6 (TK)  250032+NdNd
Fol105*00524USANorth Carolina Fol RG1/5 (TK)  200033+NdNd
Fol106*00527   Fol MN 27/9 (TK)  200033+NdNd
 F. oxysporum f. sp. radicis-lycopersiciForl1200045Morocco   Tomato1995250091100100
Forl100*00519USAFlorida CL7 (LRG) 1995320098100Nd
Forl101*00520USAFlorida PB114 (LRG) 199620009990Nd
Forl102*00528   0-1090 (TK)  250090NdNd
Forl103*00529   Forl II D (TK)  250090NdNd
Forl106*00530   C544 (TK)  250091NdNd
Forl107*00531   C758 (TK)  250091NdNd
Forl120*00532   CRNK676 (TK)  250092NdNd
Forl121*00533   CRNK678 (TK)  250092NdNd
Forl122*00534   01150-6 (TK)  320094NdNd
Forl123*00535   01152-31 (TK)  320094NdNd
Forl129*00483Israel  C204/3 (TK)  600093NdNd
Forl132*00536   C623/35  250096NdNd

Molecular characterization

Fungal DNA was extracted from cultures on potato dextrose agar (PDA) using a rapid minipreparation procedure (Edel et al., 2001). Some isolates were identified at the species level using a molecular strategy based on sequences of a portion of the translation elongation factor EF-1α gene (Geiser et al., 2004) and sequences of the internal transcribed spacer (ITS) region of the ribosomal DNA (rDNA).

Part of the EF-1α gene was amplified by PCR using the primers ef1 (5′-ATG GGT AAG GA(A/G) GAC AAG AC-3′) and ef2 (5′-GGA (G/A)GT ACC AGT (G/C)AT CAT GTT-3′) (O’Donnell et al., 1998b) in a final volume of 25 μL containing 1 μL of DNA, 0·1 μm of each primer, 150 μm dNTP, 3 U Taq DNA polymerase and PCR reaction buffer. Amplifications were conducted in a Mastercycler (Eppendorf) with an initial denaturation of 7 min at 95°C followed by 38 cycles of 60 s denaturation at 95°C, 75 s annealing at 57°C, 60 s extension at 72°C and a final extension of 10 min at 72°C. Presence of PCR products was confirmed by gel electrophoresis. EF-1α amplicons were sequenced by Beckman Coulters Genomics using the two PCR primers as sequencing primers. For each PCR product, sequences from both strands were assembled using SeqMan 6·0 (DNASTAR Lasergene, GATC Biotech). Sequence identities were determined using blast analysis from the National Center for Biotechnology Information (NCBI) available online. Sequences were aligned and compared with the Kimura’s two parameters distance model and the neighbour-joining (NJ) method using the program SeaView (Gouy et al., 2010). The topology of the resulting tree was tested by NJ bootstrapping with 1000 re-samplings of the data.

The ITS region was amplified by PCR using primers ITS1F (Gardes & Bruns, 1993) and ITS4 (White et al., 1990) in a final volume of 25 μL containing 1 μL of DNA, 0·5 μm of each primer, 150 μm dNTP, 3 U Taq DNA polymerase (Q-Biogen) and PCR reaction buffer. Amplifications were conducted in a Mastercycler with an initial denaturation of 3 min at 94°C followed by 35 cycles of 1 min denaturation at 94°C, 1 min annealing at 50°C, 1 min extension at 72°C and a final extension of 10 min at 72°C. Presence of PCR products was confirmed by gel electrophoresis. For each PCR product, sequences from both strands were assembled and sequence identities were determined as above.

All isolates were characterized by restriction fragment length polymorphism (RFLP) analysis of the IGS region as previously described (Edel et al., 1995, 1997). A 1·7 kb fragment of the IGS was amplified by PCR with oligonucleotide primers PNFo (5′-CCC GCC TGG CTG CGT CCG ACT C-3′) and PN22 (5′-CAA GCA TAT GAC TAC TGG C-3′) and digested with seven restriction enzymes: AluI, HaeIII, HinfI, MspI, RsaI, ScrFI and XhoI. Each isolate was assigned to an IGS type defined by the combined restriction patterns obtained with the seven enzymes (Edel et al., 1997, 2001; Lori et al., 2004; Abo et al., 2005).

All isolates were tested for the presence of the SIX1 gene specific to forma specialis lycopersici by PCR using primers P12-F2B (5′-TAT CCC TCC GGA TTT TGA GC-3′) and P12-R1 (5′-AAT AGA GCC TGC AAA GCA TG-3′) (van der Does et al., 2008). Amplifications were performed in a final volume of 25 μL containing 1 μL of DNA, 0·2 μm of each primer, 200 μm dNTP, 3 U Taq DNA polymerase and PCR reaction buffer. Amplifications were conducted in a Mastercycler with an initial denaturation of 2 min at 94°C followed by 30 cycles of 45 s denaturation at 94°C, 45 s annealing at 64°C, 45 s extension at 72°C and a final extension of 10 min at 72°C. Strain Fol8 of F. oxysporum f. sp. lycopersici was used as positive control in each PCR experiment. The PCR products were checked by electrophoresis. Strains were scored as SIX1+ if amplification of their DNA yielded a 990 bp fragment or as SIX1− if no amplification product was obtained.

Vegetative compatibility testing

Vegetative compatibility was determined by pairing complementary nitrate non-utilizing (nit) mutants derived from each strain, as previously described by Correll et al. (1987). Briefly, mutants were first generated for each strain on PDA-potassium chlorate medium (PDA 30 g L−1, KClO3 15 g L−1) and then classified as nit1, nit3 or nitM based on their phenotype on minimal medium (MM) containing one of three different nitrogen sources (nitrate, nitrite and hypoxanthine). The composition of MM was (per litre distilled water): sucrose 30 g, KH2PO4 1 g, MgSO4.7H2O 0·5 g, KCl 0·5 g, FeSO4.7H2O 10 mg, NaNO3 2 g, agar 20 g and trace elements solution 0·2 mL. The trace element solution contained (per 95 mL distilled water): citric acid 5 g, ZnSO4.7H2O 5 g, Fe(NH4)2.6H2O 1 g, CuSO4.5H2O 0·25 g, MnSO4.H2O 50 mg, H3BO4 50 mg and NaMoO4.2H2O 50 mg.

For some isolates for which the different types of mutants could not be obtained, the concentration of potassium chlorate in the medium was increased from 15 to 30 g L−1. Mutants of different types were paired on MM in 5 cm Petri dishes and scored as compatible if dense aerial growth was formed at the point of contact between the two paired mutants, corresponding to the formation of a heterokaryon. Vegetative compatibility was determined by pairing complementary nit mutants derived from all 106 isolates in all pairwise combinations. For each pairwise combination, at least two pairings were performed: (i) a nit1 mutant of isolate A was paired with a nit3 or nitM mutant of isolate B, and (ii) a nit3 or nitM mutant of isolate A was paired with a nit1 mutant of isolate B. When two mutants formed a visible and robust heterokaryon, indicated by the presence of dense aerial mycelium, the corresponding strains were placed in the same VCG. Once the VCGs were identified, representative isolates of each group were checked for compatibility with the tester strains of FOL and FORL (Table 1). Control pairings between complementary mutants derived from the same parental strain were made to detect any strain that might be heterokaryon self-incompatible (Correll et al., 1987).

Pathogenicity tests

The 106 isolates of Fusarium together with six reference strains were tested for their pathogenicity on tomato cv. 63·5 Montfavet (Table 1). For each isolate, 10 12-day-old seedlings of tomato were inoculated by cutting the roots 7 mm below the crown and dipping the plants for 5 min in 5 mL of a suspension of 106 conidia mL−1 in liquid minimal medium (LMM, i.e. MM without agar). Plants were transplanted into perlite and the 5 mL of conidial suspension were poured onto the perlite around the crown. For each of the isolates tested, the 10 seedlings were inoculated with 10 independent conidial suspensions. Controls consisted of plants inoculated with reference strains of F. oxysporum Fo47 (non-pathogenic strain), Fol8 (f. sp. lycopersici) and Forl12 (f. sp. radicis-lycopersici), and with LMM instead of a conidial suspension as the negative control. Plants were incubated in a growth chamber at 24°C during the day (16 h) and 20°C at night. Three weeks after inoculation, the percentage of affected plants (dead plants together with plants showing dark brown lesions on crown and roots) was determined. For all of the isolates that induced a low percentage of disease, pathogenicity tests were performed a second time, but at a daytime temperature of 28°C instead of 24°C.


Species identification

Microscopic observations showed that numerous isolates originating from Algeria originally identified as F. oxysporum were atypical. Macroconidia appeared longer and thinner than those generally observed for F. oxysporum. Thus the 48 Algerian isolates were identified by their EF-1α sequences (GenBank accession nos HM584898 to HM584901 and JN000865 to JN222908). blast analysis using 99–100% sequence identity as the threshold for species limits indicated that 27 isolates were F. oxysporum, 18 were F. commune and three were F. redolens (Table 1). Sequence alignments revealed that the 18 EF-1α sequences of F. commune were identical, as were the three EF-1α sequences of F. redolens. Representative isolates of F. commune (MIAE00101 and MIAE00112) and of F. redolens (MIAE00129 and MIAE00131) were identified by their ITS sequences (GenBank accession Nos HM584894 to HM584897). Their identification at the species level was confirmed by blast analysis with 100% sequence identity with corresponding sequences. Evolutionary relationships of the 48 EF-1α sequences were inferred in a NJ tree, which revealed eight different sequence types among the 27 Algerian isolates of F. oxysporum (Fig. 1).

Figure 1.

 Bootstrapped neighbour-joining tree (Kimura two-parameter distance) inferred from 48 partial translation elongation factor EF-1α gene sequences of Fusarium spp. isolated from tomato or soil in Algeria. The origin of the isolates is given in Table 1. Bootstrap values (>75 %) are indicated on the internodes.

Molecular characterization

Oligonucleotide primers PNFo and PN22 successfully amplified a single DNA fragment of about 1·7 kb for each of the 130 isolates (106 tomato or soil isolates and 24 reference strains). The PCR products were digested individually with each of the seven restriction enzymes. Depending on the enzyme, two to 11 different restriction patterns were obtained among the 130 isolates of Fusarium (Table 2). In all, 18 combinations of patterns representing 18 IGS types were identified among the 130 isolates (Table 3). Among the Algerian isolates, the 17 isolates of F. commune and the three isolates of F. redolens were assigned to separate unique IGS types, 78 and 80 respectively. The 27 Algerian isolates of F. oxysporum were distributed in nine out of the 16 IGS types found within F. oxysporum in this study. Among these nine, six (5, 12, 16, 51, 74 and 79) were not found among tomato isolates originating from other countries. IGS type 25 was the most abundant type, which was represented by 55 of the F. oxysporum isolates analysed. It was not found in Algeria; however it was present and dominant in its neighbouring country Tunisia.

Table 2.   Restriction patterns of PCR-amplified intergenic spacer (IGS) fragments of Fusarium isolates
EnzymesaFragments (bp)
  1. aFor each restriction enzyme, the various patterns are indicated by numbers in the first column and the sizes in base pairs (bp) of the corresponding restriction fragments in the second column. Estimates of fragment sizes were determined by electrophoresis in 4–6% Nusieve 3:1 agarose (FMC) and by comparison with the molecular weight marker VIII (Roche Diagnostic) with measurements rounded to the nearest 5 bp. This was done for purposes of comparison among isolates; the values do not reflect absolute base pair fragment sizes. Restriction fragments less than 60 bp were not taken into consideration because they were not clearly resolved by electrophoresis. Numbers corresponding to the different restriction patterns follow those previously described (Abo et al., 2005; Edel et al., 1997; Edel et al., 2001; Lori et al., 2004).

  2. *Two restriction fragments of the same size (doublet).

 1850, 500, 270, 85
 2540, 500, 320, 270, 85
 7500, 370, 300, 235, 145, 135
 21690, 500, 270, 145, 85
 23760, 500, 375, 85
 1460, 145, 130, 115*, 100, 95*, 85, 80, 60
 6460, 145, 115*, 100*, 95*, 85, 80, 60
 7350, 150, 145, 115, 110, 100, 95*, 85, 75, 60
 8460, 145, 130, 115, 100, 95*, 85, 80, 70, 60
 11480, 145, 130, 115*, 100, 95*, 85, 80, 60
 13470, 175, 145, 130, 115, 100, 95*, 85, 80, 60
 19460, 175, 145, 130, 115, 100, 95*, 85, 80, 60
 21460, 195, 145, 130, 115, 100, 95*, 85, 60
 22460, 175, 145, 130, 115, 100, 95*, 85, 60
 28460, 145, 130, 115*, 100, 95, 85, 60
 29600, 145, 130, 115, 100, 95*, 85, 80, 60
 2700, 550, 210, 120, 80
 3670, 530, 330, 160
 5700, 580, 210, 120, 80
 20700, 550, 210, 130, 80
 22700, 550, 220, 195
 1560, 275, 200, 105, 95, 85, 75, 60*
 4420, 275, 200, 135, 105, 95, 85, 75, 60*
 6560, 275, 200, 105, 95, 75, 60*
 11560, 275, 200, 115, 105, 85, 75, 60*
 18640, 275, 200, 140, 105, 75, 60*
 19590, 275, 200, 105, 95, 85, 75, 60*
 27625, 275, 200, 120, 105, 75, 60*
 28500, 275, 200, 175, 105, 95, 75, 60*
 1610, 560, 400, 90
 31200, 400, 90
 4650, 560, 400, 90
 5520, 400, 320, 275, 90
 6610, 540, 400, 90
 9650, 540, 400, 90
 22560, 400, 375, 275, 90
 1460, 215, 180, 170, 135, 110, 90, 85, 60
 2460, 305, 180, 170, 135, 110, 90, 60
 9420, 305, 180, 170, 135, 110*, 60
 11460, 215, 180, 170, 135, 110*, 85, 60
 12460, 215, 180, 170, 135, 90, 85*, 60
 13470, 215, 180, 170, 135, 120, 90, 85, 60
 21460, 215, 180, 170, 135, 120, 85*, 60
 24490, 240, 145, 135, 125, 110*, 60
 11300, 370
Table 3.   Intergenic spacer (IGS) types and restriction patterns of Fusarium isolates revealed by restriction fragment length polymorphism analysis of PCR-amplified IGS sequences
SpeciesIGS typeaRepresentative isolatebRestriction patterns of amplified IGS fragments digested with enzymescGeographic distribution
Alu IHae IIIHinfIMspIRsaIScr FIXho I
  1. aIGS types represent the combination of patterns obtained with seven restriction enzymes. The numbers assigned to IGS types 2–60 follow those previously described (Abo et al., 2005; Edel et al., 1997; Edel et al., 2001; Lori et al., 2004). The restriction patterns are described in Table 2.

  2. bIsolates are described in Table 1.

  3. cNumbers designate the various patterns obtained for each restriction enzyme and follow those previously described (Abo et al., 2005; Edel et al., 1997; Edel et al., 2001; Lori et al., 2004; Schouten et al., 2004).

Fusarium oxysporum2MIAE005111121111Tunisia
3MIAE001321121112Algeria, France, USA
20MIAE005242121312Algeria, Spain, Italy, USA
23MIAE000461121312Crete, USA
25MIAE00045113211132Tunisia, Morocco, Turkey, Israel, Spain, France, USA
32MIAE0053421213122Turkey, France, USA
60MIAE00483129213132Israel, USA
67MIAE00514212220279212Algeria, Tunisia
F. commune78MIAE00101232822182291Algeria
F. redolens80MIAE00129773285242Algeria

Of the Algerian isolates characterized by their EF-1α sequence and their IGS type, a high correspondence was found between the two grouping methods: all isolates in the same IGS type possessed the same EF-1α sequence type, with the exception of isolates belonging to IGS type 3 (MIAE00132, MIAE00135 and MIAE145) that were differentiated into two EF-1α sequence types (Fig. 1). Conversely, three isolates (MIAE00132, MIAE00134 and MIAE00142) grouped in the same EF-1α sequence type could be differentiated by their IGS type.

PCR amplification with primers P12-F2B and P12-R1 yielded a 990 bp fragment of the SIX1 gene for the 10 reference strains of FOL (SIX1+) but no PCR product for the 13 reference strains of FORL (SIX1−). Among the 48 Algerian isolates and the 58 isolates originating from seven other Mediterranean countries, 13 isolates were scored as SIX1+. These included seven isolates previously identified as FOL, and six isolates recently collected in Tunisia (Table 1). The 93 remaining isolates classified as FORL or as unknown forma specialis were found to be SIX1−.

Vegetative compatibility grouping

One isolate of F. commune, two isolates of F. redolens and three isolates of F. oxysporum were found to be heterokaryon self-incompatible (HSI) (Table 1). Among the 17 remaining isolates of F. commune originating from Algeria, 14 isolates were assigned to the same VCG and three isolates could not be assigned to any VCG because of the lack of some types of mutants, despite repeated attempts to obtain them. The 24 Algerian isolates of F. oxysporum that were not HSI sorted into 13 VCGs. None of these VCGs corresponded to previously described groups; these new VCGs were designated by letters. VCG C included four isolates, VCG F and G each included three isolates, VCG B, D, E and H each included two isolates. The six remaining VCGs were each represented by a single isolate. The 12 Tunisian isolates of F. oxysporum that were not HSI sorted into four VCGs. VCG 0030 of FOL included six isolates also scored as SIX1+, VCG 0090 of FORL included two isolates, and two new VCGs each grouped two isolates. These new VCGs (I and J) were different from the new VCGs identified among the Algerian isolates.

Among the European isolates, seven French isolates collected before 1990 and one Spanish isolate collected in 1995 were assigned to VCG 0030 or 0031 of FOL; these eight isolates were also scored as SIX1+. All of the other European isolates, previously identified as FORL or from unknown forma specialis, were assigned to one of the following VCGs of FORL: 0090, 0091, 0092, 0094 and 0096, with the exception of five isolates for which the VCG could not be determined. Isolates in the same VCG also possess the same IGS type. All of the isolates in VCG 0090, 0091, 0092, 0096 of FORL and VCG 0030, 0031, 0032 of FOL were identified as IGS type 25. Isolates in VCG 0094 and 0098 of FORL were grouped in IGS type 32. Isolates in VCG 0099 of FORL and 0033 of FOL were grouped in IGS type 20. Finally, the isolate of VCG 0093 was placed in IGS type 60.

Pathogenicity tests

The references strains of FOL and FORL induced 90 to 100% of disease (including death) in the bioassay conducted at 24°C (Table 1). No symptoms were observed on uninoculated plants and plants inoculated with strain Fo47 known to be non-pathogenic. Among the 106 isolates tested, 11 isolates induced no disease symptoms at all, 24 isolates induced 10–30% of disease, 20 isolates induced 40–60% of disease and 50 isolates induced 70–100% of disease. All of the isolates that induced a low percentage of disease were tested again in a bioassay conducted at 28°C during the day instead of 24°C. Among the 53 isolates tested at 28°C, 41 isolates induced a higher percentage of disease than that induced at 24°C. Particularly among the Algerian isolates, the eight isolates that produced no disease at all at 24°C induced 60% (one isolate), 70% (one isolate), 90% (three isolates) and 100% (three isolates) of disease at 28°C. The disease symptoms observed on plants inoculated with F. commune and F. redolens were similar to those observed on plants inoculated with FORL: plants showed dark brown lesions on crown and roots and dead plants. The pathogenicity of the latter two species was recently reported on tomato in Algeria (Hamini-Kadar et al., 2010).


A surprisingly high level of genetic diversity was found among the isolates responsible for fusarium diseases in Algeria collected between 2005 and 2008. Besides several genotypes within F. oxysporum, it was also found that F. redolens and F. commune were pathogenic to tomato. Fusarium redolens is closely related to F. oxysporum. The latter two species were thought to be conspecific by Snyder & Hansen (1940) but they were recognized as two distinct species by Gerlach (1981). Later, the two species were clearly differentiated by molecular phylogenetic studies (Baayen et al., 1997; O’Donnell et al., 1998a); however, the two species remain difficult to distinguish using only morphological characters. Apart from the present study, only one other report of F. redolens causing disease on tomato plants was found, which was in the UK 40 years ago (Wilcox & Jackson, 1970). In the study here, the three isolates identified as F. redolens all originated from Algeria. They shared the same IGS type, and all three were more pathogenic in the bioassay conducted at 28°C than at 24°C. As two of them were HSI in the VCG test, there is no more information concerning the intraspecific diversity, which might be further resolved by amplified fragment length polymorphism (AFLP) analysis (Baayen et al., 2000). Whether the concept of forma specialis could be applied to these isolates of F. redolens is not known because their host specificity has not yet been demonstrated. This concept was previously applied to Fusarium isolates responsible for carnation diseases, including F. oxysporum f. sp. dianthi and F. redolens f. sp. dianthi (Baayen et al., 1997).

Eighteen Algerian isolates were identified as F. commune. This recently described species is closely related to, but phylogenetically distinct from, F. oxysporum (Skovgaard et al., 2003). This is the first report of F. commune isolated from and pathogenic on tomato. Because F. commune and F. oxysporum are morphologically similar (Skovgaard et al., 2003), previous cases of fusarium diseases on tomato may have been due to F. commune. The species has been isolated from soil and from different host plants including carrot, barley, corn, carnation, pea, pine and Douglas-fir (Skovgaard et al., 2003; Stewart et al., 2006). In this study, unlike F. redolens which was only found in the north near Alger, F. commune was collected only in the northwest region in several towns around Oran. The 18 isolates shared the same EF-1α sequence and comparison with other sequences available in GenBank showed that they were identical with several other isolates of F. commune originating from different host plants. However, at the present time, it is not known if there is any host specificity in F. commune. In addition to the shared EF-1α sequence, the 18 isolates of F. commune also shared the same IGS type. Also 14 of the 18 isolates grouped in the same VCG, corresponding to what could be called a clonal population (Klein & Correll, 2001). However, variability in the percentages of diseased plants suggests intraspecific diversity that could be assessed by AFLP. As with isolates of F. redolens, disease induced by the F. commune isolates was generally greater at 28°C than at 24°C, which might reflect adaptation to the warmer climatic conditions in Mediterranean countries.

The 27 remaining isolates originating from Algeria were identified as F. oxysporum, which showed a high level of intraspecific diversity, including nine IGS types and 13 VCGs. All 27 isolates were pathogenic on tomato and exhibited diverse levels of disease incidence ranging from 20–100%. As none of the isolates was scored as SIX1+, the 27 isolates can be assigned to FORL (van der Does et al., 2008). However, none of the Algerian isolates could be assigned to a known VCG of FORL. They were distributed in seven multiple-member VCGs and six single-member VCGs, but are all closely related genetically to previously characterized strains of F. oxysporum with which they share the same IGS type or at least common RFLP patterns. IGS 79 was the only novel type, but it matched known restriction patterns (Edel et al., 1997, 2001; Lori et al., 2004; Abo et al., 2005).

High diversity was also found among isolates of F. oxysporum originating from the neighbouring country Tunisia, with 14 isolates distributed in four IGS types. Although they were collected at the same period in both countries, the F. oxysporum isolates pathogenic on tomato were completely different. Several Tunisian isolates were identified as FOL both by their SIX1+ PCR and their VCG (0030). Among the other isolates identified as FORL, the typical VCG 0090 was found in Tunisia. IGS type 25, that includes VCG 0030 of FOL and VCG 0090 of FORL, was the dominant type among the Tunisian isolates; the dominance of IGS type 25 in the survey corroborates similar findings by Hibar et al. (2007). In the present study, IGS type 25 was also dominant among isolates collected in other countries including France, Spain and Italy. In these countries, the forma specialis FOL was mostly detected among the oldest isolates collected before 1985. The more recent isolates were all identified as FORL with a known VCG. In contrast to what was observed in both North African countries, the diversity in IGS types was much lower in European countries, with two IGS types found within FOL (IGS types 20 and 25) and four IGS types within FORL (IGS types 20, 25, 32, 60). Some isolates of FOL and FORL shared the same IGS type, whereas isolates of the same forma specialis were distributed in different IGS types, which may reflect their polyphyletic origins (Cai et al., 2003; O’Donnell et al., 2009). In addition to the different climatic conditions, the intensive cropping of tomato in European countries could have changed the balance among populations of F. oxysporum linked to tomato crops.

The combination of the two loci EF-1α and IGS made it possible to identify the Fusarium isolates to species and to characterize them at the intraspecific level. O’Donnell et al. (2009) have recently proposed the same strategy to type F. oxysporum isolates by sequencing both loci. Thus the number of types in the survey might increase by sequencing the IGS region. Concerning the identification of the formae speciales of F. oxysporum pathogenic on tomato, the results confirm that it is now possible to identify more rapidly FOL and FORL by combining the specific SIX PCR described by van der Does et al. (2008) with the rapid bioassay described in this study. Within formae speciales, although laborious, VCG testing remains a helpful method to characterize the genetic diversity of F. oxysporum populations. It also has several limitations, the first being the fact that it depends on reference strains that could be lost from collections. This study was unable to find representative strains of one VCG of FOL (0035) and one VCG of FORL (0097). Thus, it is possible that one of the new VCG identified here is VCG 0097.

Considerable diversity was found among the isolates of Fusarium pathogenic on tomato in Algeria, including species never (F. commune) and rarely (F. redolens) isolated from tomato, and several possibly new VCGs. Also, this diversity was different from the one observed in Tunisia, and both were different from that observed in European countries. The fact that numerous isolates collected in Algeria correspond to new genotypes never detected before and not detected in other countries may be due to the tomato cultivars used in Algeria. These might be different from the ones used in other countries and are diverse because farmers have long produced their own seeds using old cultivars; it is thus difficult to identify the different cultivars used. Given the difference between pathogenic isolates in Algeria and in other Mediterranean countries, it is hypothesized that newly emergent pathogenic forms may have evolved from local non-pathogenic populations or were previously misidentified. This question could be addressed by investigating the genetic diversity of soilborne populations in Algerian tomato growing areas. Genetic similarity between pathogenic and co-occurring non-pathogenic isolates may indicate that pathogenic isolates could be recent derivatives from local non-pathogenic populations.


The authors thank N. Hamini-Kadar and M. Begin for technical assistance, A. Buffière, K. Elena, N. Hamini-Kadar, K. Hibar, M. Jacquet, I. Larena Nistal, A. Moretti, Q. Migheli, M. Raviv, D. Vakalounakis and S. Yucel for providing fungal strains and C. Alabouvette for helpful comments on the manuscript.