Australian Plants online
Index   Back Issues   ANPSA Home

Mundulla Yellows - The Cause Remains Controversial

Jim Barrow

"Mundulla Yellows" is the name given to a disease that affects at least 87 species and 29 genera of plants. However it is most commonly observed in eucalypts. It was first reported near Mundulla, South Australia in the 1970s, but has since been observed widely in Australia and also overseas. The first sign is that the area between the leaf veins goes yellow - an interveinal chlorosis. This begins on the immature leaves. The chlorosis may spread to more mature leaves and these may show necrotic spots. Branches may die back. Especially on eucalypts, epicormic growth may then occur, but the new growth is also chlorotic and dies. Symptoms mainly occur in disturbed vegetation, particularly near roadsides and paths. It does not affect all the trees in an area but typically one tree, or a group of trees, in an otherwise healthy stand. Three "causes" have been proposed.

  1. It is caused by soil properties, especially alkalinity which induces iron deficiency.
  2. It is caused by an infectious biotic agent.
  3. It is a simple case of poisoning by residues from triazine herbicides such as simazine.

I want to discuss how all well these three explanations relate to the observed occurrence and how well they are supported by the evidence put forward by the proponents.

1: Soil properties

There are three strands in the work that led to proposal that it is caused by soil properties. In one strand, (summarised here) the Victorian scientists who put forward this proposal showed that no infectious biotic agent was involved - and therefore argued it had to be something else. In the second strand, they chose five sites on which the trees were affected and five sites in which they were not. The sites were all well separated. This sampling strategy increases the probability that there will be differences in soil properties between the sites. The soil from the affected sites all had a higher pH, higher electrical conductivity and lower available iron. Aqueous extracts of soil from affected sites had high levels of carbonate and bicarbonate. From these observations they concluded that symptoms were influenced by soil properties such as increased alkalinity and salinity plus the accumulation of carbonate and bicarbonate in the soil solution. This evidence is, in itself, not convincing. There are many examples of trees growing on alkaline soils, and therefore being subjected to all of the supposedly adverse properties, but showing no symptoms. Nor is it consistent with the reported occurrence of the disease in just a few trees in a stand even though all the trees are apparently growing in similar soil. However, in the other strand of their work, they showed that plants showing severe symptoms did respond to application of iron and this has to be accepted as convincing evidence that iron is somehow involved.

2: Not soil properties but rather a biotic agent

In contrast, a South Australian group sampled paired sites, that is, sites in which affected and normal trees were growing fairly close to each other. The results show that Mundulla yellows can occur in acid soils and provide no evidence of an association between occurrence and pH.

Soil pH under paired Mundulla Yellows-affected (MY) and normal eucalypts
in the South-West of Western Australia
Species Site Soil type pH at 10 cm pH at 30 cm
MY Normal MY Normal
E. camaldulensis Northern Sand Plains Sand, limestone 7.47 7.48 7.45 7.51
E. camaldulensis Swan Coastal Plain Yellow sand 6.68 6.94 6.77 7.04
E. marginata Swan Coastal Plain Yellow sand 6.40 6.07 6.28 5.90
E. marginata Darling Range foothills Sandy loam 4.72 4.44 4.29 4.29
E. marginata Jarrah forest Lateritic gravelly sand 4.73 4.70 5.20 5.20
E. salmonophloia Wheatbelt Sandy loam 6.64 6.92 7.19 7.09
E. salubris Wheatbelt Sandy loam, granite 4.77 n.a. A 4.80 n.a. A
E. camaldulensis Wheatbelt Sandy loam 5.41 5.20 5.20 5.00
E. rudis Blackwood Valley Red loam 5.30 5.40 5.20 5.00
A: No data available.     From Hanold D. and others, Australasian Plant Pathology , 2006, 35 , 199-215  

So where does this leave us?

On the one hand, we have in Victoria one group (largely from the Department of Primary Industry) who are convinced that the cause is not a biological pathogen but rather that the disease is caused by differences in soil properties and especially by alkaline soils.

On the other hand, we have in South Australia another group (largely from the University of Adelaide) who are convinced (and who provide good evidence) that the cause is not differences in soil properties. They think that "The symptoms, incidence, distribution and pattern of spread of MY are consistent with the disease having a biotic cause, yet, as reported here, no biotic agent has been found for MY." So they hold to their biological pathogen idea despite the lack of evidence. They dismiss the idea that the disease could be caused by residual herbicides because sites have "a recorded history of no herbicide sprays for at least .....20 years on roadsides".

I have two problems with this conclusion. Firstly, since herbicides can be applied in a matter of minutes, I do not understand how it is possible to be sure that none was applied over a 20 year interval. Secondly, I describe in the next section trees that have been affected by herbicides including two cases where the herbicide was applied more than five years ago, so an extension of the effect, even up to 20 years, does not seem infeasible.

3: Herbicide residues

This explanation was originally proposed by Jim McNamara from observations in South Australia and I later showed that is this explanation seemed to fit some observations in Western Australia. More recently, we showed that application of the herbicide hexazinone had induced Mundulla-like symptoms in a range of trees. Hexazinone is closely related to the triazine herbicides. It is a triazinone. I now intend to present further observations providing further evidence that herbicides can induce Mundulla yellows and that the symptoms can last at least five years.

Case one . In June 2006, I observed that couch grass (Cynodon dactylon (L.) Pers.) in a garden bed at a school in the suburb of Claremont had been sprayed with a herbicide. Attempts to identify the ingredients of the herbicide have been unsuccessful because the grounds supervisor has left.

Photo of sprayed area   Chlorotic Lilly-pilly Photo   Chlorotic Lilly-pilly Photo
Left: By June 2007, when I again inspected the site, the tree closest to the sprayed area had died and had been removed. Note the bare ground near the base of the tree.
Centre: The affected area with a chlorotic Lilly-pilly (Syzigium sp) shrub in the foreground, in front of unaffected cyprus trees and an affected eucalypt (see below).
Right: Close up of the lilly-pilly leaves showing interveinal chlorosis plus necrotic spotting.


Dieback on Eucalyptus robusta   Dieback on Eucalyptus robusta
Typical Mundulla yellows symptoms shown on a Eucalyptus robusta tree about 10 m from the site of the herbicide application
Left: Chlorosis and die-back of the tips. Right: Chlorotic epicormic growth.

These observations are typical of many Mundulla yellows experiences. The symptoms are characteristic and affected trees are confined to a small area. By the time the symptoms are observed, the evidence of the sprayed grass has disappeared and records of the spraying are impossible to obtain. If I had not noticed the sprayed grass, this would be a another unexplained occurrence.

Case two. In June 2002 I noticed that many trees in the suburb of Mount Claremont, Western Australia, were showing symptoms consistent with Mundulla Yellows. All of these trees were close to areas that had been sprayed with herbicide. The firm responsible for turf maintenance denied having used any herbicide except glyphosate, but this was impossible to check as the local government authority had changed to a different contractor. Nevertheless, the local government authority issued strict instructions to the new contractor that in future simazine was not to be used and the problems ceased to occur on newly planted trees. Many of the trees that were showing symptoms in 2002 were judged to be unsightly and have been removed. However, some remain and symptoms are still apparent.

Spotted Gum Photo   Spotted Gum leaves
Left: Spotted gum, Eucalyptus maculata (syn. Corymbia maculata), in Cleland St. Mount Claremont W.A. showing symptoms consistent with Mundulla Yellows. Only the nearest 2 trees in the row of 38 trees show symptoms. A small area of grass between slabs in a concrete path 15 m the left of the first tree was sprayed with herbicide in winter 2002 (Photo June 2007).
Right: 1 - Leaves from the first tree; 2 - Leaves from the near side of the second tree; 3 - Leaves from the far side of the second tree; 4 - Leaves from the third tree.

This occurrence in a few trees in an otherwise unaffected stand is reported to be typical of Mundulla Yellows. It is unlikely to be associated with differences in soil properties as the soil appears to be uniform along the stand. Nor is the lack of spread over a five year interval consistent with a biological pathogen. It is more likely to be associated with the observed application of a herbicide in 2002 to a small area about 15 m to the left of the first tree. The severity of the symptoms has changed little in the intervening five years.

Case three. In June 2002 I noticed a chlorotic sheoak tree with recently sprayed grass beneath it and (fortunately) photographed it. It is likely that the chlorosis was caused by the use of herbicide and, given its common availability and the nature of the symptoms, that the herbicide contained simazine. Again, the severity of the symptoms has changed little in the intervening five years. Indeed rather than dieing, the tree has grown somewhat. I think this survival is also because there has been no re-application of herbicide. Some dwarf eucalypts planted in 2004 about 2 m from the base of a sheoak tree do not show any symptoms of chlorosis. This suggests that carryover of the herbicide has been within the plant rather than within the soil.

Sheoak Photo   Sheoak Photo   Sheoak Photo
Left: A young sheoak tree ( Allocasuarina fraseriana (Miq.) L.A.S.Johnson) near Mount Claremont library (Photo June 2002). Note the dead grass beneath the tree.
Centre: The same tree photographed in April 2007
Right: A unaffected sheoak about 30 m from the other tree in April 2007

Consider now the three positions of that might be held.

  1. Mundulla yellows can be caused by induced nutrient imbalance and especially deficiencies of iron and of manganese on alkaline soils. Although there was some initial evidence for this explanation, it is not consistent with many aspects of the occurrence of the disease and when it was specifically tested, as above, it could not be sustained. Of course, nutrient imbalances on alkaline soils can produce chlorosis but these are not Mundulla yellows.

  2. Mundulla yellows can be caused by a biological pathogen. Despite large amounts of work, no such pathogen has ever been identified. Further, when this explanation has been specifically tested, the evidence is strongly against it.

  3. Mundulla yellows can be caused by herbicides. There is strong evidence that this is so both in the cases cited here and earlier. These cases illustrate the general difficulty of associating symptoms with herbicide use. If a herbicide has been used only once, the killed plants (usually grasses) have disappeared within a few months. Yet the symptoms in the trees may persist for at least five years. An observer would therefore find it difficult to make the connection. Further, it is often difficult to obtain records of the use of herbicides, and not uncommon, as was the case for the trees in Mount Claremont, for the use of triazine herbicides to be denied.

    Yet, all of the evidence is based on observations made after the fact. It is high time for a controlled experiment in which rates of herbicide are tested.

    A difficulty with the herbicide explanation is the evidence that symptoms can be cured by the application of nutrients especially iron. This is puzzling because we would expect that residual herbicide to destroy chlorophyll, whereas shortage of nutrients means that chlorophyll cannot be made. That is, the mechanisms of effect are different. This is one of the factors that should be tested in a controlled experiment.


Index   Back Issues   ANPSA Home

Australian Plants online - 2008
Association of Societies for Growing Australian Plants