OUP user menu

Fitness of three Fusarium pathogens of wheat

Berna Tunali, Friday Obanor, Gul Erginbaş, Rhyannyn A. Westecott, Julie Nicol, Sukumar Chakraborty
DOI: http://dx.doi.org/10.1111/j.1574-6941.2012.01388.x 596-609 First published online: 1 September 2012


Crown rot and head blight of wheat are caused by the same Fusarium species. To better understand their biology, this study has compared 30 isolates of the three dominant species using 13 pathogenic and saprophytic fitness measures including aggressiveness for the two diseases, saprophytic growth and fecundity and deoxynivalenol (DON) production from saprophytic colonization of grain and straw. Pathogenic fitness was generally linked to DON production in infected tissue. The superior crown rot fitness of Fusarium pseudograminearum was linked to high DON production in the stem base tissue, while Fusarium culmorum and Fusarium graminearum had superior head blight fitness with high DON production in grains. Within each species, some isolates had similar aggressiveness for both diseases but differed in DON production in infected tissue to indicate that more than one mechanism controlled aggressiveness. All three species produced more DON when infecting living host tissue compared with saprophytic colonization of grain or straw, but there were significant links between these saprophytic fitness components and aggressiveness. As necrotrophic pathogens spend a part of their life cycle on dead organic matter, saprophytic fitness is an important component of their overall fitness. Any management strategy must target weaknesses in both pathogenic fitness and saprophytic fitness.

  • pathogen aggressiveness
  • mycotoxin production
  • saprophytic fitness


In population, biology fitness is widely used as a measure of survival and reproductive success of an entity such as an allele, individual or group. Malthusian fitness, which essentially describes instantaneous growth rate, is a convenient way to measure fitness of microorganisms such as bacteria (Pringle & Taylor, 2002) and can be used for fungi where mycelial growth is an effective predictor of spore number. In reality, fitness of filamentous fungi has been measured in a number of different ways reflecting the complexities associated with their life cycle of alternating diploid and haploid generations and the confusion over what constitutes an individual. Also, the concept of fitness has been used in a wide variety of ways, which are often not consistent with its usage in population biology (Antonovics & Alexander, 1986). Traits such as reproductive rate, infection efficiency and aggressiveness have been used to describe the fitness of plant pathogenic fungi (Leach et al., 2001). Often one key spore form such as urediniospore of rusts has been used to determine the fitness, despite the different sexual and asexual spores having specific purpose in the overall life cycle. Consequently, such fitness measures have limited usefulness in understanding evolutionary and other mechanisms.

The mode of nutrition plays a significant role in plant pathogen fitness. For biotrophic pathogens that derive nutrition only from living host tissue and do not have an active existence outside of their hosts, fitness can be determined from fecundity or the heritable latent period, which is a major component of parasitic fitness (Lehman & Shaner, 1996). In contrast, necrotrophic fungal pathogens derive nutrition by killing host tissue during infection and also establish, compete, survive and reproduce on dead organic matter to persist in the absence of living hosts. Their fitness comprises a pathogenic and a saprophytic component. Saprophytic fitness is essential to necrotrophic pathogens, and when host colonization is minimized by a resistant host, this allows the population to grow and reproduce after the host plant dies (Melloy et al., 2010). The same trait/gene can control pathogenic and saprophytic fitness in some fungal pathogens, such as Bipolaris maydis race T (Klittich & Bronson, 1986), but not in others such as Alternaria alternata (Hatta et al., 2002).

The necrotrophic Fusarium graminearum (Fg), Fusarium pseudograminearum (Fp), Fusarium culmorum (Fc) and other species are economically significant pathogens in most cereal-growing regions of the world. They infect the stem base of wheat and barley causing necrosis and dry rot of the crown, basal stem and root tissue commonly known as crown rot (CR, Backhouse et al., 2004). The same species also infect floral tissue causing blighting of the spikelet and developing grains to cause Fusarium head blight (FHB) or scab in both wheat and barley (Boutigny et al., 2011). Wet and warm weather during crop anthesis and maturation may favour FHB (McMullen et al., 1997), and hot and dry conditions promote CR (Burgess et al., 2001). Fg is the most prevalent among FHB pathogens globally (Xu & Nicholson, 2009) including in Australia (Burgess et al., 1987), whereas Fp and Fc are the predominant CR pathogens (Burgess et al., 2001; Backhouse et al., 2004; Chakraborty et al., 2006). FHB has affected vast areas of the globe with recent epidemics in Canada, China, Europe, South America and the USA (Goswami & Kistler, 2004). CR is a chronic disease of economic concern to Australia (Chakraborty et al., 2006), North America (Dyer et al., 2009), Europe, South America, West Asia and North Africa (Burgess et al., 2001). Both diseases curtail grain yield and quality, and during 1998–2000, an estimated US$2.7 billion was lost to FHB in the northern Great Plains and central USA as a result of reduced yield and price discounts from lowered grain quality. One important element of reduced quality is the production of trichothecenes nivalenol (NIV), deoxynivalenol (DON) and other mycotoxins in infected tissue that make them unsafe for human and animal consumption (Desjardins, 2006).

Studies have revealed clear difference in population genetics of the three Fusarium species (Miedaner et al., 2008). The random mating homothallic Fg (teleomorph Gibberella zeae) is driven by occasional outcrossing, high gene flow, low host-mediated directional selection and strong lineages along geographically separated lines. The heterothallic Fp (teleomorph Gibberella coronicola) is a single phylogenetic species with limited lineage development (Scott & Chakraborty, 2006), while the anamorphic Fc shows no obvious clonal structure (Obanor et al., 2010). All three species can cause CR and FHB under natural field conditions (Burgess et al., 1987; Bateman, 2005) and in pathogenicity assays (Akinsanmi et al., 2004).

The three species and their strains differ in overall DON production, and Fg strains generally produce higher levels of DON than the other two (Chakraborty et al., 2006). DON is produced during both FHB (Bai et al., 2002) and CR (Mudge et al., 2006) infections. Studies using DON-nonproducing Fg mutants or strains that differ in DON levels have shown that DON production significantly correlated with FHB severity in some studies (Bai et al., 2002; Goswami & Kistler, 2005) but not in others (Gale, 2003). Similar studies on CR have shown reduced stem colonization by Fg and delayed defence gene expression (Desmond et al., 2008), but there was no overall difference in CR severity when two wild-type strains were compared with two DON-nonproducing strains (Wang et al., 2006).

DON production by Fc and Fg also offers competitive advantage during saprophytic phase of their life cycle, for instance, by suppressing the expression of chitinase genes in Trichoderma atroviride, which is a common antagonist of many fungi (Lutz et al., 2003). Competition among Fusarium species is important during the saprophytic phase but whether DON production offers any advantage is largely unknown. Species such as Fg with a low saprophytic fitness is rapidly replaced by other competitors including Fusarium equiseti and Fusarium oxysporum (Pereyra et al., 2004).

In Australia, FHB is sporadic and limited to south-eastern Queensland and northern New South Wales (Burgess et al., 1987), but CR is widespread throughout the wheat belt (Murray & Brennan, 2009). Despite this, FHB poses a greater risk than CR to food and feed safety from DON-contaminated grains. Both Fg and Fp caused the 2010 FHB epidemics in eastern Australia, but DON levels of immature grains were significantly higher when spikes were infected with Fg than with Fp (Obanor et al., 2012). The linked aetiology, pathogen biology and epidemiology of the two diseases in Australia point to a CR-FHB continuum where macroconidia produced on stubbles from CR infection become FHB inoculum (Mitter et al., 2006a) under suitable weather conditions. The three dominant Fusarium species that drive this continuum have different ecological adaptation, pathogenicity, toxigenicity and possibly saprophytic ability. Improved understanding of biology of these three species can be used to predict their relative prevalence in response to farming and weather conditions to improve the management of FHB epidemics and mycotoxin contamination of grains. To achieve this overall goal, this study compares the fitness of Fg, Fp and Fc and explores the relationships between components of pathogenic and saprophytic fitness.

Materials and methods

Inoculum preparation

Thirty monoconidial isolates comprising 10 Fg, 13 Fp and seven Fc collected from eastern Australia were used in this study (Table 1). Isolates originating from barley or wheat tissues and identified using both molecular and morphological techniques were selected to represent a range of FHB and CR aggressiveness (Akinsanmi et al., 2004). Data on CR aggressiveness were from 56-day-old seedlings. Isolates were recovered from storage in 15% glycerol at −80 °C, and macroconidia suspensions were obtained by cultivating isolates on quarter-strength potato dextrose agar (PDA; Difco Laboratories, MD) plates. The cultures were incubated at room temperature (c. 25 °C) for 10 days under an alternating 12-h darkness/12-h combined black light (F20T9BL-B20W FL20S.SBL-B NIS, Japan) and standard cool white fluorescent light (35098 F18E/33 General Electric). Macroconidia were harvested by flooding individual cultures with 5 mL sterile distilled water and gently scraping colony surfaces with a sterile L-shaped glass rod. The resulting suspension was filtered through two layers of sterile cheese cloth to remove mycelia and concentration adjusted using a haemocytometer (Weber Scientific International, UK) to 105 macroconidia mL−1 for FHB inoculation and 106 macroconidia mL−1 for CR inoculation. Inoculum was stored at −20 °C until required.

View this table:

Source of Fusarium isolates used in this study

IsolateSpeciesHostPlant part
CS3716F. culmorumWheatCrown
CS4494F. culmorumWheatCrown
CS4495F. culmorumWheatCrown
CS4496F. culmorumWheatCrown
CS4497F. culmorumWheatCrown
CS4498F. culmorumWheatCrown
CS4501F. culmorumWheatCrown
CS3005F. graminearumBarleySpike
CS3179F. graminearumWheatSpike
CS3187F. graminearumWheatSpike
CS3192F. graminearumWheatSpike
CS3196F. graminearumWheatStubble
CS3200F. graminearumWheatSpike
CS3259F. graminearumWheatSpike
CS3375F. graminearumWheatSpike
CS3386F. graminearumWheatStubble
CS3407F. graminearumWheatFlag leaf node
CS3096F. pseudograminearumWheatCrown
CS3173F. pseudograminearumWheatCrown
CS3175F. pseudograminearumBarleySpike
CS3181F. pseudograminearumWheatCrown
CS3220F. pseudograminearumWheatSpike
CS3270F. pseudograminearumWheatCrown
CS3321F. pseudograminearumWheatCrown
CS3342F. pseudograminearumWheatCrown
CS3350F. pseudograminearumWheatCrown
CS3361F. pseudograminearumWheatCrown
CS3427F. pseudograminearumWheatCrown
CS3438F. pseudograminearumWheatCrown
CS3442F. pseudograminearumWheatCrown

Aggressiveness and DON production from Crown rot

Wheat plants of the susceptible durum cultivar ‘Tamaroi’ were grown in a naturally illuminated glasshouse with a daytime temperature of 25 °C (±5 °C) with 60% (±10%) relative humidity (RH) and a night-time temperature of 15 °C (±5 °C) with 80% (±10%) RH. For each of the 30 Fusarium isolates (Table 1), six replicate seedlings were individually raised in sterile potting mix in 5 × 5 × 5 cm punnets housed in seedling trays (Garden City Plastics, Qld, Australia). Ten days after emergence, seedlings were laid horizontally on their side, and a 10-μL droplet of inoculum was applied to the base of the stem about 5 mm from the soil surface as described previously (Mitter et al., 2006b). Inoculated seedlings were incubated at high RH in darkness for 48 h after which they were placed upright in seedling trays. Five weeks after inoculation, three replicate plants per isolate were destructively sampled to assess the CR severity of seedlings, while the remaining plants were re-potted in 250-mL square pots (Garden City Plastics) and grown to maturity. The experiment was repeated once.

CR severities on both seedling and adult plants were used as measures of pathogen aggressiveness. The length of necrotic discoloration and seedling height were recorded to express seedling CR severity as the proportion of stem length browning. The proportion of tiller length browning for each yielding tiller was similarly used as a measure of CR severity of adult plants. For each replicate plant, DON was determined from grains and a sample of stem base tissue, covering 10 cm length from the crown.

Aggressiveness and DON production from FHB

Tamaroi plants were grown in the glasshouse as described for CR. Five seeds were sown per pot (13 cm square × 15 cm) in sterile potting mix (50% sand and 50% peat), and three replicate pots were used for each isolate. All spikes at mid-anthesis were inoculated by placing a 10-μL droplet of spore suspension inside the middle spikelet using a pipette (Akinsanmi et al., 2006). Control spikes were inoculated with sterile distilled water. Inoculated spikes were covered with a moistened sealable plastic bag to maintain high RH. After 48 h, the plastic bag was removed and replaced with a paper bag until disease and other assessments. The experiment was repeated once.

FHB severity at 2 weeks after inoculation was used as a measure of pathogen aggressiveness. FHB severity was calculated from the proportion of spikelets infected by recording the number of discoloured and bleached spikelets and the total number of spikelets for each inoculated spike. Spikes were again covered with paper bags and left to mature and senesce. At grain maturity, spikes from all tillers from each plant were harvested and pooled. These pooled samples were used for mycotoxin extraction and analysis.

DON measurements

DON content of grain and stem base was determined using the method of Mirocha et al. (1998) with modifications. DON was extracted in 86% acetonitrile (Sigma-Aldrich, MO) in water from 1 to 3 g of ground grain or stem base tissue using an Accelerated Solvent Extraction (ASE 2000; DIONEX Corporation, CA) equipment. The extraction protocol had three cycles, each consisting of heating for 5 min at 40 °C under 10.3-Mpa, 5-min static phase and a flush phase. Acetonitrile was evaporated under N in a fume hood and DON resuspended in milliQ water before DON assay. A 96-well competitive ELISA kit (Beacon Analytical Systems Inc, Portland, ME) was used following manufacturer's instructions, and DON content was estimated from absorbance at 450 nm using a spectrophotometer (iEMS reader MF; Labsystems, Franklin, MA). When DON content of samples was beyond the ELISA kit detection range of 0.2 and 5 mg kg−1, extracts were concentrated by evaporating or diluted with milliQ water.

Mycelial growth rate and fecundity on PDA

Mycelial growth on 9-cm-diameter Petri plate and fecundity were studied using full-strength PDA at 25 °C. Three replicate cultures for each of the 30 isolates were initiated by inoculating a 4-mm-diameter core taken from the edge of an actively growing culture on quarter-strength PDA. Colony diameter was measured at 2, 4 and 6 days after inoculation. Macroconidia were harvested from 6-day-old cultures to prepare a suspension following methods used to make inoculum. Spore concentration was determined using a haemocytometer, and fecundity was expressed as the number of conidia per unit area of the culture. The experiment was repeated once.

DON production from saprophytic colonization of substrates

The level of DON production after saprophytic colonization of cereals by the 30 Fusarium isolates was tested using sterilized cracked corn, wheat grain and wheat straw as a substrate. All substrates were soaked in water overnight and autoclaved for three successive days. Approximately 350 g of cracked corn or wheat grain, or 55 g of wheat straw dispensed into 15-cm-diameter Petri dishes was used as a replicate. Three replicates per isolate were inoculated with a droplet of 106 macroconidia mL−1 and incubated at 25 °C for 21 days with weekly shaking to break up mycelial matting. The colonized substrates were dried in a laminar flow, ground in a coffee grinder and stored at 4 °C until used. The extraction and determination of DON from ground samples followed the methods described above. The experiment was repeated once.

Rate of saprophytic colonization of wheat straw

The rate of saprophytic colonization of wheat straw by the 30 isolates at 15 and 25 °C was determined using a recently described technique (Melloy et al., 2010). Disease-free wheat straw cut into 8-cm pieces was soaked in water for 24 h and autoclaved for three successive days. Three replicate straws were inoculated for each isolate by pressing one end onto an actively growing culture; each inoculated straw was placed in a Petri dish, sealed with parafilm and incubated at either 15 or 25 °C for 4 days. After incubation, each straw was cut into eight pieces, each one cm long, and plated on fresh PDA plates. Based on the number of pieces yielding Fusarium colonies, saprophytic fitness was expressed as the total length of straw colonized by an isolate. The experiment was repeated once.

Data analysis

The influence of Fusarium species, repeat run of the experiment and isolates on each fitness measure was analysed using analysis of variance. Isolate was nested within species using the general linear models procedure in sas (PROC GLM, SAS Institute Inc., Cary, NC). Data were transformed, as necessary to stabilize variance.

A total of 13 fitness measures were available for analysis: six describing pathogenic fitness were: (i) FHB aggressiveness, (ii) CR aggressiveness on seedlings (ACR-S), (iii) CR aggressiveness on adult plants (ACR-A), (iv) DON in grain from FHB, (v) DON in grain from CR, and (vi)DON in stem base from CR. Seven describing saprophytic fitness were: (i) DON in wheat straw from saprophytic growth, (ii) DON in wheat grain from saprophytic growth, (iii) DON in cracked corn from saprophytic growth, (iv) growth rate on media, (v) fecundity on media, (vi) saprophytic wheat straw colonization at 15 °C and (vii) saprophytic wheat straw colonization at 25 °C. Correlation and principal component analysis (PCA) were used to explore how the 13 measures explained the overall fitness of the 30 Fusarium isolates. Isolates were ranked for each of the 13 fitness measures to generate a data matrix of 13 variates (fitness measures) and 30 observations (isolates), and principal components were calculated using the princomp procedure in sas.

Two-dimensional scatter plots, also known as biplots, were generated by transposing data to produce a matrix of 13 observations (fitness measures) and 30 variates (isolates). The transposed data were monotonically transformed using prinqual procedure with the MDPREF option to maximize the proportion of variance accounted for by the first two principal components. The factor procedure was used on the transformed data to perform a PCA, and the biplot was created using the %PLOTIT macro available as part of the sas autocall library.


Aggressiveness and DON production from Crown rot

Data on ACR-S and DON in grains were arcsine-transformed, DON in stem base was log-transformed to stabilize variance, and CR aggressiveness on mature plants was analysed untransformed. Although all main effect and their interactions were significant (P< 0.05) for seedling severity (output of analysis not shown), the level of aggressiveness measured as the proportion of stem length browning was generally low with relatively small difference between species (Fig. 1). ACR-S for individual isolates ranged from 0.15 to 0.4 for Fc, 0.05 to 0.31 for Fg and 0.02 to 0.48 for Fp. For ACR-A, analysis showed a significant (P< 0.05) effect of species and isolate only. The overall difference between aggressiveness of the three species was more pronounced in adult plants than in seedlings, and Fp was significantly more aggressive than the other two species, while Fc was significantly more aggressive than Fg (output of analysis not shown). The range for adult plant aggressiveness was also wider than for seedlings for all three species: 0.12 to 0.63 for Fc, 0.02 to 0.28 for Fg and 0.07 to 0.93 for Fp. One Fc and six Fp isolates showed high level of aggressiveness causing > 50% tiller length browning of adult plants, but none of the Fg isolates had such high level of aggressiveness (Fig. 2). While 85% of Fp isolates were more aggressive on adult plants than on seedlings, this compared with 40% of Fg and 71% of Fc isolates. These data show that Fp has higher pathogenic fitness for CR than the other two species.

FHB and CR aggressiveness of Fusarium culmorum, Fusarium graminearum and Fusarium pseudograminearum on seedling and adult plants of wheat. The bar represents standard error of the mean. CR aggressiveness was measured as the proportion of stem length browning 5 weeks after inoculation for seedlings and at maturity for adult plants. FHB aggressiveness was measured as the proportion of spikelets infected 2 weeks after inoculation.

FHB and CR aggressiveness of individual isolates of Fusarium culmorum, Fusarium graminearum and Fusarium pseudograminearum on seedling and adult plants of wheat. The bar represents standard error of the mean. See Fig. 1 for description of CR and FHB aggressiveness.

With an overall mean of 7.43 mg kg−1, all three Fusarium species produced high levels of DON in the stem base tissue of CR-infected plants (Fig. 3). Only Fusarium species and isolate had significant effect on stem base DON (output of analysis not shown), and both Fg (8.51 mg kg−1) and Fp (8.49 mg kg−1) produced similar high levels, while Fc (4.13 mg kg−1) produced significantly less. In contrast, very little DON was found in grains of CR-affected plants, and despite a significant isolate effect, DON levels were generally low (overall mean 0.36 mg kg−1), and there was very little difference between the three Fusarium species (output of analysis not shown). Of the 30 isolates, only three Fc (CS4495, CS4496 and CS4501) and one Fp (CS3427) had > 1 mg kg−1 DON in grains from CR infection, but DON levels in stem base tissue were > 100-fold higher for some isolates (Fig. 4). Of the 30 isolates, 6/10 Fg, 8/13 Fp and 1/7 Fc produced > 5 mg kg−1 DON in stem base tissue.

DON content of wheat grain and stem base following FHB or CR disease caused by Fusarium culmorum, Fusarium graminearum or Fusarium pseudograminearum. Bar represents standard error of the mean.

DON content of wheat grain and stem base following FHB or CR disease caused by isolates of Fusarium culmorum, Fusarium graminearum or Fusarium pseudograminearum. Bar represents standard error of the mean.

ACR-A significantly correlated with DON in stem base (DCR-S, r = 0.41, P< 0.01) and ACR-S according to Pearson's correlation coefficient (Table 2). When data for each species were considered separately, the correlation coefficient (r= 0.68, data not shown) for Fp was highly significant (P< 0.001), indicating that DCR-S is an essential component of its CR pathogenic fitness, although the highly aggressive CS3342 was a low DON producer. In contrast, four isolates of Fg (CS3005, CS3196, CS3200 and CS3407) with low ACR-S and ACR-A produced > 10 mg kg−1 DCR-S (Figs 2 and 4), and Pearson's correlation coefficient (r= 0.45) from a species-wise analysis was nonsignificant for this species.

View this table:

Pearson's correlation coefficients (r) between measures of pathogenic fitness head blight aggressiveness (AHB), ACR-S, ACR-A, DON in grain from FHB (DHB-G), DON in grain from CR (DCR-G), DON in stem base from CR (DCR-S); and saprophytic fitness DON in wheat straw from saprophytic growth (DWS), DON in wheat grain from saprophytic growth (DWG), DON in cracked corn from saprophytic growth (DCG), fecundity on media (FEC), saprophytic wheat straw colonization at 15 °C (SG-15) and saprophytic wheat straw colonization at 25 °C (SG-25) for 30 isolates from three Fusarium species. Probability > |r| under H 0: ρ = 0 is indicated with *for P< 0.05 and ** for P< 0.01

Pathogenic fitnessSaprophytic fitness
AHB0.07−0.10.5 **0.13−*−0.10.25−0.0−0.05

Aggressiveness and DON production from FHB

FHB aggressiveness data were square-root-transformed, and DON in grain was log-transformed for an analysis of variance. Fusarium species and isolates differed significantly (P< 0.05) in FHB aggressiveness (output of analysis not shown), and Fc was significantly more aggressive than the other two species that were equally aggressive (Fig. 1). There were aggressive isolates in each Fusarium species. Eleven highly aggressive isolates, three Fg, three Fp and five Fc, infected > 20% spikelets (Fig. 2). Of these, only two Fg (CS3005 and CS3187) and one Fp (CS3220) originated from spike tissue (Table 1), and all remaining isolates were obtained from crown tissue or wheat stubble. This clearly demonstrates the inherent ability of Fusarium species and isolates commonly associated with CR to cause severe FHB under suitable conditions.

Both Fusarium species and isolates significantly influenced DON production in grains from FHB infection (DHB-G) (output of analysis not shown). As a FHB pathogen, Fp produced significantly less DHB-G compared with the other two species, which produced similar DON levels (Fig. 3). However, as a FHB pathogen, Fg is more toxigenic than Fc. Of the seven Fg isolates with > 1 mg kg−1 DHB-G, four produced between 2 and 13 mg kg−1, but the high level of DON production by Fc was because of a single isolate (CS3716) producing > 25 mg kg−1 (Fig. 4). The highest DHB-G by any Fp isolate was 2.4 mg kg−1. When data for all three species were considered, DHB-G significantly (P< 0.01) correlated (r= 0.5) with FHB aggressiveness (AHB, Table 2). From a species-wise analysis, the correlation coefficient (r = 0.71, data not shown) was significant (P< 0.001) for Fg, but nonsignificant for Fc (r= 0.48, P< 0.08) or Fp (r= 27, P= 0.19). Given the similar levels of AHB for Fg and Fp, DHB-G production appears to be only one of many aspects of FHB pathogenic fitness.

Components of saprophytic fitness

Data on mycelial growth rate per day were arcsine-transformed, and fecundity was square-root-transformed to stabilize variance. The influence of Fusarium species, repeat run of the experiment and isolate was analysed using analysis of variance with isolate nested within species. All three species had similar growth rates on PDA at 25 °C (data not shown), but conidia production on PDA was significantly different for the three species and isolates within each species. With 0.9 × 104 macroconidia per unit colony area, Fp had the lowest fecundity and this was significantly less than fecundity of the other two species (Fc 5.3 × 104 and Fg 8.2 × 104). However, there was significant difference in fecundity among isolates within each species, and there was a large variation between replicates and the two times the experiment was run (data not shown).

Data on DON production from saprophytic colonization of wheat straw (DWS), wheat grain (DWG) and cracked corn (DCG) as substrates were log-transformed before an analysis of variance to examine the effect of Fusarium species and isolates on DON levels. There was no significant difference between species, but both isolate and substrate were significant (P< 0.05). Significantly higher amount of DON was produced on wheat straw (0.19 mg kg−1), and this was followed by cracked corn (0.12 mg kg−1) and wheat grains (0.03 mg kg−1), but the species × substrate interaction was also significant (P< 0.05). However, the overall low level of DON production ranging from 0.03 to 0.3 mg kg−1 (data not shown) makes it difficult to pinpoint any obvious substrate-isolate/species combination with the potential to produce high levels of DON, and any trend has to be interpreted with caution.

An analysis of variance of square-root-transformed data on the rate of saprophytic colonization of wheat straw at 15 (SG-15) and 25 °C (SG-25) showed significant (P< 0.05) effect of Fusarium species, isolate, temperature and interactions with temperature (output of analysis not shown). Despite the significant species × temperature interaction, all three species had a higher SG-25 than SG-15 (Fig. 5). At the species level, the rate of straw colonization accelerated at a rapid rate in both Fc and Fp as temperature increased from 15 to 25 °C but the rate of acceleration was lower in Fg (Fig. 5).

The daily rate of colonization of wheat straw incubated at 15 and 25 °C by seven Fusarium culmorum (a), 10 Fusarium graminearum (b) and 14 Fusarium pseudograminearum (c) isolates. The solid line in each panel shows the mean daily rate for each species, and the bar represents the standard error of the mean.

Overall fitness of Fusarium isolates

There were three significant (P< 0.05) correlations between the six measures of pathogenic fitness, FHB aggressiveness (AHB), ACR-S, ACR-A, DON in grain from FHB (DHB-G), DON in grain from CR (DCR-G) and DON in stem base from CR (DCR-S) (Table 2), showing obvious links between DON production and aggressiveness. There was also the same number of significant correlations between the seven measures of saprophytic fitness, DON in wheat straw from saprophytic growth (DWS), DON in wheat grain from saprophytic growth (DWG), DON in cracked corn from saprophytic growth (DCG), growth rate on media (no significant correlation with any variable and not shown in Table 2), fecundity on media (FEC), saprophytic wheat straw colonization at 15 °C (SG-15) and saprophytic wheat straw colonization at 25 °C (CS-25); two of these were linked to SG-25. More importantly for this necrotrophic pathogen, there were four significant correlations between components of pathogenic and saprophytic fitness, including between DWG and DWS with AHB, DHB-G and DCR-G. SG-25 was also significantly correlated with ACR-A.

The first seven principal components explained > 86% of the variance in a PCA, with each additional component explaining < 5%, indicating that variation in fitness of the 30 isolates has seven major dimensions (output of analysis not shown). Eigenvector of the first component explained 18% of the variation with large (> 25%) positive weight on three Fg and one Fp and a large negative weight on one Fp isolate (data not shown). The second eigenvector explained a further 16% of variation with positive weights for one Fc, three Fg and one Fp isolates.

The two-dimensional scatter plot (Fig. 6) revealed a grouping of the DON production–related measures DCG, DCR-G, DHB-G, DWG, together with AHB and saprophytic growth on media along vectors of isolates with high AHB (e.g. CS3187, CS4495, CS4501) or low ACR-A and ACR-S (e.g. CS3220, CS3259, CS3350). But the fitness of 6 (CS3005, CS3200, CS3196, CS3716, CS4496 and CS 3173) of 10 isolates with high AHB could not be explained using the 13 measures used. While DCR-S was aligned with vectors of isolates with high ACR-A and ACR-S (e.g. CS3321, CS3270, CS3342), it did not group with ACR-S or ACR-A, which together with DWS were located along vectors of other isolates with moderate (e.g. CS3181, CS4498) to high (e.g. CS3096, CS3361, CS3442) ACR-A and ACR-S. This suggests that DON production in stem base may be only one of the factors essential for high CR aggressiveness. SG-15, located far left of origin of isolate vectors, did not contribute to the discrimination of isolates. The vector for Fc CS4496 did not align with any of the 13 fitness measures used. Why this low DON-producing isolate originating from wheat crown with moderate CR and high FHB aggressiveness that is generally similar to some other isolates (e.g. CS3173, CS3220) does not group with any other isolate needs further study.

Two-dimensional scatter plot to show the association between 30 isolates of Fusarium culmorum (Fc), Fusarium graminearum (Fg) and Fusarium pseudograminearum (Fp) and 13 fitness measures, head blight (FHB) aggressiveness (AHB), ACR-S, ACR-A, DON in grain from FHB (DHB-G), DON in grain from CR (DCR-G), DON in stem base from CR (DCR-S), DON in wheat straw from saprophytic growth (DWS), DON in wheat grain from saprophytic growth (DWG), DON in cracked corn from saprophytic growth (DCG), growth rate on agar media (GRO), fecundity on media (FEC), saprophytic wheat straw colonization at 15 °C (SG-15) and saprophytic wheat straw colonization at 25 °C (SG-25). The data matrix of fitness measures and isolates was monotonically transformed; principal components were analysed using the factor procedure and plotted using the %PLOTIT macro from the sas autocall library. The direction of isolate vectors indicates the most favoured association with the 13 fitness measures (open circles) based on coefficients for the first two principal components.


This work has for the first time compared important elements of saprophytic and pathogenic fitness controlling FHB and CR aggressiveness of three important Fusarium pathogens of wheat and barley. Results show that these necrotrophic species differ in their overall fitness as pathogens of both diseases. Pathogenic fitness was generally linked to DON production in infected tissue. The superior fitness of Fp as a CR pathogen was linked to high DON production in the stem base tissue, while Fc and Fg had superior FHB fitness with high DON production in grains. While some isolates of all three species have similar CR or FHB aggressiveness, they have different toxigenicity and do not produce the same level of DON in infected tissue. There is a significant correlation between some saprophytic and pathogenic fitness components such as saprophytic DON production in sterile grain and straw with FHB aggressiveness or DON production in infected grain, and between the rates of saprophytic colonization of straw at 25 °C with ACR-A.

Overall, Fp had superior fitness as a CR pathogen with the majority of isolates causing severe CR of adult plants. Our findings on the superiority of Fp over Fc and Fg as a CR pathogen of mature plants are in line with published literature (Backhouse et al., 2004; Chakraborty et al., 2006; Dyer et al., 2009). Although there was a significant correlation between CR severity of seedling and adult plants, all three Fusarium species had largely similar ACR-S and the relatively small difference in aggressiveness between the species magnified on mature plants. Severity of adult plants is a more meaningful measure because of its practical relevance to CR resistance in the field. Most of the highly aggressive Fp isolates produced high levels of DON in the infected stem base tissue of adult plants. Although DON production in straw by Fusarium species has been reported earlier (e.g. Brinkmeyer et al., 2006), this is the first report linking CR aggressiveness of Fp to DON production in straw. Some Fc isolates (CS4495 and CS4497) with high ACR-A did not produce high levels of DON in the stem base. At the species level, ACR-A was significantly higher for Fp isolates than for Fg, although there was no significant difference between these species in DON production in the stem base. This suggests more than one pathogenicity mechanisms operating in the two species.

As a species, Fc was significantly more aggressive than the other two species for FHB, but as with CR, some isolates within each species were equally aggressive. This is somewhat surprising because the environmental conditions used in these studies for plant growth and infections were warmer than in the natural range for Fc (Backhouse et al., 2004). Also, in common with the other two species, Fc was a faster saprophytic colonizer of straw at 25 °C than at 15 °C. Both Fc and Fp accelerated saprophytic colonization of straw faster than Fg when temperature was increased from 15 to 25 °C. However, despite being an important FHB pathogen in Europe (Bateman, 2005) and Australian isolates being capable of causing FHB in pathogenicity assays (Akinsanmi et al., 2004), a Fc-induced FHB epidemic has never been reported in Australia. It remains to be seen if this species will pose a FHB risk for Australia under suitable weather and farming conditions.

Overall, there was a significant correlation between DHB-G and AHB, indicating an important role of DON as a component of pathogenic fitness. However, at the species level, the correlation was significant for Fg and Fc, but not for Fp, supporting previous findings that the role of trichothecenes in FHB pathogenesis differs among species (Langevin et al., 2004). For Fg, DON production enhanced FHB aggressiveness in some studies (Bai et al., 2002) but not in others (Gale, 2003), while DON production by Fc isolates has been implicated in pathogenesis (Muthomi et al., 2000). However, data on DON from small samples of grains need to be treated with caution because of inherent variability among grains from the same plant/spike. DON levels need to be confirmed from large samples of bulked grains.

Our findings are similar to published research (Akinsanmi et al., 2004) which showed that a small group of Fp, Fg and Fc isolates have high aggressiveness for both CR and FHB. In this study, Fc CS4495 and CS4497; Fg CS3375 and CS3407; and Fp CS3173 and CS3361 had moderate to high aggressiveness for both CR and FHB. However, there was no overall link between FHB and CR aggressiveness, and the correlation coefficient was not significant for any species. Recent studies on pathogenesis mechanisms have highlighted differences between the two diseases. Similarities during early stages of pathogenesis for the two diseases were replaced with distinct phases of colonization during later stages of CR infection, and each phase was associated with a different fungal gene expression profile (Stephens et al., 2008).

This study has clearly established Fp as the least toxigenic of the three Fusarium species. As a FHB pathogen, Fp produced just over 1 mg kg−1 DON in grains, while the other two species each produced well over 3 mg kg−1, although for Fc, this was because of a single isolate CS3716 producing over 25 mg kg−1. Difference in DON production in Australian Fg and Fp isolates has been reported before (Blaney & Dodman, 2002). In common with published literature (Mudge et al., 2006), we have detected DON in grains as a result of CR infection by all three Fusarium species, but levels are very low with only one Fp isolate CS3427 producing 1.8 mg kg−1. This is consistent with negligible levels of DON in Australian wheat where CR caused by the widely distributed Fp is a chronic disease in much of this moisture-limiting environment (Backhouse et al., 2004; Chakraborty et al., 2006). Although both Fp and Fg have caused FHB epidemics in Australia (Burgess et al., 1987; Southwell et al., 2003), our findings show that FHB epidemics caused by Fp will lead to lower DON in grains compared with epidemics because of Fg or Fc, with obvious grain quality, food and feed safety implications (Desjardins, 2006). These findings are confirmed by field data from the latest Australian FHB epidemics in 2010, where high level of DON in grains occurred only when Fg was responsible for FHB symptoms (Obanor et al., 2012). Hence, findings from these studies using standardized glasshouse conditions can be used to streamline surveillance and management of DON by targeting areas where FHB epidemics are caused by Fg or Fc.

The level of DON production in infected tissue generally increased by several orders of magnitude when compared with DON production in vitro or from saprophytic colonization of substrates in these studies. This is in common with previous findings (Mudge et al., 2006; Voigt et al., 2007). Nevertheless, some isolates can produce high levels of DON in wheat grains from saprophytic colonization (Chakraborty et al., 2006). Also, DON level in grains from CR infection was much lower than DON level in grains from a direct infection of spike tissue during FHB. Recent in vitro nutrient profiling studies with Fg have shown that treatments with a variety of polyamines increased DON production levels to > 1000 mg kg−1 (Gardiner et al., 2009). Polyamines are well-known plant metabolites produced in response to abiotic stress including water stress, but their role in pathogenicity and DON production is not well understood. Spray inoculation of Fg at mid-anthesis in an FHB infection assay led to concomitant increases in polyamines, Fg biomass and DON in spike tissue (Gardiner et al., 2010). These authors suggested that Fusarium pathogens may have evolved to recognize these metabolites and this may explain severe CR epidemics in drought-stressed environments. However, the overall low DON levels of Australian wheat indicate that factors in addition to polyamines influence DON levels.

There is an ecological adaptation among the three pathogens that further influence their fitness. Macroconidia, not ascospores, are the main inoculum source for FHB in Australia, and Australian Fg isolates produce fewer perithecia in culture or in the field compared with isolates from the USA (Mitter et al., 2006a). With winter snow covering wheat fields postharvest in the USA, the production of perithecia on crop residues is critical to pathogen survival. These perithecia release massive quantities of ascospores in spring and summer to cause FHB (De Luna et al., 2002). As a spike-infecting inoculum for FHB, long-distance aerial dispersal of ascospores has a distinct advantage over short-distance splash-dispersed macroconidia, and this may be a further reason for the limited distribution of FHB in Australia. In this study, isolates with high FHB aggressiveness were obtained from crown, spike and stubble tissue, suggesting that FHB infection may be opportunistic. In contrast, all isolates except Fp CS3175 with high CR aggressiveness originated from the crown tissue, which may indicate an adaptation in the Australian Fusarium pathogens towards CR.

The biology and ecological adaptation of the three Fusarium species in Australia (Chakraborty et al., 2006; Miedaner et al., 2008) also prompted us to consider fitness traits associated with only the vegetative part of their life cycle. While Fc does not have a known sexual stage, in Australia, the sexual stage is infrequent to rare for the other two species. However, a comprehensive understanding of fitness of these important Fusarium species will require studying large number of isolates from around the globe.

This study shows that both pathogenic fitness and saprophytic fitness contribute to the overall fitness of these necrotrophic pathogens, although only a limited number of components were assessed. Saprophytic growth on media and DON production from saprophytic colonization of straw and grain were linked to high FHB aggressiveness. Similarly, the high rate of saprophytic colonization of wheat straw by Fp correlated with its high ACR-A. However, these findings from sterile wheat straw do not represent true competitive saprophytic ability in the field. Field studies in the USA have shown that Fg is a weak saprophyte and is rapidly replaced from colonized straw by other organisms including other Fusarium species (Pereyra et al., 2004), while Fc may persist as chlamydospores in soil. Studies using isolates with different toxigenic potential from each Fusarium species are needed to better assess their competitive saprophytic fitness on crop residues, including any potential role of DON and how this may be linked to their pathogenic fitness. Saprophytic fitness is essential in pathogen life cycle to enable the population to bounce back when partially resistant host varieties restrict their pathogenic growth and reproduction (Melloy et al., 2010). Existing crop residue management tools such as crop rotation is widely used to reduce CR and FHB inoculum. Improved knowledge of overall fitness of necrotrophic pathogens through further studies may lead to novel management strategies.


The senior author was awarded an Endeavour Research Fellowship, and G.E. received a Crawford Fund training award for this research. Funding for this work was provided by the Grains Research and Development Corporation of Australia and CSIRO Plant Industry. Ross Perrott, Alan Tan and Swati Kothari provided technical assistance. All assistance is gratefully acknowledged.


  • Editor: Wietse de Boer


View Abstract