Pathogenicity of Fusarium spp. to chickpea seed and seedlings (Cicer arietinum L.)

Abstract

Damping-off disease can cause severe stand loss and reduce seedling vigour of kabuli chickpea (Cicer arietinum L.). While Pythium spp. are thought to be the primary cause of damping-off in chickpea, diversity and pathogenicity of other genera have not been widely explored. In surveys of three Montana fields of chickpea affected by damping-off, Fusarium spp. were isolated from 83% of the seeds tested, and were more common than the other taxa recovered, namely Pythium and Rhizopus. Fusarium spp. were identified to species using morphological characters and sequence analysis of two loci, the internal transcribed spacer (ITS) region of ribosomal DNA and the translation elongation factor (TEF) 1-α gene. Consistently identified species were F. oxysporum Schltdl. and F. redolens Wollenw. While discrepancies between identification methods made the identity of other isolates unclear, Fusaria infecting chickpea were diverse and varied across sites. The pathogenicity of eight isolates to kabuli chickpea was tested across a gradient of soil moisture and initial propagule density. All isolates were able to cause disease on kabuli chickpea seeds and seedlings. Effects of soil moisture levels on disease incidence varied among isolates, even isolates whose TEF 1-α sequences were 100% identical. This indicates that strains within a species can vary in pathogenicity. Additionally, most isolates caused more severe disease symptoms on cotyledons than roots.

Résumé

La fonte des semis peut causer de lourdes pertes au champ et réduire la vigueur des semis des pois chiches Kabuli (Cicer arietinum L.). Bien que Pythium spp. semblent être la cause première de cette maladie chez le pois chiche, la diversité et la pathogénicité d'autres genres n'ont pas été évaluées extensivement. Dans des études menées dans trois champs de pois chiches du Montana touchés par la fonte des semis, on a isolé Fusarium spp. de 83 % des semences testées et ceux-ci étaient plus courants que les autres taxons décelés, entre autres, Pythium et Rhizopus. Fusarium spp. ont été identifiés à des espèces qui utilisent les caractères morphologiques et l'analyse de séquence de deux locus, la région de l'espaceur transcrit interne (ETI) de l'ADN-r et le gène 1-α du facteur d'élongation de la traduction (FET). Les espèces qui y étaient identifiées invariablement incluaient F. oxysporum Schltdl. et F. redolens Wollenw. Tandis que des différences entre les méthodes d'identification rendaient l'identité des autres isolats incertaine, les fusariums infectant les pois chiches étaient différents et variaient d'un site à un autre. La pathogénicité de huit isolats de pois chiches Kabuli a été testée le long d'un gradient d'humidité de sol et de densité initiale de propagules. Tous les isolats ont pu transmettre la maladie aux semences et aux semis de pois chiches. Les effets des taux d'humidité du sol sur l'incidence de la maladie variaient chez les isolats, même chez ceux dont les séquences du gène 1-α du FET étaient identiques à 100 %. Ceci indique que des souches appartenant à une espèce peuvent varier quant à la pathogénicité. De plus, la plupart des isolats ont provoqué des symptômes plus graves sur les cotylédons que sur les racines.

Keywords:

• damping-off
• Fusarium oxysporum
• Fusarium redolens
• kabuli chickpea
• root infection
• seed infection

Mots clés:

• Fonte des semis
• Fusarium oxysporum
• Fusarium redolens
• infection de la semence
• infection racinaire
• pois chiche Kabuli

Introduction

In North America, chickpea is grown primarily in California, Idaho, Montana, North Dakota, Oregon, South Dakota, and Washington, in the provinces Alberta and Saskatchewan, as well as parts of Mexico. Seed rot, pre-emergence damping-off and post-emergence damping-off (collectively called ‘damping-off’) can cause substantial stand loss and decreased plant vigor (Haware, 1998) and yields (Kaiser & Hannan, 1983; Trapero-Casas & Jiménez-Diaz, 1986; Trapero-Casas et al., 1990). This disease is characterized primarily by pre-emergence seed decay and cotyledon degradation, as well as discoloration and necrosis of roots (Kaiser & Hannan, 1983). Damping-off of chickpea has been primarily attributed to Pythium spp., in particular Pythium ultimum Trow (Kaiser & Hannan, 1983; Trapero-Casas et al., 1990; Haware, 1998), although Fusarium solani Mart. has also been reported to be associated with kabuli chickpea damping-off (Kaiser & Hannan, 1983; Brick et al., 1998; Haware, 1998; Hammon & Berrada, 2001; Chen et al., 2004). As with many seedling diseases, chickpea damping-off appears to be a function of environmental conditions (cool, wet weather) and inoculum density of the pathogens. Although studies testing seed treatments for damping-off in field settings have been conducted in the Great Plains states of the USA (e.g. Hammon & Berrada, 2001; Chen et al., 2004; Leisso et al., 2009), little published work exists addressing the identity of pathogens that cause damping-off in chickpea.

Both Fusarium spp. and Pythium spp. were isolated from kabuli chickpea affected by damping-off in Montana field trials at three locations in 2007 (Leisso et al., 2009). While root rot in chickpea can be caused by F. solani and F. oxysporum Schltdl. (Trapero-Casas & Jiménez-Díaz, 1985; Nene & Reddy, 1987; Nene & Sheila, 1996; Chen, 2008; Hasanzade et al., 2008), little is known about the ability of Fusarium spp. to cause damping-off of chickpea, with the exception of F. solani (Kaiser & Hannan, 1983; Chen et al., 2004) and F. equiseti (Corda) Sacc. (Goswami et al., 2008). Past research has indicated Fusarium spp. are most pathogenic in dry soils, and are frequently reported as important disease agents in dry conditions (Stover, 1953; Cook & Papendick, 1972; Cook, 1981).

The objectives of this study were to assess which pathogens were associated with damping-off of kabuli chickpea in Montana, determine the pathogenicity of the most commonly isolated Fusarium spp., and to determine how soil moisture and Fusarium spp. inoculum density affect disease incidence and severity.

Materials and methods

Recovery and identification of Fusarium spp. isolates

To assess the incidence and diversity of fungi associated with damping-off in chickpea, decaying kabuli chickpea (Cicer arietinum ‘Sierra’) seeds and seedlings were collected three to four weeks after planting from control field plots where no fungicide seed treatment was applied. These plots were part of a study of seed treatments to prevent damping-off (Leisso et al., 2009). Diseased seeds and seedlings were collected from three locations in Montana – Bozeman, Huntley, and Sidney – on 16 May, 17 May and 23 May 2007, respectively. During the three weeks following planting and prior to seed and seedling collection, Bozeman soil temperatures averaged 14.1 °C and plots received 42 mm of precipitation. At Huntley, the average soil temperature during the three weeks following planting was 15.4 °C and plots received 7 mm of precipitation. At Sidney, mean soil temperature was 14.2 °C, and plots received 49 mm of precipitation. Two rotting seeds or seedlings were collected from each of the six control treatment plots for a total of 12 seeds per location; seedlings showed evidence of germination such as elongation of the radical. Seeds or seedlings were collected from row sections showing low seedling emergence. Each seed or seedling was washed to remove excess soil, surface disinfested in 0.013% NaOCl for 30 s, rinsed three times with sterile de-ionized water, and divided into eight pieces; four pieces were plated on potato dextrose agar (PDA) amended with 50 ppm streptomycin and PARP (17 g corn meal agar, 5 ppm pimaricin, 250 ppm ampicillin, 10 ppm rifampin and 100 ppm pentachloronitrobenzene L⁻¹). Each dish contained two pieces from each of the two cotyledons. Hyphal tips were subcultured to PDA and initially identified using morphological criteria (Van der Plaats-Niterink, 1981; Nelson et al., 1983; Barnett & Hunter, 1998; Leslie & Summerell, 2006).

For common genera of fungi/oomycetes (defined here as isolated from more than 10% of seeds per site), rates of infection and coinfection were calculated based on morphological identifications. Fusarium species were sequenced (see below) and identified with a BLAST search. To determine the composition of Fusarium species within each site, morphologically distinct isolates were identified using BLAST based on the internal transcribed spacer region (ITS) rDNA sequences. The number of unique isolates (with different ITS sequence) in each seed was used to calculate the richness and evenness of Fusarium species for each site using Simpson's index of diversity (Simpson, 1949). A sub-set of BLAST searches returned multiple species that matched the sequence to the same degree. These samples could represent one or several species. Community compositions of each site were compared using the later assumption. Differences among sites in frequency of genera/ITS isolates were tested using chi-square statistics.

Molecular identification

Eight Fusarium isolates were selected for characterization as representing the most commonly isolated morphological type from each site. Single spores of isolates were obtained by performing serial dilution plating of spores and transferring a single germinating spore to PDA. DNA was extracted from fungal mycelia using the E.Z.N.A. Fungal DNA kit (Omega Bio-tek, Doraville, GA) according to manufacturer's instructions. Approximately 250 mg of mycelia were aseptically removed from a one-wk-old colony of each isolate grown on PDA. Mycelia were placed in a sterile mortar, frozen with liquid nitrogen, and ground into fine powder. Approximately 100 mg of the pulverized mycelia was used for DNA extraction.

Primers ITS1F (Gardes & Bruns, 1993) and ITS4 (White et al., 1990) were used to amplify the rDNA ITS region. The 20 μL reaction mixture consisted of 1X GoTaq Master mix (Promega Corporation, Madison, WI) 1.25 μM primer ITS 1F, 1.25 μM primer ITS 4, and 85 ± 20 ng of extracted DNA. PCR conditions were denaturation at 95 °C for 5 min, followed by 35 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1 min with a final extension step at 72 °C for 7 min. The amplification was performed using an iCycler Thermal Cycler (Bio-Rad Laboratories Ltd, Hemel Hempstead, UK). Amplified DNA was purified using QIAquick PCR purification kit (Qiagen Sciences Inc., Germantown, MD) and directly sequenced by Functional BioSciences of Madison, WI, using forward primer ITS1F.

The translation elongation factor (TEF) 1-α gene, which has the highest phylogenetic utility for identification of Fusarium spp. was also amplified and sequenced. Primers used were EF1 and EF2 (Geiser et al., 2004) and the reaction conditions were the same as described for amplification of the ITS region, with the exception of an annealing temperature of 53 °C. Amplified DNA was purified using QIAquick PCR purification kit (Qiagen Sciences Inc., Germantown, MD) and sequenced by Functional BioSciences of Madison, WI using forward primer EF1.

ITS region and TEF 1-α sequences were compared to sequence collections on GenBank (Benson et al., 2005) using BLAST (Altschul et al., 1990) algorithm ‘blastn.’. Translation elongation factor sequences were queried to the FUSARIUM-ID server at http://fusarium.cbio.psu.edu (Geiser et al., 2004) which contains a BLAST search tool. Sequences generated in this study have been deposited in the GenBank database under accession numbers listed in Table 4.

Effects of soil moisture and inoculum density of Fusarium spp. on disease incidence and severity

Eight Fusarium isolates were tested for pathogenicity to chickpea seeds across a range of soil moisture and inoculum densities in a growth chamber experiment. A factorial treatment design was used with four moisture levels (10, 20, 30 and 40%, v/v [volume water to volume sand]) and five inoculum levels (0, 10, 100, 1000 and 10 000 propagules g⁻¹ of sand) for a total of 20 treatments. Published data relating soil volumetric water content and water/matric potential (Blank & Young, 1992) was used to parameterize the van Genuchten model (1980) and estimate the matric potential of our moisture treatments. The 0.4, 0.3, 0.2 and 0.1 volumetric water content treatments were estimated to have matrix potentials of 0, −0.00038, −0.00157, and −0.00677 MPa. Moisture and inoculum levels were chosen based on results of preliminary experiments and previous legume studies (Kraft, 1975; Kaiser & Hannan, 1983; Trapero-Casas & Jiménez-Díaz, 1985). Sand (silica sand 20/30, Lane Mountain Co., Valley, WA) was chosen as the growth medium due to its neutral pH and inert properties (Dhingra & Sinclair, 1985).

Fusarium inoculum was produced in 25% potato dextrose broth (PDB, one-quarter of manufacturer's recommendations for g L⁻¹, EMD Chemicals USA, Rockland, MA) by inoculating with three 7-mm diameter circular agar plugs cut from the edges of actively growing three to five day-old-cultures and filtering broth cultures through four layers of sterile cheesecloth after 15–20 days of growth. Conidia, chlamydospores, and hyphal fragments in solution were quantified using a haemacytometer (Hausser Scientific, Horsham, PA). Prior to sand inoculation, 250 μL of propagule solution was plated onto PDA and the percentage of conidia, chlamydospores, or hyphal fragments that were germinating and actively growing was quantified after 12 h incubation at 25 °C in the dark. Volumes of propagule solutions added to sand were adjusted by adding sterile tap water so that the inoculum potential was proportional to the number of viable propagules.

Sand moisture levels and inoculum concentration were achieved by aseptically mixing 70 g autoclaved sand (autoclaved twice, 24 h between each cycle) in individual sterile Petri dishes with sterile distilled water containing the appropriate quantity of inoculum using a flame-sterilized glass rod. Ten seeds of kabuli chickpea ‘Sierra’ with low resistance to seed-seedling disease were placed in each dish after treatments had been established. Seeds were surface-disinfested in 0.013% NaOCl for 3 min, rinsed for 30 s in 70% ethanol, and then rinsed thrice in sterilized de-ionized water and dried overnight (at least 10 h at room temperature ∼23 °C) on autoclaved paper towels in a laminar flow hood prior to use. Dishes were sealed with Parafilm to prevent moisture loss. No additional water was added to the dishes during the experiment. Dishes were placed in a growth chamber set at 15 °C/10 °C day/night regime, an optimal temperature range for chickpea germination (Ellis et al., 1986), and a 14 h photoperiod. Three replicates of the 20 treatments were arranged in a completely randomized block design with one replicate per block. The experiment was performed twice for each isolate.

Disease incidence and severity was measured 10 days after planting. The number of diseased seeds or seedlings was recorded, and cotyledon and root disease severity assessed visually according to percentage of necrotic cotyledon, epicotyl and hypocotyl tissue on a 1–5 scale modified from rating scales for legume seed and root diseases (Kraft, 1975; Trapero-Casas & Jiménez-Díaz, 1985; Peters & Grau, 2002) (Table 1).

Table 1.  Cotyledon and root disease severity rating scale used in evaluating disease symptoms of kabuli chickpea seeds and seedlings infected by Fusarium spp

Cotyledon rating scale	Description	Root rating scale	Description
1	Seed germinated, no disease apparent on cotyledons	1	Seed germinated, no disease apparent on roots
2	Seed germinated, < 25% of cotyledons necrotic or covered by mycelia	2	Seed germinated, < 25% of roots necrotic
3	Seed germinated, 25–50% of cotyledons necrotic or covered by mycelia	3	Seed germinated, 25–50% of roots necrotic
4	Seed germinated, 25–50% of cotyledons necrotic or covered by mycelia	4	Seed germinated, 50–75% of roots necrotic
5	Seed failed to germinate (due to disease) or 75–100% of cotyledons necrotic	5	Seed failed to germinate (due to disease) or 75–100% of roots necrotic

To complete Koch's postulates, five surface-disinfested, diseased seeds from each trial were plated on PDA for the re-isolation and identification of the inoculated-Fusarium spp.

Statistical analyses

Data analysis was performed using Statistical Analysis System (version 9.1, SAS Institute, Cary, NC). Residuals for each experimental parameter were examined for normality and homogeneity of variances. Assumptions of normality were examined using the Shapiro-Wilk's test (Shapiro & Wilk, 1965) performed with PROC UNIVARIATE. Homogeneity of variance between isolates and between experiments was examined using Hartley's F-max test (Hartley, 1950). Data were log transformed for analysis as necessary. Orthogonal polynomial statements were constructed in PROC GLM to compare disease incidence levels within the levels of soil moisture and inoculum density as well as to determine linear and quadratic trends in disease incidence as affected by soil moisture and propagule density. Multilinear regression using PROC REG was used to examine the linear relationship between moisture combined with inoculum density on disease incidence. Cotyledon and root disease severity data were analyzed using analysis of variance in PROC GLM. Letter groupings delineating equal means were generated by comparing results of the lsmeans pdiff option, which performs pair-wise comparisons of least squares means. Pearson's correlation coefficients were generated by comparing least square means of affected isolates of disease incidence, cotyledon disease severity and root disease severity. For all analyses, P < 0.05 was considered significant.

Results

Pathogen isolation and Fusarium spp. identification

Genera commonly isolated from kabuli chickpea seeds affected by damping-off were Fusarium, Pythium and Rhizopus. Multiple genera were isolated from 67% of seeds. Fusarium was the most common genus recovered, occurring on 83% of all seeds tested (Table 2). Pythium spp. and Rhizopus spp. were found on 50% and 47% of all seeds, respectively. Co-infection of Fusarium and Pythium was most common (28%), followed by Fusarium and Rhizopus (22%) and all three genera (14%). Interestingly, Pythium and Rhizopus were only isolated from the same seed once. Among seeds where a single genus was isolated, Fusarium was the most common (58%) while Pythium was less frequent (17%). Frequency of Fusarium recovery was consistently high across sites (χ² = 1.2 w/2 d.f., P > 0.5). However, incidence of genera in the Huntley site, where Pythium was not isolated from any seeds and Rhizopus was isolated from twice as many seeds compared with the other sites, differed significantly (χ² = 22.65 w/12 d.f., P < 0.05) from the other two sites, Bozeman and Sidney, where compositions of genera were similar (χ² = 2.26 w/5 d.f., P > 0.1).

Table 2.  Pathogens isolated from kabuli chickpea affected by damping-off in Montana field trials. Pathogens were isolated from 12 diseased seeds at each location

Location	Fusarium only	Pythium only	Rhizopus only	F+P*	F+R*	P+R*	F+P+R*	Total Fusarium	Total Pythium	Total Rhizopus
Bozeman	2	1	0	4	1	1	3	10	9	5
Huntley	4	0	3	0	5	0	0	9	0	8
Sidney	1	1	0	6	2	0	2	11	9	4
Total	7	2	3	10	8	1	5	30	18	17
% Seed	19.4	5.6	8.3	27.8	22.2	2.8	13.9	83.3	50.0	47.2
Note: * Co-infections of Fusarium and Pythium (F+P); Fusarium and Rhizopus (F+R); Pythium and Rhizopus (P+R); and Fusarium, Pythium and Rhizopus (F+P+R).

Communities of Fusarium species were similarly diverse across sites but differences in composition followed the same pattern seen in the analysis of common genera. Seven to eleven ITS species were isolated in each site and evenness (a measure of the degree to which communities are dominated by a single species) was similarly high across sites (Table 3). Distribution of Fusarium ITS differed significantly among sites (χ² = 48.3 w/28 d.f., P < 0.001). Site differences in distribution are driven by, in order of statistical importance: (1) F. equiseti was commonly isolated in Huntley but not from the other two sites, (2) F. redolens Wollenw. was isolated only from Sidney and (3) F. oxysporum was absent in Huntley. Distribution was similar between the Bozeman and Sidney isolates (χ² = 19.82 w/14 d.f., P = 0.071).

Table 3.  Diversity of Fusarium isolates across three sites in Montana

Site	Total no. seeds tested	Richness*	Evenness^†	Dominant ITS species^‡
Bozeman	9	9	0.84	Giberella avenacea (67), F. tricinctum (67), F. oxysporum (33)
Huntley	10	8	0.82	Giberella avenacea (80), F. equiseti (50), F. tricinctum (30)
Sidney	8	9	0.83	F. oxysporum (63), F. solani (38), F. redolens (20)
Notes: * Number of unique ITS sequences per site.
† Simpson's index of diversity. Values close to 1 indicate the abundance of species is evenly distributed across the community.
‡ Per cent of seeds from which species were isolated is in parentheses.

Of the Fusarium isolates obtained from each location, several morpho-species were associated with damping-off at high frequency, and from these species, a subset of eight isolates was selected for pathogenicity testing (Table 4). The morphological, ITS and TEF identifications agreed to the species level for isolates 2.6 and 3.5 only. The morphological and TEF/Fusarium-ID identifications agreed for isolates 1.7, 2.2 and 3.2. Accession numbers and closest identities to Fusarium spp. of the isolates according to sequence comparison of the ITS rDNA region and TEF 1-α gene sequence to GenBank using BLAST as well as the Fusarium-ID database are listed in Table 4.

Table 4.  Morphological and sequence-based identification of Fusarium spp. isolated from kabuli chickpea affected by damping-off in Montana

Isolatedesignation	1.3	3.5	3.2	4.6	DI	1.7	2.2	2.6
Location of origin	Bozeman	Bozeman	Huntley	Huntley	Huntley	Sidney	Sidney	Sidney
Morphological identification	F. solani	F. oxysporum	F. trincintum	F. culmorum	F. sambucinum	F. solani	F. solani	F. redolens
ITS sequence identification*
Accession no.	EU910075	EU910080	EU910079	EU910081	EU910082	EU910077	EU910076	EU910078
GenBank search return; accession number	Gibberella avenacea FA12; EU255802.1	Fusarium oxysporum isolate 25; EU839377.1	Fusarium acuminatum strain NRRL 54217; HM068325.1	Fusarium equiseti strain NRRL 13402; GQ505681.1	Fusarium equiseti strain NRRL 36478; GQ505743.1	Fusarium sp. SS-r3; GU797142.1	Fusarium oxysporum f. sp. phaseoli strain DM091019-1; HM756257.1	Fusarium redolens strain DH-A3; HM346540.1
Query coverage	100%	99%	100%	100%	100%	100%	99%	100%
E value	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
TEF sequence identification
^†Accession no.	EU910069	EU910072	FJ752547	EU910073	EU910074	EU910068	EU910070	EU910071
GenBank search return; accession number	Fusarium reticulatum var. negundinis isolate 15/2.3.1; DQ854864.1	Fusarium oxysporum f.sp. vasinfectum isolate IMI-338122; EU246573.1	Fusarium sp. K5225-2; AB294874.1	^§ Fusarium sp. DI; EU910074.1	^§ Fusarium sp. 4.6; EU910073.1	^‖ Fusarium sp. NRRL 44903; GU170620.1	^‖ Fusarium sp. NRRL 44903; GU170620.1	Fusarium redolens strain NRRL 52657; GU250584.1
Query coverage	95.42%	100%	100%	97%	97%	99%	99%	100%
E value	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
‡Fusarium-ID search return; isolate identification no.	Fusarium sp.; FC_01861_EF1a	Fusarium oxysporum species complex isolate NRLL 38608; MLST22	Fusarium tricinctum sp. complex isolate NRRL 36147; MLST2-a	No matching sequences	No matching sequences	Fusarium solani sp. complex isolate NRRL 32810; MLST5-m	Fusarium solani sp. complex isolate NRRL 32810; MLST5-m	Fusarium redolens; FD_01103_EF-1a
Query coverage	99.36%	99.83%	99.85%			100%	100%	99.845
Score	999.99	999.99	999.99			999.9	999.9	999.99
Notes: * ITS sequence amplified using primer pair ITS 1F and ITS 4. Blastn nr/nt query.
† Transcription elongation factor (TEF) 1-α sequence amplified using primer pair EF1 and EF2.blastn (nr/nt) query.
‡ TEF 1- α. Fusarium- ID query.
§ Closest identification other than isolates in this study had less than 25% coverage.
‖ Closest identification other than isolates in this study.

Effects of soil moisture and inoculum density of Fusarium spp. on disease incidence and severity

All Fusarium spp. isolates tested infected kabuli chickpea seeds and seedlings. Disease occurred at all moisture and inoculum density levels (Fig. 1). Low inoculum concentrations (10 propagules g⁻¹ of sand) caused disease for all of the isolates, except for isolate 3.5 (F. oxysporum) whose disease incidence was not different from zero at P < 0.05 (Fig. 1). Disease incidence varied among isolates. Maximum incidence ranged from 100% for isolates 1.3 and 3.2, approximately 80% for isolates 1.7, 2.2 and 4.6, and 40–50% for isolates 2.6, 3.5 and D.I.

Fig. 1. Effects of moisture, inoculum density on disease incidence in chickpea seedlings across eight Fusarium spp. isolates under controlled conditions in growth chamber experiments. Inoculum densities are in propagules per gram of dry sand.

With few exceptions, high levels of disease incidence were observed by Fusarium isolates at relatively low inoculum density (< 100 propagules g⁻¹ of sand). All of the isolates showed linear relationships between log propagule density and disease incidence at low inoculum densities (< 100 propagules g⁻¹ of sand) but at higher inoculum densities, disease incidence was unaffected or reduced (quadratic term, P < 0.05) for six of the eight isolates (Table 5 and Fig. 1).

Table 5.  Analysis of variance and significance of contrasts for linear and quadratic trends of damping-off incidence of kabuli chickpea cause by Fusarium spp. in relation to moisture and inoculum density in growth chamber trials

	Source of variation			Soil moisture		Propagules per g soil*
Isolate	Soil moisture	Inoculum density	Soil moist. × inoc. density	Linear trend	Quadratic trend	Linear trend	Quadratic trend
1.3	< 0.0001	< 0.0001	< 0.0001	< 0.0001	0.0033	< 0.0001	< 0.0001
3.5	0.0005	< 0.0001	0.71	0.0166	0.1873	< 0.0001	0.9471
3.2	< 0.0001	< 0.0001	0.01	< 0.0001	0.0045	< 0.0001	< 0.0001
4.6	< 0.0001	< 0.0001	0.09	0.0005	0.2851	< 0.0001	0.0002
D.I.	0.17	< 0.0001	0.77	0.8474	0.2014	< 0.0001	0.0017
1.7	< 0.0001	< 0.0001	0.08	< 0.0001	0.7391	< 0.0001	< 0.0001
2.2	< 0.0001	< 0.0001	0.44	< 0.0001	0.1586	< 0.0001	0.0155
2.6	0.334	< 0.0001	0.10	0.8022	0.0881	< 0.0001	0.2013
Notes: For all isolates, the source of variation for soil moisture had 3 df, inoculum density had 4 df, and soil moisture × inoculum density had 12 df. All linear and quadratic trend contrasts have 1 df. Significant contrasts are in bold.
*Log propagules per g soil used in analysis.

Moisture and inoculum density had highly significant and independent (interaction, P > 0.05) effects on disease incidence for most isolates (Table 5). The relationship between soil moisture and disease incidence varied among isolates. Five of the eight isolates (1.3, 1.7, 2.2, 3.2 and 4.6) showed an increase in disease incidence with increasing soil moisture, while isolate 3.5 had an inverse relationship, and isolates DI and 2.6 exhibited a neutral relationship (Table 5 and Fig. 1).

Cotyledons had higher severity ratings than roots for six of the eight isolates for data pooled over all moisture and inoculum levels (Table 6). Isolate 3.2 caused the most severe disease on cotyledons, followed by isolate DI. Isolate 3.2 caused the most severe disease on roots, followed by isolate 1.7.

Table 6.  Disease severity of cotyledons and roots infected by Fusarium spp. isolates and correlation between disease incidence and severity of damping-off of kabuli chickpea caused by Fusarium spp. in growth chamber trials

Isolate	Cotyledon*	Root*	Corr. incidence × disease severity (cotyledons) ^†	P-value	Corr. incidence × disease severity (roots) ^†	P-value
1.3	3.6 c	2.1  de	0.75	< 0.001	0.54	0.014
3.5	2.8 e	1.6 f	0.73	< 0.001	0.48	0.033
3.2	4.3 a	4.3 a	0.98	< 0.001	0.98	< 0.001
4.6	3.3 d	2.2 d	0.52	0.004	0.19	0.427
D.I.	4.0 b	2.6 c	0.73	< 0.001	0.50	0.009
1.7	3.2 d	3.0 b	0.65	0.002	0.88	< 0.001
2.2	2.3 f	2.4 cd	0.60	0.005	0.56	0.010
2.6	2.8 e	1.7 ef	0.45	0.047	0.31	0.180
Notes: * Numbers in a column for each isolate followed by the same letter are not significantly different at P < 0.05 (Fisher's LSD). Significant contrasts are in bold.
†Pearson's correlation coefficient (r).

Across treatments, conditions that facilitated infection were associated with increases in disease severity. Pearson's correlation coefficients indicated a positive relationship between disease incidence and severity for cotyledons and roots for all isolates except isolate 2.6 for cotyledons and isolate 2.6 and 4.6 for roots (Table 6).

Discussion

A range of Fusarium species appear to be common seed and seedling pathogens on chickpea. Diversity and incidence of infection was consistently high across sites, suggesting that disease was not affected by the range in abiotic conditions observed in this study. Furthermore, Fusarium was the most commonly isolated genus from seeds where only one genus was recovered. Rhizopus, generally considered as a secondary or opportunistic pathogen, was present in low numbers and usually (in 14 of 17 isolations) found in association with Fusarium. The pathogenicity of Rhizopus was not tested in this study.

Soil temperature and growing season precipitation may affect community composition of seed/seedling pathogens of chickpea. Huntley, which is warmer and drier than the other two sites that were wet and cool, showed a different composition of recovered genera and of Fusarium species. We also did not isolate Pythium from chickpeas at Huntley, probably due to the prevailing dry conditions, which is not conducive to Pythium infection (Agrios, 1997).

Disease severity was highest for isolates identified based on morphology as F. tricinctum (Corda) Sacc. and F. solani. This is consistent with previous reports of F. solani as a pathogen of chickpea (Kaiser & Hannan, 1983; Brick et al., 1998; Haware, 1998; Hammon & Berrada, 2001; Chen et al., 2004). Isolates obtained from Huntley had ITS sequences consistent with F. equisiti, which has been reported as a pathogen of several legume species including chickpea, where it was shown to cause seed decay and seedling disease (Goswami et al., 2008). Other studies have confirmed that diverse Fusarium species can infect chickpea seeds and seedlings. Bayraktar and Dolar (2009) isolated several of the same morphological species as in this study, and F. redolens and F. acuminatum Ellis & Everh. have been reported to infect chickpea (Gambhir et al., 2010). Furthermore, the positive relationship, exhibited by most but not all isolates, between disease incidence and soil moisture, suggests that optimal conditions for infection differ among Fusarium spp. Past research has indicated that Fusarium spp. are favoured by dry soils, and are frequently reported as important pathogens under dry conditions (Stover, 1953; Cook & Papendick, 1972; Cook, 1981). However, our results indicate that for more than half of the isolates, disease incidence increased with increasing soil moisture. Three of these isolates were identified as F. solani according to morphological characters (1.3, 1.7 and 2.2). Other isolates were non-responsive (DI and 2.6; F. sambucinum Fuckel and F. redolens) or inversely affected by soil moisture (3.5; F. oxysporum).

Extrapolating the results of the pathogenicity experiment to field conditions is difficult. However, biological activity in the soil is often a function of water potential rather than soil water content. This suggests that while similar relationships between soil moisture and pathogenicity are expected in field soils, observed levels of pathogenicity are predicted to occur at higher soil volumetric water contents in field soils compared with our sand moisture treatments since increasing loam and clay content results in higher matric potentials for a given soil water content.

The increase in disease incidence with soil moisture may be due to effects of soil moisture on seed exudates. Soil moisture increases the quantity of seed exudates released from the seed during germination. As seed moisture increases, chickpeas likely exude higher quantities of compounds such as carbohydrates and amino acids, similar to peas (Cook & Flentje, 1967). These nutrient-rich exudates may stimulate pathogen propagule germination (Schroth & Snyder, 1961) and serve as a food source for both Fusarium spp. and other microorganisms (Singh & Merhota, 1980; Nelson, 1990). This process may also explain the increased disease severity on cotyledons compared with roots. Since the cotyledons represent the greatest volume of seed tissues, they would be the primary tissue releasing exudates and the first site of infection for Fusarium. In this study, cotyledon disease symptoms were more severe than root symptoms, and disease severity on cotyledons increased with moisture increase, while it did not increase on roots.

Kabuli chickpeas are predisposed to infection during the early seedling growth stage due to low seed defences (Kumar et al., 1991), a juvenile tissue-based susceptibility in the first 48 h of germination (Kaiser & Hannan, 1983), and the nature of their seed exudates, which, as a nutrient source, can increase pathogen inoculum potential. High soil moisture levels, which stimulate high seed exudate release, combined with low soil temperatures at planting (Ellis et al., 1986), likely increase the length of the juvenile tissue susceptibility (Kaiser & Hannan, 1983). These factors create a favourable environment for the initiation of damping-off disease.

While results from morphological studies, molecular identification, pathogenicity, and response to soil moisture suggest that multiple Fusarium species can cause seed and seedling disease in chickpea, further research is still needed to determine the diversity and identity of the Fusarium species associated with damping-off in chickpea. The Fusarium spp. tested for pathogenicity in this study were identified as F. solani, F. oxysporum, F. tricinctum, F. culmorum, F. sambucinum and F. redolens using morphological characters. Molecular identification was attempted with two commonly used sequences, the ITS region of rDNA and the TEF gene. However, with the exception of F. oxysporum and F. redolens, identification was inconsistent with the different methods and sequences were submitted to GenBank as ‘Fusarium spp.’ due to this lack of consistency. Furthermore, strain diversity within species appears to be present. For example, isolates with identical TEF 1-α gene sequences (isolates 4.6 and DI; isolates 1.7 and 2.2) responded differently in the pathogenicity assay. This is consistent with results from Hasanzade et al. (2008) who tested 20 isolates of F. solani f. sp. pisi W.C. Snyder & H.N. Hansen on a susceptible cultivar of chickpea and found pathogenicity varied significantly between isolates.

Results of seed treatment studies in Montana in 2007 demonstrated that mefenoxam (marketed as Apron XL LS), which targets Pythium, was highly effective in controlling damping-off on chickpea, while seed treatments containing fluidoxonil (Maxim), targeting other fungi, were less effective (Leisso et al., 2009). This suggests that Pythium spp. (Kaiser & Hannan, 1983; Trapero-Casas et al., 1990) may be more important in causing damping-off in Montana than pathogens such as Fusarium spp. or that fludioxonil was not effective in controlling Fusarium. The interaction between Fusarium spp. and Pythium spp. in causing damping-off disease on chickpea was not explored in this study, but previous work has indicated that simultaneous infection by more one than one pathogen can cause higher levels of disease severity than either organism alone (Peters & Grau, 2002). We therefore recommend growers use a mix of fungicides to target both Pythium and Fusarium on chickpea.

Acknowledgements

The authors would like to extend their appreciation to O. Neher, E. Grimme, J. Holmes, M. Burgess, S. King, J. Eckhoff and B. Schaff for their helpful input and technical assistance. This work was supported by Interregional Project – 4 Biopesticide Grant No. 407-1409.