Previous PageTable Of ContentsNext Page

Crop sequence to manage crown rot in cereals in south-eastern Australia

Margaret Evans1, Grant Hollaway2, Hugh Wallwork1, Ray Correll3

1 Field Crop Pathology Unit, South Australian Research and Development Institute, GPO Box 397, Adelaide, SA 5001
2
Department of Primary Industries, Private Bag 260, Horsham, Vic 3401
3
RHO Environmetrics, PO Box 366, Highgate, SA 5063

Abstract

Crown rot is a fungal disease which affects all winter-growing cereals and is estimated to cost the Australian grains industry up to $56 million per annum. In south-eastern Australia, the main pathogens causing crown rot are Fusarium pseudograminearum (Fp) and F. culmorum (Fc). Management of this disease relies heavily on a non-host break but the relative effects of each crop/pasture type on Fp and Fc levels are not well understood. Quantitative DNA assays specific to Fp and Fc were used to monitor pathogen levels in plant material on and in soil. Effects of a range of cereals and non-cereals on Fp and Fc inoculum were assessed at two sites in Victoria and one in South Australia during the period 2003 to 2005. Fp and Fc responded to treatments in a similar manner. The relative effects of the different previous crop and pasture treatments were consistent at all sites, although some sites were more responsive than others to treatments. All commercial cereals increased levels of Fp and Fc - durum wheat increased levels most, followed by barley, oats and bread wheat. All non-cereals decreased levels of Fp and Fc - fallow was most effective, followed by peas and canola. There is potential to predict the effect of sequence options at a site if the background Fp and Fc levels and their variability are known.

Key Words

Disease, pathogen, inoculum, functional relationships, ribosomal DNA probe.

Introduction

Crown rot is a fungal disease which affects all winter growing cereals and is estimated to cost the Australian grains industry up to $56 million per annum (Brennan and Murray 1998). The main pathogens causing crown rot in Victoria (Vic) and South Australia (SA) are Fusarium pseudograminearum (Fp) and Fusarium culmorum (Fc) (Backhouse et al. 2004).

Both Fp and Fc can survive saprophytically on infected plant residues for at least 3-4 years. Neither pathogen grows through soil and infection normally occurs by direct contact between infected plant residues and uninfected plants. Crop residue retention, direct seeding and an increase in cereal (particularly durum wheat) in cropping sequences all contribute to increased incidence and severity of crown rot (Duveiller et al. 2007).

Since varietal resistance, seed treatment and fungicidal sprays are not available for crown rot management, farmers rely principally on crop sequences (Lamprecht et al. 2006) and other cultural techniques to minimise the impact of crown rot on cereal production. Non-cereal breaks such as canola, chickpeas and legume pasture are known to contribute to reduced crown rot expression in subsequent cereal crops (Sturz and Bernier 1989, Kirkegaard et al. 2004, Lamprecht et al. 2006). For good crown rot management, it is important that the relative effects of each crop/pasture type on Fp and Fc are understood and quantified under the conditions and farming systems of south-eastern Australia.

The purpose of this study was to compare the effects of a range of cereals and non-cereals on levels of Fp and Fc in Vic and SA to determine if crop sequence can be used to manage crown rot.

Materials and methods

Treatments (Table 1) were applied in paddocks either naturally infected with high levels of Fusarium spp. (Birchip, Vic and Cambrai, SA) or artificially infected in the previous year by sowing durum wheat seed artificially infected with either Fc or Fp (Longerenong, Vic). Treatments were applied using a randomised block design with 4 (Longerenong) or 6 (Cambrai and Birchip) replicates. Sites were managed according to local agronomic recommendations. Field experiments were direct drilled (except for one treatment at Longerenong) and plant residues were retained. Herbicides were applied as needed (no residual herbicides were used) and weeds were effectively controlled.

During March or April of the year following treatment, samples were collected to quantify Fusarium spp. levels. From each plot, forty cores were taken to a depth of 100 mm using a 15 mm diameter AccucoreTM soil sampler (Spurr Soil Probes, Brighton North, South Australia). All cores were taken over crop rows and included plant residues on the soil surface as well as the soil and the plant residues in the soil. Samples were sent to the Root Disease Testing Service (South Australian Research and Development Institute, Adelaide, South Australia) where PCR assays using ribosomal DNA probe sequences specific to Fp and Fc were applied to total DNA extracted from the samples (Ophel-Keller et al. 2008). Results were converted via a quantitative DNA standard and are expressed as picograms of Fp and Fc present per gram of sample.

Table 1. Crop, variety and non-crop treatments used at each site to determine their effects on levels of Fusarium spp. MS - moderately susceptible (bread wheat); S - susceptible (bread wheat); VS - very susceptible (durum wheat). H, host; NH, non-host. na –treatment not applied at this site. Note – results for vetch, medic, triticale and cultivated peas are not presented as they were sown at one site only.

Treatment

Description

Cambrai

Birchip

Longerenong

Barley

H

Schooner

SloopVic

SloopVic

Oats

H

Mortlock

na

Marloo

Wheat MS

H, mod. susc.

Kukri

na

na

Wheat S

H, susceptible

Frame

Yitpi

Yitpi

Wheat VS

H, very susc.

Tamaroi

Tamaroi

Tamaroi

Fallow

No host

Chemical

Mechanical

Mechanical

Canola

NH (oilseed)

Outback

Saphire

Saphire

Field peas

NH (legume)

Parafield

Kaspa

Kaspa

Log10 transformed DNA levels of Fp and Fc after treatment from individual sites were subjected to analysis of variance by GenStat Version 8.2. Multiple comparisons were conducted using Fisher’s protected l.s.d. test. A meta-analysis was performed for Fp data from all sites by the use of estimating equations following the method of Morton et al. (2007) to combine the information from several experiments. This model assumes the treatment effects on the log scale have a linear relationship. The meta-analysis was performed for Fp only, as there was most information for this species and because at Birchip, Fp and Fc (present naturally in the paddock) behaved in a similar manner in response to treatments as shown by a strong linear correlation (R2 = 0.776) between the species.

Results and discussion

In the year after treatment, highest levels of Fusarium spp. occurred after cereals and lower levels occurred after no host or non-hosts (Table 2). Apart from this difference between cereals and non-cereals, no other treatment differences were evident.

Table 2. Effect of cereal and non-cereal treatments on levels of Fusarium spp. (Log10 pg fungal DNA/g of sample) the season after treatments were applied. Fp, F. pseudograminearum; Fc, F. culmorum; na –treatment not applied at this site. MS - moderately susceptible (bread wheat); S - susceptible (bread wheat); VS - very susceptible (durum wheat). Means in the same column followed by the same letter are not significantly different (P=0.05). Note – results for vetch, medic, triticale and cultivated peas are not presented as they were sown at one site only.

Treatment

Cambrai

Birchip

Longerenong

 

Fp

Fp

Fc

Fp

Fc

Fallow

2.19 a

2.36 a

1.29 a

2.29 a

2.38 ab

Canola

2.29 ab

2.78 ab

1.33 a

2.91 cd

2.51 ab

Peas

2.57 abcd

2.32 a

1.24 a

2.61 bc

2.42 ab

Wheat MS

2.78 cdef

na

na

3.11 d

2.57 bc

Wheat S

2.94 def

3.23 bcd

1.53 a

3.02 cd

2.56 abc

Wheat VS

3.09 ef

3.62 d

1.94 bc

3.54 e

2.85 d

Barley

3.18 f

3.48 cd

2.20 c

2.88 cd

2.36 ab

Oats

3.24 f

na

na

3.01 d

2.79 cd

LSD (P=0.05)

0.45

0.54

0.45

0.39

0.26

This is inconsistent with previous studies, where durum wheat has been a critical trigger for crown rot problems (Duveiller et al. 2007) and some non-cereals (e.g. canola) reduced levels of Fp significantly more than others (Kirkegaard et al. 2004). The highly variable distribution of Fusarium spp. (Heap and McKay 2004) may be contributing to this inability to differentiate statistically between specific treatments at individual sites. For this reason and also to assess the consistency of response to treatments across sites, meta-analysis of data from all sites was undertaken to further explore the practical implications for industry of results from this research.

Meta-analysis of F. pseudograminearum data from all sites

There was a good relationship between actual (measured) and fitted (predicted) values of Fp levels (Table 3) which indicates that treatment effects on a Log10 scale have a linear relationship. This validated the underlying assumption of the meta-analysis model.

Table 3. Comparison of fitted (predicted) and actual (measured) F. pseudograminearum levels (Log10 pg fungal DNA/g of sample) the season after treatments were applied. The fitted value for each treatment at a site is a weighted average determined by the variability of the trial in relation to the other trials. Differences of ≤0.1 between fitted and observed values are within the error from normal “noise” seen in field results. na –treatment not applied at this site. MS - moderately susceptible (bread wheat); S - susceptible (bread wheat); VS - very susceptible (durum wheat). Note – results for vetch, medic, triticale and cultivated peas are not presented as they were sown at one site only.

Treatment

Cambrai

Birchip

Longerenong

 

Actual

Fitted

Actual

Fitted

Actual

Fitted

Fallow

2.19

2.31

2.37

2.33

2.29

2.44

Peas

2.58

2.45

2.33

2.53

2.61

2.58

Canola

2.29

2.57

2.78

2.69

2.92

2.68

Wheat MS

2.79

2.87

na

3.10

3.11

2.95

Wheat S

2.94

2.95

3.23

3.21

3.02

3.03

Oats

3.24

3.04

na

3.33

3.01

3.11

Barley

3.19

3.11

3.48

3.43

2.88

3.18

Wheat VS

3.10

3.23

3.62

3.59

3.54

3.28

Fallow reduced levels of Fp to approximately a third, while peas and canola approximately halved Fp levels (Table 4). Durum wheat more than trebled levels of Fp (Table 4), with other cereal types increasing levels much less. These findings support those already reported in the literature for durum wheat (Duveiller et al. 2007).

Table 4. Effects of treatments on levels of F. pseudograminearum, assessed using a standardised measure of treatment effects (ξ) estimated over all the sites. Converting ξ to a multiplicative scale allows calculation of the magnitude of changes in fungal DNA levels. MS - moderately susceptible (bread wheat); S - susceptible (bread wheat); VS - very susceptible (durum wheat). Note – results for vetch, medic, triticale and cultivated peas are not presented as they were sown at one site only.

Treatment

Treatment effects

 

ξ (Log10 scale)

Multiplicative scale

Fallow

-0.42

0.38

Peas

-0.27

0.53

Canola

-0.16

0.69

Wheat MS

0.14

1.39

Wheat S

0.23

1.68

Oats

0.31

2.01

Barley

0.39

2.44

Wheat VS

0.50

3.18

Although it is implicit in these results that treatments will behave similarly in relation to one another at all sites, the magnitude of the overall response to treatments varied between sites. Birchip was the most responsive site to treatments, having a responsiveness rating of 1.36 compared with ratings of 0.9 and 1.0 for the other sites. Responsiveness may be related to high starting levels of Fp (Birchip had the highest starting levels of Fp), site characteristics (e.g. soil type), environment (e.g. rainfall) or a combination of factors.

Predicting responses of sites to treatments

The good relationship between actual and fitted values of Fp levels (Table 4) gives an assurance that predictions for those treatments not tested at all sites are well estimated by the fitted values. This gives rise to the possibility of predicting the effect of crop sequences on levels of Fp. Being able to predict the effects of different non-cereals on changes in Fp levels would have particular practical value in assessing the length of non-cereal break needed prior to sowing highly susceptible crops such as durum wheat. The reliability of predicting changes in Fp levels needs to be validated in the field before it is considered for commercial use. Validation will need to include assessment of the effects of site responsiveness and seasonal conditions.

Conclusions

Non-cereals decrease Fp levels (fallow more than peas and canola) and cereals increase Fp levels (durum wheat more than barley and oats more than bread wheat). Relative to one another, cereal and non-cereal treatments behave similarly across sites and seasons. The magnitude of reduction or increase in Fusarium spp. levels will depend on the responsiveness of the site. Once the responsiveness of a site is known it may be possible to predict the changes in Fp levels over the next cropping season under different cereal and non-cereal options. The reliability of the predictive process needs to be validated in the field to assess its commercial usefulness.

Acknowledgements

This work was supported by the GRDC. Graham Exell, Greg Naglis, Jonathan Bretag, Dennis Ward and Jose Alvarado are thanked for their technical assistance and Alan McKay for provision of Fusarium quantification. Much of this work was initiated by Jeremy Dennis.

References

Backhouse D, Abubakar AA, Burgess LW, Dennis JI, Hollaway GJ, Wildermuth GB, Wallwork H and Henry F (2004). Survey of Fusarium species associated with crown rot of wheat and barley in eastern Australia. Australasian Plant Pathology 33, 255-261.

Brennan JP and Murray GM (1998). Economic importance of wheat diseases in Australia. NSW Agriculture, Wagga Wagga.

Duveiller E, Singh RP and Nicol JM (2007). The Challenges of maintaining wheat productivity: pests, diseases, and potential epidemics. Euphytica 157, 417-430.

Heap JW and McKay AC (2004). Spatial distribution of soilborne disease inoculum DNA in cereal crops and implications for soil sampling protocols. In ‘3rd Australasian Soilborne Diseases Symposium’ p. 106. (Rowland Flat, South Australia)

Kirkegaard JA, Simpfendorfer S, Holland J, Bambach R, Moore KJ and Rebetzke GJ (2004). Effect of previous crops on crown rot and yield of durum and bread wheat in northern NSW. Australian Journal of Agricultural Research 55, 321-334.

Lamprecht SC, Marasas WFO, Hardy MB and Calitz FJ (2006). Effect of crop rotation on crown rot and the incidence of Fusarium pseudogramiearum on wheat in the Western Cape, South Africa. Australasian Plant Pathology 35, 419-426.

Morton R, Warne MStJ and Correll RL (2007). Simultaneous prediction of toxicity of multiple chemicals to multiple species using multi-dimensional functional relationships. Environmetrics 18, 1–20.

Ophel-Keller K, McKay A, Hartley D, Herdina, and Curran J (2008). Development of a routine DNA-based testing service for soilborne diseases in Australia. Australasian Plant Pathology 37, 243-253.

Sturz AV and Bernier CC (1989). Influence of crop rotations on winter wheat growth and yield in relation to the dynamics of pathogenic crown and root rot fungal complexes. Canadian Journal of Plant Pathology 11, 114-121.

Previous PageTop Of PageNext Page