CSIRO Plant Industry, GPO Box 1600, Canberra, ACT, 2601.

ABSTRACT

Brassicas are generally regarded as non-hosts for vesicular arbuscular mycorrhizal (VAM) fungi. As colonisation by VAM fungi may substantially increase growth of many crops, it is important to assess whether inclusion of brassicas in a rotation may negatively affect growth of a following crop through lowering the level of VAM colonisation. Hence, the colonisation levels of VAM fungi in 1998 wheat and linola crops sown after a range of Brassica and non-Brassica crops in 1997 were examined at two sites in southern Australia. VAM colonisation was generally lower following Brassica crops, however, this did not negatively affect crop uptake of P or Zn, or crop yield. The general applicability of these results is discussed.

INTRODUCTION

Growth of Brassica crops, particularly B. napus annua (canola) and B. juncea (Indian mustard), in the Australian wheatbelt may enhance yields of following cereal crops through reducing the incidence of root diseases to a greater extent than non-Brassica break-crops (Angus et al. 1991). This effect has been termed ‘biofumigation’ and is thought to result from biocidal compounds being released when glucosinolates (GSLs) in Brassica tissues are hydrolysed to form compounds including isothiocyanates (ITCs). However, the impacts of biofumigation on beneficial organisms such as rhizobia and vesicular-arbuscular mycorrhizal (VAM) fungi has not been investigated.

VAM fungi colonise most crop and pasture species (Thompson 1987) where, in return for photosynthate, they provide the plant with nutrients, particularly P and Zn (Thompson 1994). The potential for VAM fungi to enhance plant growth varies greatly. Plants with finely-branched root systems and copious root hairs (eg cereals and grasses) generally do not rely heavily on VAM fungi to access and absorb nutrients, while plants with relatively coarse, poorly-branched root systems and few root hairs (eg many legumes) may acquire a substantial portion of their nutrients through the actions of VAM fungi (Schweiger et al. 1995).

In the northern wheatbelt of Australia, crops such as linseed are highly dependent on VAM fungi for acquiring Zn and P, with yield reductions if VAM colonisation is inadequate (Thompson 1994). In Australia, low VAM colonisation in crops may result from long bare fallows (Thompson 1994), severe drought (Ryan and Ash 1996), or applications of high rates of soluble P fertilisers (Ryan et al. 1994). However, brassicas may also negatively affect VAM colonisation in subsequent crops. Brassicas are generally considered to be non-hosts for VAM fungi and while a field is planted with a Brassica no new VAM inoculum is produced and the viability of existing inoculum declines. In addition, the release of ITCs by brassicas may directly affect VAM fungi through reducing the viability of VAM inoculum or affecting growth of the fungi (Schreiner and Koide 1993).

This paper investigates whether brassicas are consistently non-mycorrhizal across a range of sites in eastern Australia and reports results from two field trials which examined the effects of brassicas on VAM colonisation, nutrient uptake and grain yield in two following crops thought to differ in VAM dependence; wheat (low) and linola (high).

BRASSICAS: NON-VAM CROPS?

As there are reports from other countries of brassicas being heavily colonised by VAM fungi (Purakayastha et al. 1998), roots from a range of Brassica crops and field sites in eastern Australia were collected in 1998. These roots were stored in 80% ethanol, stained (Grace and Stribley 1991) and the percentage of root length colonised by VAM fungi  VAM (%)  assessed (Giovannetti and Mosse 1980). While colonisation in non-brassicas indicated that VAM fungi were present at all sites, there was no evidence of colonisation in any of the brassicas (Table 1). Thus in this region, brassicas can be considered as non-hosts for VAM fungi.

Table 1: The percentage of root length colonised by VAM fungi in Brassica and non-Brassica crops at four sites in the eastern Australian wheatbelt in 1998 (n=3-5 plants).

Location	Crop		Variety	VAM(%)
Narrabri	Canola	B. napus annua	Karoo, Monty, Oscar	0
(northern NSW)	Indian mustard	B. juncea	JL2	0
	Chickpeas			50
Ginninderra	Canola		Dunkeld	0
(ACT)	Fodder rape	B. napus biennis	Striker, Arran	0, 0
	Indian mustard		Siromo	0
	White mustard	Sinapis alba	Maxi	0
	Fodder radish	Raphanus sativa	Nemex	0
	Linola			15
Cootamundra	12 lines of Indian mustard			all 0
(southern NSW)	Subterranean clover			38
Horsham	30 lines of canola and Indian mustard			all 0
(VIC)	Perennial ryegrass			47

EFFECT OF BRASSICAS ON FOLLOWING CROPS

Morangarell Field Site

Materials and methods. To minimise disease effects, the trial was situated on an uncultivated stock-route at Morangarell, near Temora (34º27’S, 147º32’E). Grass was burned and the area cultivated in February 1997. The soil was a grey cracking clay with a pH in water of 7.8, 1.2% organic carbon, 7 mg/kg of extractable P (Colwell), 5 mg/kg nitrate N and 0.5 mg/kg DTPA Zn.

In 1997, 20 x 2 m plots of several crops were sown on May 14: canola cvs. Monty (low root GSLs) and Karoo (high GSLs); Indian mustard cvs. Siromo (low GSLs) and yellow mustard (high GSLs); linola cv. Argyle, a Karoo/Argyle mixture and wheat cv. Janz. These were sown with two rates of P fertiliser (5 and 20 kg/ha of P as triple superphosphate). An unfertilised bare fallow treatment was also included. Treatments were replicated three times, each replicate forming a block within which treatments were randomly arranged, but with the low and high P plots for each crop adjacent. The trial was harvested in December 1997 and the site cultivated again in April 1998.

On May 16, 1998, 10 m of each plot was sown to durum wheat cv. Wallaroy and 10 m to linola cv. Argyle. Linola was included as it was thought to be a more dependent host than wheat. Five kg/ha of P as triple superphosphate and 50 kg/ha of N as urea were applied at sowing and N reapplied at the end of August. Heavy rain in July and August resulted in sections of the trial being under water for a number of days on several occasions, causing patchiness which persisted through to harvest. In August (tillering), 20 plants per plot were harvested. Shoots were dried at 70°C for three days, weighed, ground, and nutrient concentrations determined by X-ray spectrometry. Roots were stored and assessed for VAM as described above. Plots were mechanically harvested for grain yield. Results were analysed by ANOVA using Genstat®, with a visual estimation of the severity of water-logging in each plot included as a co-variate. The co-variate was always significant, however a large amount of variation in the data remained unaccounted for in all models.

Results. Yields in 1997 were 0.3-0.5 t/ha for the brassicas, 0.4-0.5 t/ha for the linola and 0.5-1.0 t/ha for the wheat; they were limited by lack of N and a dry end to the season. While the brassicas were not colonised by VAM fungi, VAM% was >30% for the linola and >40% for the wheat. In 1998, precrop did not have a significant effect on VAM colonisation, shoot P or Zn concentration or final crop yield of wheat or linola (Table 2). However, there was a tendency for VAM colonisation to be highest after wheat and lowest after brassicas.

Table 2: VAM colonisation, shoot P and shoot Zn concentrations at week 12, and final grain yields of wheat and linola at Morangarell in 1998 following a range of 1997 pre-crops (estimated means, precrop did not have a significant effect on any variable at p< 0.1).

Precrop	Wheat	Linola	Canola		+ linola	Mustard		Fallow
GSL level	-	-	low	high	high	low	high
Wheat
VAM (%)	34	19	16	18	23	34	16	18
Shoot P (%)	0.19	0.20	0.18	0.18	0.18	0.18	0.19	0.17
Shoot Zn (%)	19.5	19.2	17.5	18.2	15.9	17.3	13.8	19.2
Yield (t/ha)	2.4	2.3	2.5	2.4	2.1	2.1	2.4	2.3
Linola
VAM (%)	31	24	16	18	20	18	22	14
Shoot P (%)	0.20	0.21	0.21	0.23	0.20	0.21	0.22	0.19
Shoot Zn (%)	15.4	20.3	22.2	20.4	20.5	17.4	20.9	20.3
Yield (t/ha)	0.71	0.70	0.81	0.86	0.71	0.64	0.68	0.77

Cowra Field Site

Materials and methods. A NSW Agriculture field trial at Cowra (148°41’E, 33°50’S) designed to assess the effects of brassicas on the cereal disease take-all (Gaeumannomyces graminis Var. tritici) was also sampled. The trial was situated on a clay loam with a pH in water of 7.3, 2.7% organic carbon and 0.11% total N. The site was inoculated with take-all in 1996 to establish high levels of inoculum.

In 1997, 11 x 2 m plots of the following crops were sown on May 9: wheat cv. Janz and wheat + metham; canola cvs. Oscar (low root GSLs), Dunkeld (medium GSLs), Tamara (very high GSLs) and Tamara + wheat; linola cv. Argyle; and faba beans. Twenty-eight kg/ha of N and 16 kg/ha of P were applied at sowing. Treatments were arranged in a randomised block design with eight replicates. The crops were irrigated when necessary. In 1998, stubble was burnt and the site cultivated twice. On April 9, each plot was sown to wheat cv. Whistler with 28 kg/ha of N and 16 kg/ha of P. In August (end of tillering), 25 plants per plot were removed. Shoots were dried at 70°C for 3 days, weighed, ground, and nutrient concentrations determined as above. Roots were assessed for VAM colonisation as described above. Plots were harvested on December 8 and the grain threshed and weighed. Results were analysed by ANOVA using Genstat®.

Results. VAM colonisation was highest in the wheat-wheat treatment and this was halved by application of the fumigant metham (Table 3). The canola-wheat treatment had very low VAM colonisation levels and this did not differ with the level of root GSL. Including wheat with the 1997 canola did not increase colonisation in the subsequent wheat crop. In spite of being VAM hosts, there were lower levels of VAM colonisation in wheat after faba beans and linola than in wheat after wheat. Thus, with the exception of fabas, mycorrhizal precrops substantially increased the level of VAM colonisation in subsequent crops compared with a non-mycorrhizal precrop.

The most obvious trends in the shoot P and Zn data were the lower concentrations in the wheat-wheat treatment; grain yields in 1997 do not suggest that this was due to greater removal of nutrients in these plots. Yield was also lowest in the wheat-wheat treatment, corresponding with a far greater incidence of take-all (results not shown). The higher VAM colonisation in the wheat-wheat treatment was unable to compensate for this effect

Table 3: VAM colonisation and shoot P and Zn concentrations at anthesis, and final grain yield of wheat at Cowra in 1998 following a range of 1997 pre-crops (estimated means, LSD at p=0.05 and a p value for the effect of precrop).

Precrop	Wheat	+ metham	Linola	Faba beans	Canola			+ wheat	LSD	p
GSLs					low	medium	very high	very high
VAM (%)	29	14	16	6	7	5	6	6	6.5	<0.001
P (%)	0.23	0.27	0.26	0.32	0.28	0.26	0.29	0.30	0.02	<0.001
Zn (%)	12.0	15.5	14.2	14.4	13.8	13.9	15.7	14.1	1.73	<0.003
Yield (t/ha)	4.1	4.6	4.9	4.8	4.8	4.9	4.6	4.4	0.4	<0.001

DISCUSSION

Overall it appeared that the non-mycorrhizal nature of the brassicas resulted in a reduced level of VAM colonisation in following crops. While the effects of all brassicas were similar, despite different types and levels of GSLs (Kirkegaard and Sarwar 1998), it is possible that differences due to GSLs may have been more apparent if soil inoculum levels, and therefore colonisation levels, had been higher.

At Cowra there was lower colonisation in wheat following the mycorrhizal linola and faba beans than in the wheat-wheat treatment. This was unexpected, as both crops should be dependent hosts (Thompson 1994). Higher shoot P concentrations in these treatments could have been responsible, as high levels of soil and plant P are often reported to suppress VAM colonisation (Ryan and Ash 1994). Although, curiously, the roots of wheat grown after faba beans at Cowra looked different to wheat roots in all the other treatments, with the cortex sloughing off during staining. Perhaps also, in 1997 the linola and fabas may have supported a smaller root biomass than the wheat and therefore left a smaller biomass of colonised root to act as inoculum.

The reduction in VAM colonisation in crops grown after brassicas did not reduce uptake of P or Zn or negatively affect crop yield. Indeed, at Cowra, P and Zn concentrations were lowest in the wheat-wheat treatment which had the highest level of VAM colonisation. This contrasts with the results of Thompson (1987) where, on cracking clay soils in the northern wheatbelt, reduced VAM colonisation after long fallows was shown to contribute to nutrient deficiencies in following crops. It is possible that these contrasting results are due to the waterlogging at Morangarell and high disease levels at Cowra masking the effects of reduced VAM colonisation; for instance, at Cowra, the take-all may have substantially reduced the ability of the root systems to access nutrients.

Differences in factors such as VAM species, soil type or other environmental variables may also result in the importance of VAM fungi for crop growth varying between sites. For instance, Australian studies using three different soil types have reported the effect on wheat growth of removing VAM fungi as negative (Thompson 1990), neutral (Ryan 1998) and positive (Graham and Abbott 1998); the last study indicating that under some circumstances the presence of VAM fungi may actually have a negative effect on wheat growth. Further work is being conducted to clarify the importance of soil type, region and dependency of following crop on the effect of reduced VAM levels after Brassica crops on crop nutrition and yield.

ACKNOWLEDGEMENTS

Kate McCormick, Kate Light (Ag-Seeds), John Holland and Rex Oram generously allowed us to sample their field trials. The Morangarell Landcare Group assisted with the trials at Morangarell and the Cowra trials were managed by NSW Agriculture. This project is funded by the GRDC.

REFERENCES

1. Angus, J. F., Gardner, P. A., Kirkegaard, J. A., and Howe, G. N. (1991). Productivity and break-crop effect of winter growing oilseeds. Australian Journal of Experimental Agriculture 31, 669-77.

2. Giovannetti, M., and Mosse, B. (1980). An evaluation of techniques for measuring VAM infection in roots. New Phytologist 84, 489-500.

3. Grace, C., and Stribley, D. P. (1991). A safer procedure for routine staining of vesicular arbuscular mycorrhizal fungi. Mycological Research 95, 1160-2.

4. Graham, J. H., and Abbott, L. (1998). Functional diversity of mycorrhizal fungi in the wheat rhizosphere. In 'Proceedings of the ICOM II Conference'. p72-3. (Uppsala, Sweden.)

5. Kirkegaard, J. A., and Sarwar, M. (1998). Biofumigation potential of brassicas I. Variation in glucosinolate profiles of diverse field-grown brassicas. Plant and Soil 201, 71-89.

6. Purakayastha, T. J., Singh, C. S., and Chhonkar, P. K. (1998). Growth and iron nutrition of broccoli (Brassica oleracea L. var. italica Plenck), grown in a Typic Ustochrept, as influenced by vesicular-arbuscular mycorrhizal fungi in the presence of pyrite and farmyard manure. Biology and Fertility of Soils 27, 35-8.

7. Ryan, M. H. (1998). The Ecology of VAM Fungi in Contrasting Australian Agricultural Systems. PhD. Australian National University.

8. Ryan, M. H., and Ash, J. E. (1996). Colonisation of wheat in southern NSW by VA-mycorrhizal (VAM) fungi is significantly reduced by drought. Australian Journal of Experimental Agriculture 36, 563-9.

9. Ryan, M. H., Chilvers, G. A., and Dumaresq, D. C. (1994). Colonisation of wheat by VA-mycorrhizal fungi was found to be higher on a farm managed in an organic manner than on a conventional neighbour. Plant and Soil 160, 33-40.

10. Schreiner, R. P., and Koide, R. T. (1993). Mustards, mustard oils and mycorrhizas. New Phytologist 123, 107-13.

11. Schweiger, P. F., Robson, A. D., and Barrow, N. J. (1995). Root hair length determines beneficial effect of a Glomus species on shoot growth of some pasture species. New Phytologist 131, 247-54.

12. Thompson, J. P. (1987). Decline of vesicular-arbuscular mycorrhizas in long fallow disorder of field crops and its expression in phosphorus deficiency in sunflower. Australian Journal of Agricultural Research 38, 847-67.

13. Thompson, J. P. (1990). Soil sterilisation methods to show VA-mycorrhizal aid P and Zn nutrition of wheat in vertisols. Soil Biology and Biochemistry 22, 229-40.

14. Thompson, J. P. (1994). Inoculation with vesicular-arbuscular mycorrhizal fungi from cropped soil overcomes long-fallow disorder of linseed (Linum usitatissimum L.) by improving P and Zn uptake. Soil Biology and Biochemistry 26, 1133-43.