Overview Objectives

This project studied the ‘biofumigation’ properties of brassicas as rotation, companion or green manure crops, and the use of these properties in managing two important pathogens of tropical vegetable production systems - bacterial wilt (Ralstonia solanacearum) and root-knot nematodes (Meloidogyne spp.) - in Australia and the Philippines.

Project Background and Objectives

This project arose from recent research to investigate the potential of biofumigation by brassicaceous crops for pest and disease suppression in various farming systems. In Australia observations of superior wheat growth following Brassica rotation crops led to the recognition that root-derived biocides released in soil can suppress soil-borne pathogens. From broadacre agriculture an interest in the concept spread into horticultural industries - driven mainly by the search for alternatives to methyl bromide (MB) and other synthetic fumigants, reports of enhanced biodegradation of the major chemical alternatives to MB, and a general desire to move towards more environmentally benign production systems.

The most serious soil-borne diseases of solanaceous vegetable crops such as eggplant, tomato and potatoes in tropical environments are bacterial wilt (BW) caused by Ralstonia solanacearum and to a lesser extent root knot nematodes (Meloidogyne spp.). Biofumigation using Indian mustard (Brassica juncea) as green manure has been found effective in reducing the level of BW in the soil, and lessening the severity of the disease in a following tobacco crop. A commercially available Indian mustard biofumigant green manure was shown to significantly reduce bacterial wilt in a following potato crop, resulting in spectacular yield increases (from 0.3 to 22 t/ha). The fact that such results were obtained with no purposeful selection of brassicas for high biofumigant properties indicates there may be significant scope to improve the level of suppression achieved.

Progress Reports (Year 1, 2, 3 etc)

This project aims to evaluate the potential of “biofumigation” using Brassicaceous rotation, companion or green manure crops as a component of the integrated management of important soil-borne pests of solanaceous vegetables in tropical production systems. Biofumigation refers to the process by which soil-borne pests and pathogens are suppressed by naturally occurring biocides released in soil when tissues of Brassicaceous plants decompose in soil. The major compounds implicated are isothiocyanates (ITCs) released when glucosinolates (GSLs) in the tissues are hydrolysed.

The focus for this Bilateral project is two important pathogens of tropical vegetable production systems in Australia and The Philippines, namely Bacterial Wilt (Ralstonia solanacearum) and root-knot nematodes (Meloidogyne spp.). The project involves collaboration between groups at QDPI (Mareeba and South Johnstone) Australia, National Crop Protection Centre at UPLB, The Philippines and CSIRO Plant Industry, Canberra.

The key objectives in the first year of the project were (1) to identify brassicas suitable for evaluation as biofumigants in tropical production systems; (2) to evaluate the suppressive potential of their tissues in laboratory and glasshouse studies and (3) identify the compounds responsible for pest suppression. Preliminary field trials were also carried out in Australia.

In the Philippines, selection of suitable brassicas for preliminary assessments was based on (1) commonly grown species that were adapted to various growing regions, (2) species with a wide spectrum of GSL types, and (3) availability of good quality seed. In Australia, previous studies had indicated which brassicas were suited to tropical conditions and some commercial biofumigants are already available. We used this information together with that from previous studies on bacterial wilt in tobacco to guide selection for field experiments. We also included other varieties to broaden the range of GSL types.

The chopped leaf tissues of all Brassica species tested significantly reduced the populations of bacterial wilt (BW) in sealed laboratory containers when compared with sorghum, tomato and soil only controls. There were differences in the suppression by tissues of different Brassica species, but no clear correlation between suppression and the type or levels of GSL in the leaf tissues. In studies using tissues incorporated into soil and incubated in muslin/nylon bags buried in the field, there was also significant suppression by the Brassica tissues and differences in the pattern and rate of suppression. The correlation between suppression and GSL concentration was not strong, however there were indications that species high in 2-propenyl ITC (mustard/mustaza) caused more rapid suppression. However the significant suppression caused by tissues with very low GSL content, and by leaves of tobacco and sweet potato indicate other compounds released from chopped leaves are suppressive to BW.

Studies on suppression of BW using pure propenyl-ITC added in a canola oil formulation indicate that an equivalent of 200 - 1000 nmole ITC per g of soil was required to reduce the BW population to very low levels. The amount of potential ITC present in incorporated mustard tissues in our incorporation experiments (40 umole GSL/g tissue incorporated at 5%) was around 2000 nmol ITC per g of soil. This suggests that ITC-related suppression is likely if a significant proportion of the GSL was released as ITC. However our studies on ITC release efficiency from incorporated tissues indicate that less than 5 % of ITC is released by the tissue disruption achieved by a rotary hoe or rough chopping used in our experiments. Thus poor correlation between GSL content and suppression may arise from low ITC release efficiency rather than the pathogens insensitivity. Release efficieny was increased to 40% by treatments such as mulching and irrigating to increase the GSL hydrolysis indicating significant potential to improve ITC release by focussing on incorporation strategies.

Brassicas were generally poorer hosts for nematodes than tomatoes (good host) but only a few species, particularly raddish was as effective in reducing nematode populations to the same extent as poor hosts such as sorghum. However incorporated leaf tissues of all brassicas reduced nematode numbers by 40% or more, and the most suppressive such as raddish reduced the population by 95%. Significant non-GSL related toxicity of Brassica leaves to nematodes has been reported previously and our data would support those findings.

The results indicate the need for further laboratory experiments to establish the role of GSL-hydrolysis products in pathogen suppression, and to investigate other suppressive compounds that are released by chopped leaf tissues. Improving the release efficiency of ITCs from the incorporated tissues may reveal better correlations that are currently masked by inefficient ITC release following incorporation.

In summary, the high level of suppression achieved in inoculated soil for both pathogens is encouraging, irrespective of the role of GSLs. However these results will need to be confirmed in naturally infected soil under field conditions, and this will be a focus of the studies in Year 2. The results of these studies should provide information that will allow us to more confidently extend the concept into farmer’s fields for testing in the final Year.

The two key activities during this reporting period were:

(1) stage 2 laboratory and glasshouse studies to evaluate the biocidal activity of various Brassica tissues and to identify the most effective strategies to enhance biofumigation in the field (Objectives 2.2, 3, and 4).

(2) to evaluate biofumigation for control of Bacterial Wilt and Root Knot Nematodes in the field (Objective 5).

(1) Stage 2 laboratory and glasshouse studies

In general, the studies confirmed the suppressive potential of Brassica tissues reported in year 1 with significant reductions in both BW and RKN in soils amended with Brassica tissues compared with non-Brassica controls. In some cases, macerated Brassica tissues amended at realistic field application rates (5% W/W) were as effective as commercial fumigants, reducing both pests to undetectable levels, while non-Brassica controls had no impact.

We found good evidence that ITCs and other volatile GSL-derived compounds played a significant role in the short-term (<10 days) suppressive impacts of Brassica tissues on both pathogens. For example the effectiveness of suppression by mustards was rapid (occurred within 3 days) and was increased by better tissue disruption (freeze/thaw or blend), higher GSL levels in tissues and/or sealing of containers. In addition, the suppressive effect was not evident if blended tissue was left open to the air for 24 hrs before incorporation. However suppression by some brassicas low in GSLs suggests other compounds unique to brassicas (probably S-containing) may also play a role.

In longer-term incubation studies (30 days) or where tissues were not macerated, non-Brassica amendments (sweet potato, spinach, bean, tomato, sorghum) were sometimes as effective as Brassica amendments against BW, and invariably reduced RKN. Thus the general longer-term impacts of organic amendments can also play a role in pest suppression presumably due to induced changes in soil microbial community favouring pest antagonists.

(2) Field studies

In the Philippines, three field studies were conducted to assess the impacts of selected brassicas on BW, and RKN were also assessed in one of these experiments. In the first experiment at low elevation, incorporated radish, cabbage and sweet potato leaves had no impact on bacterial wilt (35-45% plants wilted). In a second experiment at mid-elevation (Mayjayjay), radish and mustard leaves and mustard roots all reduced wilting significantly (from 60% in controls to 32-45%). In the third experiment at Mayjayjay, tissues from cabbage, broccoli, cauliflower, and radish reduced wilt incidence at 6 weeks in tomatoes (from 21% to 2.8-8.4%) while at 8 weeks the broccoli and cauliflower treatments remained effective (11-19% compared with 46% in sweet potato control). In the same experiment, RKN numbers were reduced by all brassicas (57-87% reduction), but this was not different from the effects of a sweet potato control (84% reduction).

In Australia, three field experiments were conducted, two on sites without disease (Walkamin and Kairi) to determine impacts of a mustard green manure in the absence of disease, and one at Southedge Research Station BW nursery under very high BW pressure. At Walkamin and Kairi the yield of potatoes, tomatoes and capsicums were up to 20% higher following mustard than following either legume or weed-fallow controls indicating positive rather than negative impacts of Brassica green manures in the absence of disease.

At Southedge, five currently available Brassica biofumigant green manures were compared with soybean and weedy fallow for their impact on BW and RKN and yield of a following eggplant crop. No plants in the weedy fallow control plots reach flowering or fruiting stages due to severe wilting. At the end of the experiment, the incidence of disease in the soybean control ranged from 69-100%, the severity from 3.4 - 4.9 (0-5 scale) and the mean eggplant yield was 4 kg. The impact of the five brassicas compared to soybean ranged from no difference (2 varieties), to significant reductions in incidence (down to 34%), severity (down to 1.68), and a mean yield of 22 kg in the best Brassica treatment. Radish, fodder rape and mustard biofumigants were all effective in delaying disease onset, reducing the incidence and severity of disease and increasing eggplant yield. RKN numbers were not different to controls following two of the biofumigants (radish and mustard) and increased after the other three.

In summary, incorporation of some Brassica biofumigants has reduced bacterial wilt by 50-60% in three of the four field experiments reported compared to non-Brassica controls. RKN numbers in the roots of vegetables following brassicas was similar to non-Brassica controls in one experiment but increased after some brassicas in a second experiment. The results indicate promising potential to develop biofumigation as a component of integrated control of these pathogens, however significant variation among replicates in the field studies and inconsistencies in the performance of some Brassica species will need to be addressed to provide more confidence in promoting robust strategies to growers.

The major activity during this reporting period was field evaluation of the most promising brassicas for biofumigation of bacterial wilt and root knot nematodes and development of protocols to improve field efficacy. This included some preliminary experiments on commercial farms and small-scale farmer fields. Some laboratory experiments to finalise investigations to identify the active components of incorporated green manures were conducted.
Bacterial Wilt: The 2003 field experiment conducted at Southedge Research farm near Mareeba, Australia provided the most promising results to indicate the effectiveness of selected brassica biofumigants against bacterial wilt. The incorporated brassica green manures delayed the onset, significantly reduced the incidence and severity of wilt and increased tomato yield compared with the weedy fallow or soybean controls. The mustard (Brassica juncea) treatments were the most effective and the rape treatments (B. napus) the least, consistent with previous laboratory experiments on suppression by volatiles released from leaf tissues of these species. The best mustard treatments (Nemfix) reduced wilt severity from 4.8 to 2 (0-5 scale) and increased tomato yield from 2.5 t/ha to 20 t/ha. These results suggest that where Brassica material is grown in sufficient quantity (5 kg/m2 fresh material), and managed to achieve high levels of ITC release (i.e. tissue maceration, adequate and rapid incorporation and irrigation), that significant suppression of bacterial wilt can be achieved in the field, consistent with that demonstrated in laboratory and glasshouse studies.
In the Philippines, where it was more difficult to achieve all of these conditions in the field experiments, there were no consistent effects of Brassica amendments on the incidence of wilting in following vegetable crops in the five field experiments conducted in 2003-04. Generally, non-brassica and brassica treatments had similar impacts on wilt incidence, and no treatment performed consistently well across sites. Of the brassicas tested, radish and broccoli showed the most promise, each providing significantly lower wilt incidence than the non-brassica control at one site (each reduced wilt by 50 per cent). Covering the plots with black plastic following incorporation did not influence bacterial wilt levels. The reduced and variable impacts of brassicas in these experiments may have arisen from insufficient mixing of the material through the soil, as the macerated Brassica material was restricted to a band and covered, which would leave significant amounts of soil ‘untreated’ in the vicinity of subsequent crop roots.
Work to date has demonstrated that brassica green manures can significantly reduce bacterial wilt of following vegetable crops when managed to maximise the biofumigation impacts of ITCs, and we have developed a ‘best-bet’ approach to achieve this outcome. Other mechanisms of suppression also arise from incorporation of green material irrespective of its origin, and this may explain the lower and more variable results observed where incorporation techniques do not maximise ITC release.
Root-knot nematodes: Generally incorporation of all green manures including non-brassica controls reduced nematode numbers (by 47-85 per cent) compared with increases in nematodes in untreated controls (12.6 - 55.7 per cent). In most cases the suppression was similar for brassica or non-brassica treatments although the effectiveness of radish, which has been observed in previous years, was again apparent at some sites. The laboratory experiments confirmed that all components of macerated green manures (volatile, water soluble and organic matter) could cause significant independent suppression of nematode numbers. Overall the results indicate that selection of green manures which are poor hosts of nematodes (e.g. among the brassicas, radish and broccoli) combined with tissues suppressive upon incorporation (radish and mustard) is desirable, but that significant suppression in the field was observed for several non-brassica species.
The project has developed ‘best-bet’ strategies for the use of biofumigant green manures as a component of control strategies for these pest organisms, however further evaluation in commercial paddocks and farmer fields is required to establish robust and consistent approaches compatible with various production systems.

All of the planned objectives for this extension period were achieved with few exceptions. The major activities during this 6 month extension were (1) to publish key results arising from the project to date, (2) conduct an on-farm demonstration experiment in North Queensland to evaluate biofumigation on a commercial farm and showcase the “best-bet” biofumigation strategy to key Industry stakeholders (3) develop the forward plan for the Philippine work to develop, test, and extend practical biofumigation strategies in The Philippines and (4) prepare for the Australian experiments in 2005 aimed at separating the ITC impacts of biofumigants from the general suppression arising from organic matter. Objectives (2) - (4) were completed according to plans, and one of the planned publications in (1) has been submitted to an international scientific journal.

The most notable outcome was the poor control of bacterial wilt in cherry tomatoes observed in the “best-bet” demonstration experiment on Endeavour Farm near Mareeba which was in contrast to the excellent results reported using the same strategy on tomatoes at Southedge Research Station in 2003. Although there were some possible issues due to the size of the tomato plants at transplanting (too old) which may have made them particularly sensitive to bacterial wilt during a very hot period in early October, a more plausible explanation was the higher clay and organic matter content of the soil at Endeavour Farm compared with the very sandy soil at the Southedge Research Station. Recent published data from CSIRO highlights the impact of clay and organic matter on increasing the ITC sorption in soils so that much higher levels are required for pest control. This important practical outcome has been followed up in subsequent experiments on the two soils (Southedge and Endeavour) in the CSIRO Canberra laboratory and the hypothesis appears to be valid. This is an important further refinement to the practical “best-bet” advice on biofumigation as it indicates that biofumigation (i.e. the ITC impact) of brassicas may be most effective on lighter textured soils, and may not be successful on heavier soils. The result has delayed one of the planned publications of the laboratory screening methodology for bacterial wilt, as we considered it prudent to check it on a greater range of soil types prior to publication.

The plans for the 2005 work in both countries were finalised and at the time of writing the new extension proposal (January 2005 - June 2006) had been submitted and accepted by ACIAR and work was well under way.

All of the planned objectives for this extension period have been achieved with few exceptions.

The major focus of the extension was in the Philippines, where practical farm-based biofumigation technologies were evaluated, developed and tested. The approach was to focus on potato production systems in Benguet and Mindanao and combine farmer surveys and an Industry workshop, with replicated experiments on research farms and participatory on-farm research and extension. Two phases of experiments supported by laboratory studies were completed (Phase 2 still being finalised at time of writing). The results from replicated experiments support previous observations regarding significant suppression of bacterial wilt using biofumigation (50-80%), and associated increases in yield although variability in the response remains an issue. A mid-year Industry Workshop was attended by 40 advisors, growers and farm trainers including FAO staff from Thailand, Cambodia, Laos, Vietnam and Indonesia where opportunities and constraints to adoption of biofumigation were discussed. On-farm participatory experiments and farmer field schools have been carried out at 2 sites in Mindanao, and in Benguet, and in some cases the farmers have already integrated biofumigation principles into their vegetable production systems. A Biofumigation manual is being prepared on the basis of work on the project, and an FAO booklet has already been compiled and distributed which contains many outcomes from this project. An engineer based at UPLB has completed 2 preliminary student projects on identifying and modifying small-scale incorporation equipment suitable for Filipino farmers and a machine is now in the process of modification and testing. Dr Kirkegaard visited the group in 23-27 January 2006 and the project is on track to achieve the milestones as planned.

Supporting experiments at NCPC during this phase have demonstrated the importance of soil type on the effectiveness of biofumigation (less effective on heavier textured soils), demonstrated that significant suppression can be achieved without complete tissue maceration, demonstrated that 10% ethanol added during the preparation of “brews” increases their effectiveness (unrelated to direct effects of ethanol), and that organic matter additions do increase the population of antagonistic organisms, particularly on heavy textured soils. These experiments have all provided important insights to assist in design of practical approaches to biofumigation.

In Australia the focus was on laboratory (CSIRO Canberra) and field studies (QDPI Mareeba) to evaluate the relative contribution of ITCs and organic matter to the suppressive capacity of brassicas. Isolines of mustard with varying concentrations of ITC were used in the studies and these were compared with non-brassica tissues and synthetic fumigants. The Canberra studies confirmed a rapid impact (days) of brassicas related to ITCs and a slower suppression (weeks) associated with organic matter addition. The rapid ITC effect was shown to be highly dependent on soil type with much more effective suppression on light textured soils with lower organic matter. We demonstrated that the poorer effect of biofumigation observed at Endeavour farm in 2004 compared to the excellent results at Southedge in 2003 (and 2005), may have been partly due to these soils differences.

The Southedge experiment in 2005 has demonstrated that a mustard high in ITC was more effective in reducing bacterial wilt (9.8% infection) and increasing yield (55 kg/plot) than mustard with no GSL (33% infection, 28 kg/plot), clearly demonstrating a role for ITCs in the biofumigation response on this soil. The impact of soybean and weedy fallow were similar to the low ITC mustard. Interestingly the synthetic fumigant (metham sodium) provided a level of control similar to the low GSL mustard and other green manures. The observations support the hypothesis that both a short-term ITC effect and an organic matter effect operate in disease suppression, and both of these effects are combined in a high GSL green manure. These data support the 2003 Southedge experiment in which most of the green manures provided some level of suppression, but the high GSL mustards provided the most effective control. The lighter textured soil at the site also contributes to the effectiveness of biofumigation at the site. In Australia the focus now will be on compiling and publishing the results of the project and finalising the cost/benefit analysis now that a second year of yield data have confirmed the level of suppression which is achievable.

Project Outcomes

The team demonstrated that significant suppression of both organisms could be achieved using Brassica green manures, and that this arose from a combination of two separate mechanisms. The first mechanism was short-term suppression (over 2-3 days) related to the release of isothiocyanates (ITCs) from the Brassica tissues. To maximise the impact of this mechanism, strategies to increase ITC release from the plant tissues and their residence time in the soil were developed into a ‘best-bet’ approach. This involved:
use of brassica tissues which release high concentrations of toxic ITCs (e.g. mustard);
incorporating around 5 kg/m2 (5% W/W) of fresh tissue into the soil;
adequate maceration of the tissues at a cellular level to release the ITCs;
adequate water to facilitate hydrolysis of ITCs;
rapid and thorough incorporation for complete mixing of ITCs through the soil;
watering or covering to retain the volatile ITCs in the soil;
targeting light-textured soils with low organic matter to reduce inactivation of ITCs.

This approach could be readily adopted in highly mechanised Australian farming systems and was very effective on light-textured soils in north Queensland. Field experiments in Mareeba in 2003 and 2005 showed mustard green manures reduced bacterial wilt incidence in tomatoes from 80 to 15%, with a 10-fold increase in yield (from 2.5 to 20 t/ha). On loam and clay soils with more organic matter in Australia and in the Philippines there was less evidence that ITC-related suppression was occurring, and it was more difficult to achieve items 1 to 7 (above) in the context of small-scale Asian farming systems.

However a second mechanism of suppression operating over a longer period was also identified.
This mechanism, occurring over a period of 3 to 4 weeks, was related to the effects on the two pest organisms of added organic amendments. This mechanism was observed on soils with higher clay and organic-matter and did not require the strict protocols listed above. Reductions of bacterial wilt incidence in potatoes in the Philippines (49 down to 17%) and associated yield increases (9.1 to 18 t/ha) in eight on-farm experiments were largely attributable to this process, as was the significant suppression of nematodes observed (50 to 80%) irrespective of the type of green manure used.

The scientists confirmed that the nitrogenous components of organic amendments could generate the same pattern of suppression as the green manures when added in pure forms to soil, and evidence suggests it is related to the mobilisation of specific antagonistic soil bacterial populations which are yet to be identified. More work is needed to develop strategies to fully exploit this more general mechanism of suppression.

Biofumigant green manures should not host the pest organisms of interest and although most of the brassicas tested were moderate hosts of root knot nematodes (Meloidogyne spp.) providing potential for population increases, radish (Raphinus sativus) was found relatively resistant to nematodes. Radish is cheap, readily available, has large seed which can establish well and grow rapidly, also radish leaves represent a waste product that farmers can utilise. It offers the most promising option as a biofumigant green manure where nematodes levels are high.
Brassica species are generally considered non-hosts of bacterial wilt and although the team observed one incidence of bacterial wilt infection in a mustard line from Australia in a nursery in the Philippines, subsequent investigations in Taiwan suggest root damage is a prerequisite. However continual vigilance to ensure biofumigants do not become susceptible to bacterial wilt is warranted.

The ‘best-bet’ biofumigation approaches were presented and discussed with farmers, farm trainers and extension specialists in both countries in participatory on-farm learning activities, field days and workshops. The team developed strategies to integrate biofumigation as a component of integrated disease management in both countries.

Project ID
Project Country
Inactive project countries
Commissioned Organisation
CSIRO Plant Industry, Australia
Project Leader
Dr John Kirkegaard
02 6246 5080
02 6246 5399
Collaborating Institutions
Benguet State University, Philippines
University of the Philippines at Los Banos, Philippines
Queensland Department of Primary Industries and Fisheries, Australia
Project Budget
Start Date
Finish Date
Extension Start Date
Extension Finish Date
ACIAR Research Program Manager
Dr Gamini Keerthisinghe