European Journal of Plant Pathology

, Volume 136, Issue 1, pp 159–170

Development of Fusarium langsethiae in commercial cereal production

Article

DOI: 10.1007/s10658-012-0151-x

Cite this article as:
Opoku, N., Back, M. & Edwards, S.G. Eur J Plant Pathol (2013) 136: 159. doi:10.1007/s10658-012-0151-x

Abstract

A field survey was performed to study the infection and development of Fusarium langsethiae in the growing season of wheat, barley, oats and triticale under commercial (2009 – 2011) production. Sampling was completed over three years from fields within the counties of Shropshire and Staffordshire in the UK. Plants sampled (from tillering to harvest) were divided into roots, leaves, lower stem, upper stem and inflorescence/head sub-samples depending on the growth stage of the cereal. DNA was extracted and F. langsethiae DNA quantified using real-time PCR. Fusarium mycotoxins HT-2 and T-2 were quantified from head samples at harvest. Three years of data showed oat to contain the highest levels of both F. langsethiae biomass and HT-2 + T-2 mycotoxins in harvested heads of the cereals studied. The development of F. langsethiae in all three cereals appeared to be similar. Fusarium langsethiae DNA was not detected in the roots and seedlings of all three cereals suggesting F. langsethiae is not a seedling pathogen of cereals. Head infection if it occurs, is at head emergence but before flowering. Seemingly symptomless heads had high levels of F. langsethiae DNA and HT-2 + T-2, confirming previous suggestions that F. langsethiae is a symptomless pathogen of oats.

Keywords

Field survey Head blight Head infection Real-time PCR Symptomless infection 

Introduction

Fusarium langsethiae was first described in 1999 as ‘powdery’ Fusarium poae due its close morphological resemblance to F. poae, but with a powdery appearance on artificial growth media (Torp and Langseth 1999). However, it was not until 2004 that it was classified as a new species and named F. langsethiae (Torp and Nirenberg 2004). Morphologically, both F. langsethiae and F. poae produce microconidia that are globose to napiform in shape; however, F. langsethiae is differentiated from F. poae by its slower growth rate, producing less aerial mycelium. Conidia of F. langsethiae are borne on bent phialides as compared with straight monophialides of F. poae. When cultured on synthetic low-nutrient agar, F. poae produces napiform conidia in combination with falcate sporodochial conidia while F. langsethiae produces only napiform conidia. It also lacks the characteristic peach-like odour of F. poae when grown on artificial medium (Torp and Nirenberg 2004). Geographically, F. langsethiae has been reported mainly in Europe; Austria, Czech Republic, Denmark, England, Germany and Norway (Torp and Adler 2004; Torp and Nirenberg 2004) and more recently in Italy (Infantino et al. 2007), Poland (Lukanowski et al. 2008) and Serbia (Bocarov-Stancic et al. 2008).

Fusarium langsethiae has been identified as the primary producer of HT-2 and T-2 in European cereals (Langseth and Rundberget, 1999; Fredlund et al. 2010; Edwards et al. 2012). HT-2 and T-2 are considered as two of the most potent type A trichothecene mycotoxins and are a public health concern in Europe. The European Commission is currently discussing setting investigative limits for HT-2 + T-2 concentration in cereals for both human and animal consumption. Current discussion limits for cereals intended for human consumption are 100 μg kg−1 for barley, 1000 μg kg−1 for oat and 50 μg kg−1 for wheat. If legislation is set for HT-2 + T-2, farmers will have to control F. langsethiae in their cereals if they are to be marketed for human consumption. However, the F. langsethiae-cereal interaction is poorly understood making the development of control strategies difficult. Attempts have been made to understand its pathogenicity and aggressiveness towards wheat, barley and oats using in vitro assays (Imathiu et al. 2009; Opoku et al. 2011) but studies are still at the early stages and available information is not adequate to fully understand this interaction.

To understand the interaction between F. langsethiae and the different cereal species, assessment of the fungus at different physiological growth stages of the cereal is very important and according to Cooke (2006), data from field samples provide reliable information on the incidence, severity and spatial pattern of a disease under study. Data from field samples on the development of F. langsethiae in cereals is lacking. Researchers in the past few years have concentrated their efforts on the identification and quantification of F. langsethiae DNA and T-2 + HT-2 mycotoxin levels in harvested cereal grains from the field (Edwards 2009a, b, c; Lukanowski et al. 2008; Infantino et al. 2007; Torp and Nirenberg 2004; Torp and Langseth 1999).

Apart from common problems associated with field studies such as the difficulty in explaining data as a result of a number of variables which the researcher has no or little control over, there are no known symptoms of F. langsethiae infection in the field, compounding the problems of field study of this fungus.

The aim of this study was to assess the infection and development of F. langsethiae in the growing season of cereals under commercial cultivation.

Materials and methods

Study fields and general sampling

To assess the infection and development of F. langsethiae in cereals in the UK, a survey was carried out in commercial fields of wheat, barley, oats and triticale in Shropshire and Staffordshire during the 2009, 2010 and 2011 cropping seasons. All fields were within 30 km of Harper Adams University, Newport, Shropshire UK. In all cropping seasons sampling was done between April and August. Survey was carried out in 25, 27 and 26 different fields in 2009, 2010 and 2011 cropping seasons respectively (Table 1). Samples were taken between Zadoks growth stages (GS) 22 to GS92 (Zadoks et al. 1974).
Table 1

Number of commercial cereal fields sampled within Shropshire and Staffordshire in 2009, 2010 and 2011 cropping seasons

Cereal

Year

Winter oat

Winter wheat

Winter barley

Spring barley

Triticale

2009

10

5

5

5

0

2010

7

6

5

5

4

2011

7

5

5

5

4

In each field, a plot of 10 × 10 m was marked out from which samples were taken. Twenty whole plants were taken randomly from each plot. The average growth stage of the plants was determined. At harvest (GS92) weeds from three oat fields, predominantly Italian rye grass, were also sampled. Samples were visually assessed for any disease symptoms before processing.

Processing of samples

Plant stem bases and roots were washed free of soil and divided into sub-samples depending on the growth stage of the plant. At GS23 to GS29 about 4 cm of the lower stem was taken as a sub-sample. Between GS30 and GS39, two sub-samples were taken, lower stem base and leaves. Leaves, lower stem base and upper stem were taken for plants at GS40 to GS49. For plants at GS59–GS92 lower stem base, leaves and inflorescence or heads were taken as sub-samples. All cereal heads were examined for Fusarium head blight symptoms before processing. Sub-samples were put in labelled paper bags, frozen (−18 °C) for at least 24 h after which they were freeze-dried for five days in a Modulyo freeze drier (Edwards, Sussex, UK). Freeze-dried samples were milled in a laboratory mill with a 1 mm screen (Cyclotec 1093 or IKA MF10).

DNA extraction

Milled samples were mixed thoroughly and 2.5 g (leaf and stem samples) or 5 g (head samples) weighed into 50 ml centrifuge tubes for DNA extraction. To each 50 ml centrifuge tube, 30 ml of CTAB buffer (87.7 g NaCl, 23 g sorbitol, 10 g N-lauryl sarcosine, 8 g hexadecyl trimethylammonium bromide, 7.5 g ethylenediamine tetraacetic acid and 10 g polyvinylpolypyrolidone, made up to 1 L with distilled water) were added. Tube contents were mixed thoroughly by hand and with an Hs501 digital shaker (IKA Labortechnik, Staufen, Germany) for 20 min and then incubated at 65 °C for 1 h. Tubes were then shaken by hand and centrifuged at 3,000 × g for 15 min after which 0.9 ml of the supernatant was removed and added to 0.3 ml potassium acetate (5 M) in a sterile 1.9 ml Eppendorf tube, mixed for 1 min and frozen at −20 °C for 1 h. Tube contents were thawed at room temperature before 0.6 ml chloroform was added. The contents of the tubes were then mixed for 1 min and centrifuged at 12,000 × g for 15 min. One milliliter of the aqueous phase was removed and added to a sterile 1.9 ml Eppendorf tube containing 0.8 ml of 100 % isopropanol and mixed for 1 min before centrifuging at 12,000 × g for 15 min. Resulting DNA pellets were washed twice with 1 ml 44 % isopropanol. Pellets were air dried before re-suspending in 0.2 ml TE buffer and incubating at 65 °C for 25 min. Tube contents were vortexed and left at room temperature overnight before spinning at 12,000 × g for 5 min. DNA concentrations were determined using a Nanodrop 2000 spectrophotometer (Thermo Scientific, Basingstoke, UK). DNA extractions were diluted to 40 ng μl−1 except for DNA from stems and leaves of plants at GS90 and GS92 which were diluted to 20 ng μl−1 due to the low levels of DNA recovered. After dilution, DNA concentrations were measured again to determine the final working concentration and stored at 4 °C.

Internal transcribed spacer (ITS) amplification

An initial control PCR was carried out on all DNA samples prior to real-time PCR to ensure the presence and quality of DNA in samples. This involved amplification with ITS4 and ITS5 primers (TCC TCC GCT TAT TGA TAT GC and GGA AGT AAA AGT CGT AAC AAG G respectively) (White et al. 1990). These primers amplify both fungal and plant DNA present in a sample at an anneal temperature of 50 °C. PCR was carried out using a 25 μl reaction mixture made up of 100 μM of each nucleotide, 100 nM of each primer, 20 U of Taq polymerase (ABgene, Epsom, UK) ml−1, 10 mM Tris–HCl (pH 8.3), 1.5 mM MgCl2, 50 mM KCl, 100 μg of gelatine ml−1, 0.5 mg ml−1 of Tween 20, 0.5 mg ml−1 of Nonidet P-40 and 5 μl of DNA sample. Water (PCR grade) was used as negative control. Samples were amplified using a PTC-100 thermal cycler (MJ Research Inc., Minnesota, USA) programmed for initial denaturation at 94 °C for 75 s followed by 35 cycles of 15 s at 94 °C, 15 s at 50 °C and 45 s at 72 °C. There was a final extension step at 72 °C for 4 min 25 s before cooling to 4 °C until sample recovery. Amplicon gel electrophoresis was carried out on 2 % agarose gels stained with ethidium bromide (0.05 %). PCR products (ca. 650 bp) were viewed on a Gel Doc 1000 system (Bio-Rad, Buckinghamshire, UK) under UV light.

Real-time PCR

DNA samples were amplified with a real-time PCR instrument (iCycler Bio-Rad, UK) with F. langsethiae primers; FlangF3 5′-CAAAGTTCAGGGCGAAAACT-3′.and LanspoR1 5′-TACAAGAAGACGTGGCGATAT-3′ (Wilson et al. 2004) as detailed previously (Edwards et al. 2012). qPCR MasterMix Plus for SYBR® Green I with flourescein (Eurogentec, USA) reagent was used according to manufacturer’s instructions with a 25 μl reaction volume which included 5 μl template DNA. PCR water (5 μl) was used instead of template as negative control. A 10-fold dilution series of F. langsethiae DNA (100– 10−4 ng μl−1) was included in each PCR run to provide a standard curve. Concentrations of F. langsethiae DNA were divided by the total DNA concentration within a sample to give values per ng of total DNA.

The real-time PCR conditions an initial activation step (UNG) of 50 °C for 2:30 min and an initial melt of 10 min at 95 °C followed by 45 cycles with a melting step of 95 °C for 10 s, annealing temperature of 65 °C for 10 s, extension at 72 °C for 30 s, and a hold at 82 °C for 10 s during which fluorescence was measured. Melting curve florescence was determined by holding at 95 °C for 1 min, cooling to 55 °C for 1 min and then raising the temperature to 95 °C at a ramp rate of 0.05 °C s−1.

HT-2 + T-2 estimation

T-2 in milled cereal head sub-samples was measured using Ridascreen® T-2 ELISA assay (R-Biopharm AG,Darmstadt, Germany) following manufacturer’s instructions. Total HT-2 + T-2 was estimated based on the known ratio of HT-2 and T-2 in UK oats and the cross-reactivity of the T-2 antibody with HT-2 (Edwards et al. 2012).

Data analysis

Mean DNA values were calculated using Microsoft Excel (V. 2010, Microsoft). In sub-samples with detectable but not quantifiable (i.e. below the LoQ) F. langsethiae DNA were given a value of half the LoQ (0.0005 pg ng−1) to allow for the estimation of mean values. A linear regression of F. langsethiae DNA against HT-2 + T-2 concentration grouped by cereal and year of sampling was carried out to determine if there was a relationship between the parameters under study using Genstat (V.13 VSN International Ltd.).

Weather data was obtained from the Harper Adams University weather station (http://weather.harper-adams.ac.uk/).

Results

Head blight symptoms were not observed on any of the cereal head sub-samples.

The initial control PCR (ITS4 and ITS5) showed that DNA from all stem, leaf and head samples was amplifiable. However, DNA could only be amplified in 60 % and 65 % of the root samples from 2010 and 2011 respectively. There was no detectable F. langsethiae DNA in any of the root samples with amplifiable DNA. There was also no F. langsethiae DNA in any of the weeds sampled at harvest. Fusarium langsethiae DNA could be detected in a small number of sub-samples but could not be quantified in any sub-samples before GS30.

There was a yearly variation in F. langsethiae DNA levels in head sub-samples at harvest. Levels in oat decreased from 2009 to 2011 whilst levels in the other cereals did not follow a particular pattern over time Fig. 1
Fig. 1

Mean F. langsethiae DNA in cereal head sub-samples of oat, winter wheat, spring barley, winter barley and triticale at harvest (GS92) in 2009, 2010 and 2011 cropping seasons. Bars represent standard error

Development of F. langsethiae in all cereals studied was similar (based on pattern of F. langsethiae DNA in different plant parts over three years) but most differences were most distinct for oat due to the higher levels of F. langsethiae DNA observed in this cereal (Fig. 2).
Fig. 2

Mean F. langsethiae DNA in oat plants under commercial cultivation in 2009, 2010 and 2011 cropping seasons (n = 24). Bars represent standard error

In each year of sampling, the level of HT-2 + T-2 in the heads of wheat, barley and oat followed the same trend as F. langsethiae DNA that was amplified in the heads. Analysis of variance (at 95 % confidence level and using Tukey test to separate the means) revealed significant differences in HT-2 + T-2 concentration for each year of sampling with oats (P = 0.005), winter wheat (P = 0.033) and spring barley (P = 0.007). In winter barley, however, the difference in HT-2 + T-2 concentrations over the three years was not statistically significant (P = 0.11).

Weather data showed yearly differences in maximum and minimum temperatures as well as relative humidity values, especially during sampling weeks of the two important growth stages (GS59 and GS92, based on F. langsethiae DNA concentrations in head sub-samples). Due to the relevance of these two growth stages (GS59 and GS92), weekly mean, maximum and minimum temperature and relative humidity values were calculated by taking the day of sampling as the last day of the week (Table 2).
Table 2

Average daily maximum and minimum temperatures and mean relative humidity for one week before GS59 and GS92

Week of sampling

Max. temperature (°C)

Min. temperature (°C)

Relative humidity (%)

 

2009

2010

2011

2009

2010

2011

2009

2010

2011

Oat

GS59

24.2

23.9

17.7

13.9

9.0

7.9

85.6

69.9

68.6

GS92

19.1

20.9

20.5

10.5

12.8

9.4

85.2

87.1

83.3

Winter barley

GS59

22.6

21.9

18.4

7.9

8.3

10.2

64.7

75.0

82.4

GS92

21.9

21.0

20.6

10.6

13.0

10.5

79.7

74.0

83.8

Oat

In all three years of sampling F. langsethiae DNA in sub-samples between GS22 and GS39 were below the LoQ except in upper stem sub-samples in 2009 and leaf sub-samples in 2011 (Table 3).
Table 3

Mean F. langsethiae DNA (pg ng−1) in oat sub-samples in 2009, 2010 and 2011

 

2009

2010

2011

Growth stage

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

22

<LoQ

N/A

N/A

_

_

_

_

_

_

32

<LoQ

<LoQ

N/A

<LoQ

<LoQ

<LoQ

0.0024

<LoQ

N/A

39

<LoQ

<LoQ

0.0230

_

_

_

_

_

_

49/51

_

_

_

0.0630

<LoQ

<LoQ

<LoQ

0.0023

0.0006

59

0.0410

0.0240

0.2830

0.0280

0.0010

0.0860

0.0235

0.0022

0.0031

71

_

_

_

_

_

_

<LoQ

0.0019

0.0180

80/85

<LoQ

0.0150

0.4300

0.0340

<LoQ

0.0760

<LoQ

<LoQ

0.0546

92

0.0700

<LoQ

0.8050

0.0586

0.0038

0.7120

0.0084

0.0014

0.4152

N/A = no sub-sample taken, - = no sampling

At GS59, relative to the previous growth stage sampled, there was an increase in F. langsethiae DNA in all sub-samples in all sampling years except in 2010 (Table 3). Mean F. langsethiae DNA in head sub-samples at GS59 was 0.283 pg ng−1 in 2009, 0.086 pg ng−1 in 2010 and 0.0031 in 2011 (Table 3). After GS59, an increase in F. langsethiae DNA in head sub-samples was observed with highest levels at GS92 with mean values of 0.805 pg ng−1 in 2009, 0.712 pg ng−1 in 2010 and 0.415 pg ng−1 in 2011. In leaf and stem sub-samples however, although relatively high F. langsethiae DNA was recorded at GS92 increases over the growing period were not linear. For example, in 2009 there was a decrease in F. langsethiae DNA in stem sub-samples from GS59 through to GS92 (Table 3). F. langsethiae DNA in head sub-samples at harvest decreased gradually over the sampling years with the lowest levels in 2011.

HT-2 + T-2 values ranged between 1232–10777, 75–2905 and 425–2598 μg kg−1 with means of 4300, 1505 and 1229 μg kg−1 in 2009, 2010 and 2011 respectively. Simple linear regression analysis between log-transformed HT-2 + T-2 and F. langsethiae DNA concentration revealed a highly significant regression (P < 0.001, r2 = 0.52) (Fig. 3a). Regression with log-transformed HT-2 + T-2 and F. langsethiae DNA grouped by year showed that year of sampling had a significant effect (P = 0.003) on HT-2 + T-2 levels in milled oat heads. Year of sampling however, contributed only 2.1 % of the variance accounted for. The interaction between year of sampling and F. langsethiae DNA was not significant (P = 0.203).
Fig. 3

Relationship between F. langsethiae DNA and HT-2 + T-2 in cereal heads of a oat, b winter wheat, c winter barley, d spring barley and e triticale from commercial fields in Shropshire and Staffordshire from 2009 to 2011

Winter wheat

With the exception of 2009, sampling for winter wheat started at GS33. In 2009, F. langsethiae DNA recovered from all sub-samples between GS22 and GS55 were below the LoQ. In 2010 and 2011, however, a number of fields had levels above the LoQ at these growth stages (Table 4). In 2009 quantifiable levels of F. langsethiae DNA was detected in sub-samples at GS59. At GS59, mean F. langsethiae DNA was 0.065 and 0.003 pg ng−1 in leaf sub-samples in 2009 and 2011 respectively; and was 0.01 pg ng−1 in stem sub-samples and 0.024 pg ng−1 in head sub-samples in both 2009 and 2011.
Table 4

Mean F. langsethiae DNA (pg ng−1) in winter wheat sub-samples in 2009, 2010 and 2011

 

2009

2010

2011

Growth stage

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

22

 

<LoQ

N/A

_

_

_

_

_

_

33/35

<LoQ

<LoQ

N/A

0.0200

0.0180

N/A

<LoQ

<LoQ

N/A

51/55

<LoQ

<LoQ

<LoQ

0.0530

0.0500

<LoQ

0.0032

<LoQ

<LoQ

59/60

0.0650

0.0110

0.0240

_

_

_

0.0030

0.0119

0.0239

69

_

_

_

0.2290

0.0030

0.0210

0.0190

<LoQ

0.1278

80/85

0.0165

0.0140

0.1030

0.0210

<LoQ

0.1260

0.0006

0.0059

0.0129

92

0.0290

0.0370

0.1040

0.1600

0.0200

0.2400

0.0168

0.0037

0.0835

N/A = no sub-sample taken, - = no sampling

F. langsethiae DNA increased gradually from GS59 through to GS92 in stem and head sub-samples in 2009 and in head sub-samples in 2010 (Table 4). This was not the case in the other sub-samples during the sampling years. However, there was a consistent increase in F. langsethiae DNA between GS80/85 and GS92 in all sub-samples in all sampling years except in stem sub-samples in 2011(Table 4).

The highest mean F. langsethiae DNA in head sub-samples at harvest was recorded in 2010 (0.24 pg ng−1), with samples collected in 2011 having the lowest F. langsethiae DNA concentration with a mean value of 0.08 pg ng−1.

HT-2 and T-2 levels followed a similar trend as that observed in F. langsethiae DNA over the sampling years. Mean HT-2 + T-2 values were 363, 556 and 96 μg kg−1 and ranged between 111–600, 142–920 and 51–142 μg kg−1 in 2009, 2010 and 2011 respectively. A positive correlation was observed between log-transformed F. langsethiae DNA levels and HT-2 + T-2 levels in harvested head with a significant regression (P = 0.017, r2 = 0.42) (Fig. 3b).

A regression with log-transformed HT-2 + T-2 and F. langsethiae DNA grouped by year showed a significant effect of sampling year on HT-2 + T-2 levels in winter wheat heads at harvest (P < 0.001) accounting for 15 % of the observed variance. The interaction between year of sampling and F. langsethiae DNA was not significant (P = 0.553).

Winter barley

Sampling of winter barley fields started at different growth stages over the three year period. The first sampling was done at GS32 in 2009, at GS49/51 in 2010 and GS59 in 2011 (Table 5).
Table 5

Mean F. langsethiae DNA (pg ng−1) in winter barley sub-samples in 2009, 2010 and 2011

 

2009

2010

2011

Growth stage

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

32

<LoQ

<LoQ

N/A

_

_

_

_

_

_

49/51

0.0148

<LoQ

<LoQ

0.0270

0.0080

0.0010

_

_

_

59

0.0698

<LoQ

0.0070

0.0805

<LoQ

0.1395

0.0295

<LoQ

0.0514

69

_

_

_

0.0078

0.0026

0.0088

<LoQ

0.0035

0.0143

85

<LoQ

<LoQ

0.0042

0.1070

0.0430

0.0140

0.0028

0.0164

0.0128

92

0.2200

<LoQ

0.0650

0.0320

0.0001

0.0120

0.0051

0.0340

0.1185

N/A = no sub-sample taken, - = no sampling

Although there was a general increase in F. langsethiae DNA over the growing season in the three years of sampling (except in stem sub-samples in 2009) (Table 5), this increase was not linear in any sub-sample. However, there was a consistent increase in F. langsethiae in leaf and head sub-samples between GS49/51 and 59 and GS85 and GS92 except in 2010 (Table 5). F. langsethiae DNA in winter barley heads at harvest was highest in 2011 with a mean value of 0.119 pg ng−1 with samples collected in 2010 having the lowest concentration with a mean value of 0.012 pg ng−1 (Table 5.).

The highest HT-2 + T-2 values were recorded in 2010 with the lowest recorded in 2009. Mean HT-2 + T-2 values were 278, 559 and 403 μg kg−1 with values ranging between 124–639, 152–735 and 246–462 μg kg−1 in 2009, 2010 and 2011 respectively. Although a positive correlation was seen between log-transformed F. langsethiae DNA and HT-2 + T-2 levels in heads, the regression was not significant (P = 0.075, r2 = 0.22) (Fig. 3c).

A regression with log-transformed HT-2 + T-2 and F. langsethiae DNA grouped by year showed a significant effect of sampling year with no interaction between sampling year and F. langsethiae DNA on HT-2 + T-2 levels in winter wheat heads at harvest (P = 0.047 and 0.489 respectively). Year of sampling accounted for 17.4 % of the observed variance.

Spring barley

In spring barley, samples taken before GS59 had F. langsethiae DNA below the LoQ except in leaf sup-samples taken in 2010 (Table 6).
Table 6

Mean F. langsethiae DNA (pg ng−1) in spring barley sub-samples in 2009, 2010 and 2011

 

2009

2010

2011

Growth stage

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

32

<LoQ

<LoQ

N/A

_

_

N/A

<LoQ

<LoQ

N/A

51

_

_

_

0.0150

<LoQ

<LoQ

<LoQ

<LoQ

<LoQ

59

<LoQ

<LoQ

0.0100

_

_

_

0.0347

0.0072

0.0114

69

_

_

_

0.1970

0.0200

0.0060

<LoQ

0.0053

0.1073

85

0.0230

0.0340

0.0090

0.2410

<LoQ

0.0070

0.0189

0.0045

0.1677

92

0.2730

0.0010

0.0110

0.5190

0.1180

0.1470

0.0328

0.0276

0.1330

N/A = no sub-sample taken, - = no sampling

From GS59, when quantifiable F. langsethiae DNA was observed, there was an increase in all sub-samples over the sampling period (although this was not linear) with the highest at GS92 except in 2009 (Table 6).

F. langsethiae DNA in head sub-samples at harvest was highest in samples collected in 2010 with a mean value of 0.147 pg ng−1 with samples collected in 2009 having the lowest levels with a mean value of 0.011 pg ng−1.

HT-2 + T-2 values over the sampling years followed a similar pattern to F. langsethiae DNA values at harvest. Values ranged between 75–568, 531–625 and 75–568 μg kg−1 with mean values of 186, 573 and 208 μg kg−1 in 2009, 2010 and 2011 respectively. A good correlation was obtained between log-transformed F. langsethiae DNA in head sub-samples at maturity and HT-2 + T-2 levels (Fig. 3d), with a statistically significant regression (P = 0.004, r2 = 0.48).

A regression with log-transformed HT-2 + T-2 and F. langsethiae DNA grouped by year showed a significant effect of sampling year on HT-2 + T-2 levels in spring barley heads at harvest (P = 0.008) accounting for 7 % of the total variance observed. The interaction between year of sampling and F. langsethiae DNA was not significant (P = 0.125).

Triticale

Triticale fields were sampled in 2010 and 2011 only. In 2010, F. langsethiae DNA in samples collected before GS59 had values below the LoQ except in stem sub-samples whilst all samples collected in 2010 had values below the LoQ even at GS59 (Table 7).
Table 7

Mean F. langsethiae DNA (pg ng−1) in triticale sub-samples in 2010 and 2011

Triticale

2010

2011

Growth stage

Leaves

Stem

Upper stem/head

Leaves

Stem

Upper stem/head

49

_

_

_

<LoQ

<LoQ

<LoQ

55

<LoQ

0.0030

<LoQ

_

_

_

59

0.0270

<LoQ

0.0010

<LoQ

<LoQ

<LoQ

69

0.3270

0.0020

0.0020

<LoQ

0.0018

<LoQ

71

_

_

_

0.1658

<LoQ

0.0055

85

0.0040

<LoQ

0.0100

<LoQ

0.0018

0.0078

92

0.4470

0.0120

0.3200

0.0268

0.0198

0.0552

- = no sampling

From the growth stage when quantifiable F. langsethiae DNA was observed in sub-samples, it increased in all sub-samples over time with highest values being recorded at GS92 in all sampling years. These increases were however, not linear in leaf and stem but they were in head sub-samples (Table 7). At harvest F. langsethiae DNA in head sub-samples was higher in 2010 with a mean value of 0.32 pg ng−1 and 0.055 pg ng−1 in 2011.

Recorded HT-2 + T-2 values were highest in 2010 with values ranging from 320–599 μg kg−1 and a mean of 510 μg kg−1 HT-2 + T-2 values for 2011 ranged between 97 and 212 μg kg−1 , with a mean of 124 μg kg−1. The regression between F. langsethiae DNA and HT-2 + T-2 was not significant (P = 0.13, r2 = 0.34) (Fig. 3e).

A regression with log-transformed HT-2 + T-2 and F. langsethiae DNA grouped by year showed a significant effect of sampling year on HT-2 + T-2 levels in triticale heads at harvest (P = 0.002) accounting for only 2.4 % of the total variance. The interaction between year of sampling and F. langsethiae DNA was not significant (P = 0.097).

Discussion

Results from this study have highlighted three important issues; 1. Fusarium langsethiae biomass and HT-2 + T-2 in harvested heads of the cereals studied differed, 2. There was a yearly difference in F. langsethiae biomass as well as HT-2 + T-2 concentrations and 3. The development of F. langsethiae in the cereals studied over the sampling periods followed a similar pattern.

The absence of F. langsethiae DNA in sampled weeds suggests that the weeds may not be alternative host to F. langsethiae. However, since the sampling size was small, there is the need for further sampling of a broader range of weed species to confirm this.

Fusarium langsethiae biomass in the cereals was estimated using the amount of fungal DNA quantified per unit of plant DNA. The amount of F. langsethiae DNA in the heads of cereals at harvest was found to be highest in oat followed by that in wheat, barley and triticale. These levels correlated well with HT-2 + T-2 levels. This indicates that there may be a stronger association between F. langsethiae and oats than other cereal species. Associations between Fusarium species and different plant species is not a new phenomenon and this trend may even vary from region to region. Fusarium verticillioides (syn. F. moniliforme) and F. subglutinans tend to be more associated with infected maize heads than wheat, barley or oat heads (Doohan et al. 2003). Again it is on record that in the USA F. graminearum is the predominant causal agent of FHB in wheat (McMullen et al. 2012). However, in Europe it is caused by a number of Fusarium species under different conditions (Xu et al. 2005). Dry and warm conditions favour infection by F. poae, whereas warm and humid conditions tend to favour F. graminearum infection and F. culmorum infection is favoured by cool, wet or humid conditions. The two Microdochium species that are involved in head blight infection occur in cool to moderate temperatures with frequent rain showers (Xu et al. 2008).

Field data regarding the occurrence of F. langsethiae is lacking, but there is some data on levels of HT-2 + T-2 in cereals. Occurrence data from three European countries (Sweden, Finland and Norway) in 2007 show that among wheat, barley and oat, lower levels of HT-2 and T-2 occurred in wheat, moderate levels in barley and high levels in oat (Edwards et al. 2009). In contrast, a study carried out in the UK between 2001 and 2005 showed that the concentration of HT-2 + T-2 was highest in oats followed by wheat and then barley (Edwards 2009a,b,c). This is in agreement with the findings from the present study.

The reason for the stronger association between F. langsethiae and oats is not yet understood, but some speculations can be made. The occurrence of FHB is not as obvious in oat as it is in wheat and barley (Browne and Cooke 2005). Some authors have attributed this to the architecture of the oat panicle. The oat panicle is composed of open rachis from which loose spikelets arise. This structure results in single spikelet infection and thus a slower rate of spread of Fusarium infection compared to the denser head architecture found in wheat and barley (Kosova et al., 2009). F. langsethiae is thought to be a poor competitor compared to the other Fusarium species responsible for FHB (Yli-Mattila et al. 2009) and it has been suggested that it takes advantage of the absence of these species. This means that in the event of competition for infection in these three cereals, single spikelet infection coupled with a slow rate of infection in oat provides F. langsethiae a greater scope to infect and colonize. Consequently, this may be one of the reasons for the observed trend.

The effect of the previous crop in a rotation is known to impact on the occurrence of FHB pathogens and mycotoxin production. Deoxynivalenol production and subsequent loss of grain quality is high when wheat, barley and oat follow maize in a rotation (Dill-Macky and Jones 2000). Edwards (2007) also showed that previous crop was an important factor in levels of HT-2 + T-2 in UK oats and indicated that the lowest levels were detected when oats followed a non-cereal in a rotation. In the case of this study, however, rotations in all fields studied were similar and so rotation is unlikely to be the reason for the high level of F. langsethiae biomass and corresponding high HT-2 + T-2 levels in oats.

Observed yearly differences in F. langsethiae DNA were not the same in all cereal species studied. Whilst there was a clear pattern of F. langsethiae DNA levels in oat head sub-samples at harvest, levels in wheat, barley and triticale did not show a similar pattern.

In oat, F. langsethiae biomass found in the heads at harvest was highest in 2009 and gradually decreased over the subsequent sampling years. The regression analysis showed significant effect of year of sampling as well as F. langsethiae DNA concentration on HT-2 + T-2 concentration in cereal heads at harvest. Xu et al. (2008), quoted De Wolf et al. (2000) that weather conditions, often 7–14 days before flowering are known to generally influence sporulation of FHB pathogens and subsequent disease occurrence. To understand the change in F. langsethiae biomass over the sampling years in this study, consideration was given to the temperature and relative humidity conditions during the week of sampling at GS59 and GS92 when a large increase in F. langsethiae DNA occurred. Climatic factors coupled with host factors have a profound influence on the growth and survival of Fusarium species (Doohan et al. 2003). Moreover, temperature and moisture are the most important climatic conditions influencing the infection and distribution of FHB pathogens. Fusarium head blight infection in cereals tends to be severe when anthesis coincides with warm and humid conditions (Edwards 2004; Xu 2003). It is also known that inoculum production and dispersal are also influenced by climatic conditions such as temperature, humidity and wind (Parry et al. 1995). Fusarium langsethiae infection takes place at head emergence but before flowering (around GS59). At this growth stage mean weekly temperature and humidity were found to be highest in 2009 followed by 2010 with the lowest being recorded in 2011. It is of interest to note that the range between the maximum and minimum temperatures in 2009 was relatively narrow compared to the other two years at GS59 (Table 1). Fusarium langsethiae biomass at GS59 followed the same pattern. This suggests that warm temperatures coupled with a relatively high relative humidity favour F. langsethiae infection in oats at GS59. Other field data to support this trend is lacking. However, Medina and Megan (2010) showed that the growth rate of F. langsethiae isolates from England, Finland, Norway and Sweden were faster at 20–25 °C and at 0.98–0.995 water activity (aw) in in vitro studies. These authors also indicated that at low aw (0.90), F. langsethiae did not grow irrespective of temperature, highlighting the importance of moisture for the growth of F. langsethiae. This is in agreement with data from this work where high F. langsethiae biomass in oat heads corresponded well with high temperatures (ca. 24 °C) and relative humidity (70–85 %) during the infection period.

The temperature and relative humidity data at harvest does not appear to explain the observed F. langsethiae biomass and corresponding HT-2 + T-2 levels in oat heads at harvest. There is therefore the need for data from longer period of sampling to determine the association between pathogen development and weather parameters.

Triticale was sampled in 2010 and 2011, and the pattern of F. langsethiae biomass in heads at harvest followed a similar pattern to that observed with oats, therefore it is reasonable to say that factors that accounted for this observation are likely to be the same as that already discussed. However, in wheat and barley, the pattern observed did not match that observed in oat, especially in 2009. This was not as expected as, apart from winter barley which reached GS59 and GS92 earlier in all sampling years, all the other cereals studied reached these two growth stages at almost a similar date in all sampling years and thus it could have been expected that the pattern of F. langsethiae in the heads to be similar even if concentrations present were different.

At the time of sampling when winter barley was at GS59 in all three years, the average temperature was highest (although not much different) in 2009 followed by that in 2010 and 2011, just like it was when oat plants were being sampled. Relative humidity at this growth stage on the other hand increased over the years unlike the case of oat where relative humidity decreased over the years. This can, however, not be used to fully explain the pattern of F. langsethiae biomass observed in winter barley heads over the sampling years. This is because if relative humidity was the deciding factor, then one would expect a gradual increase in F. langsethiae biomass in heads of winter barley from 2009 to 2011, but this was not the case. The temperature data at harvest again does not explain the observed F. langsethiae biomass in winter barley, but rather follows a similar trend as the relative humidity data. The relative humidity data alone cannot be used to adequately explain the observed F. langsethiae and HT-2 + T-2 levels in winter barley heads at harvest. This is because although moisture is known to influence fungal growth and type A trichothecene production, at plant maturity fungal growth is minimal and the effect of environmental condition on the production of mycotoxin, is not entirely a direct effect of moisture but a function of temperature and moisture (Doohan et al. 2003)

On the basis of the findings from this study the development of F. langsethiae in cereals is postulated to follow a specific pattern: At growth stages before head emergence concentrations of F. langsethiae DNA in leaf and stem sub-samples were low and in almost all cases were below the limit of quantification. This suggests the presence of fungal spores on these plant parts rather than mycelial growth. As the cereals develop through head emergence, F. langsethiae DNA concentrations reach quantifiable levels especially in the leaves and head sub-samples with high levels at GS59. This would suggest that head infection takes place at head emergence.. Fusarium head blight infection in cereals generally occurs during anthesis (Parry et al. 1995). The infection of cereal heads by F. langsethiae before flowering could be one of the major differences between the infection process of F. langsethiae and the other FHB pathogens.. In head sub-samples, F. langsethiae DNA concentration continue to increase after GS59 (although this increase was not linear) reaching a maximum at GS92 when plants are about to be harvested. Levels of F. langsethiae DNA tend to increase slightly on leaf sub-samples as well and to a lesser extent on stem sub-samples at GS92. This pattern suggest that in heads, once infection has started, the fungus grows and colonizes the head tissues as the plant develops resulting in an increase in fungal biomass in the head. As the plant senesces (GS92) spore production increases, further increasing DNA concentrations in the head. This is understandable in that as a plant dies, phytopathogenic fungi sporulate and disseminate its spores for propagation the following season (Agrios 2005). The resulting spore dissemination at GS92 is what contributes to the increase in F. langsethiae biomass on leaves and stems at this growth stage. Further studies such as microscopy and spore trap monitoring would be necessary to confirm this proposed epidemiology.

This is the first detailed report on the development of F. langsethiae in wheat, barley, oat and triticale under commercial field cultivation. Apart from confirming the strong association between F. langsethiae and oats among the cereals studied and its ability to produce high levels of HT-2 + T-2 in symptomless oat heads, this study has for the first time, provided information on the development of F. langsethiae in cereals with detailed insight into the timing of infection and development in cereal heads under commercial cultivation. This information will help in the development of new control strategies against F. langsethiae infection in cereals.

Acknowledgement

The first author acknowledges funding of a PhD studentship from HGCA-AHDB and Harper Adams University.

Copyright information

© KNPV 2013

Authors and Affiliations

  1. 1.Harper Adams UniversityNewportUK

Personalised recommendations