In 2010, Craig Allen and colleagues published ‘the first global assessment of recent tree mortality attributed to drought and heat stress’ (Allen et al. 2010). In this fantastic paper, the authors collated examples of tree die-off (or mortality) from around the world and – in a very long sentence – they concluded:
… studies compiled here suggest that at least some of the world’s forested ecosystems already may be responding to climate change, and raise concern that forests may become increasingly vulnerable to higher background tree mortality rates and die-off in response to future warming and drought, even in environments that are not normally considered water-limited.
Given that pessimistic prognosis, it’s worth asking; how are trees faring here, in our own backyard?
In July, I invited readers to send in observations of how their local vegetation changed after the Big Wet years of 2010-2011. I expected to receive lots of comments about changes caused by the heavy rains, which I did (see recent blog posts: here, here and here). Unexpectedly, many readers also submitted observations of trees that died after the Big Wet had passed. What caused these deaths? Let’s see, by first reviewing all of the relevant observations.
Dying trees
Most reports of dieback were from dry hill slopes, and occurred last summer. In north-east Victoria, one reader noted, ‘conspicuous bands on north and west aspects where major die off occurred over the summer of 2012/13 … mostly Stringybark and Red Box. There are a range of very obvious bands which showed following the heat, so … this may have been due to the sustained wet, followed by the major dry period.’ The same reader had also seen, ‘many old isolated paddock trees die off rapidly this summer … right across the region.’ Others made similar observations: ‘there has been considerable, albeit patchy, dieback in the Strathbogies, most notably amongst stands of Messmate Stringybark (E. obliqua) on shallower soils and sheet rock.’ Further west, in central Victoria, a reader noted:
‘On western slopes abutting Mount Tarrengower in nearby Maldon there has been considerable dieback of mature eucalypts and wattles, including both natural regeneration and planted stock that had thrived under the wet conditions of 2010-11. This notably includes hardy species such as Grey Box, Drooping Sheoak and less hardy Black and Golden Wattles. It appears that successive dryness over spring-summer-autumn 2012-13 has notably impacted native vegetation in grassy woodland and dry forest vegetation types throughout Victoria, from casual observations. Following minor winter respite some plants have ‘hung on’ and are greening up despite the prevalence of dead plants.’
Trees didn’t only die in the hills: ‘across western Victoria there has been considerable dieback of Banksia marginata stands, many of truly veteran status…. the phenomena appears to be widespread and apparently associated with dry conditions.’
Hungry bugs
In their global review, Craig Allen and colleagues found that, ‘tree mortality commonly involves multiple, interacting factors, ranging from particular sequences of climate stress and stand life histories to insect pests and diseases.’ Notably, many readers found that dieback was exacerbated by hungry insects.
‘In the Daylesford area… the most noticeable change after the big wet was an explosion of cup moth caterpillars. We are having another explosion of cup moth caterpillars currently and the canopies of Messmates, Long-leaved Box and Candlebarks are looking pretty bad. Last spring the regrowth of the canopies was beautiful. But I fear the long dry summer and autumn recently will stress and kill some of those affected.’
Another reader wrote:
‘I live in Box-Ironbark country in Taradale…. The forest suffered from a massive cup-moth infestation and consequent leaf loss last spring, during the driest 6 months – i.e. lowest moisture levels on record, according to the DSE/DEPI, worse than any time during the 12 year drought that ended in 2010. The stringybarks are now looking very sick, many covered with brown leaves or more advanced stages of dying. It’s looking a lot worse than in the 1982 drought. I wonder what future there is for stringybarks in this forest.’
Other insects were abundant in other regions. Catherine wrote, ‘I have just finished my honours thesis looking at Eucalyptus viminalis dieback in the Monaro region, which appears to be the result of drought along with an infestation of the eucalyptus weevil (Gonipterus sp.). However there doesn’t seem to have been any recovery in the last few years despite the comparatively wet conditions.’ And closer to the coast…
‘… we have had a massive dieback event of E. moluccana (Grey Box) in Western Sydney during the last few years. The dieback is due to psyllid attack and has been specific to the one species of eucalypt. This is a dominant tree in threatened Cumberland Plain Woodland and has been the subject of several dieback psyllid attacks since the 1990′s. Opinion seems to be that it is related to the extreme dry/wet climate events in recent years and that the recent event is the worst on record. … This is the most significant event in our region in the recent wet years I think.’
Rainfall patterns
Many of the comments describe unusually dry weather last summer. To see how dry it was, I downloaded the Bureau of Meteorology data for Malmsbury Reservoir in central Victoria, not far from Taradale and Daylesford.
To make the patterns easy to see, I added up the rainfall for each quarter of each year (summer = Dec to Feb, autumn = March to May, etc). I then subtracted the actual rainfall that fell in each quarter from the expected quarterly rainfall, based on the long-term average. This allowed me to create a chart that shows the deviation from the expected (or average) amount of rainfall in each quarter. In the chart below, the bars point down if the quarterly rainfall was less than average, and up if the quarterly rainfall was above average. For example, in summer 2011, Malmsbury received 289 mm more rain than would fall in an average summer. The black bars show each summer (Dec-Feb) so you can clearly see the start of each year.
When compared to the average rainfall in this way, the 2012-2013 dry spell was one of the driest 9-month periods in the last 30 years. At Malmsbury, similar dry spells occurred in 1982, 1994, 2006 and 2009, as can be seen on the chart above.
The future?
How will the forests change in the future? Unfortunately, we have very few long-term studies of how alternating dry and wet periods affect Australian ecosystems. In one long-term study, the late, great David Ashton found that dry rocky slopes on Mt Towrong in the Macedon Ranges were hit by six droughts in 25 years. The vegetation waxed and waned during the dry and wet periods. On shallow soils, the most drought-sensitive species, Messmate Stringybark (Eucalyptus obliqua), died back in each drought and then re-colonised in the next wet period, then died off and re-colonized again.
Many of the observations above mention dieback of stringybark species. The fate of the stringybarks will depend on how well they recover in future wet periods. As climate change intensifies, we can expect drought-sensitive species like these to decline and be replaced by more drought-tolerant species. Perhaps that’s already occurring in some places.
Unfortunately, we have no long-term monitoring system to document how forests across south-east Australia change over time. That’s why it’s important to collate observations like these. To make these observations really useful, we need to record how the forests continue to change. We often notice short, sharp changes, especially deaths, but tend to overlook slow, gradual changes between the periods of dieback. Will dying trees recover or will they die off completely? Will new plants re-colonize the dry slopes in wet years? If so, which species, the same ones or new species?
I wonder how the dieback forests have fared after this winter’s good rains. Have they recovered? Once again, the only way we’ll know is if lots of people record and share their observations. Thanks once again to everyone who wrote in earlier, and to Anna Foley for her great Banksia photo. If you have more observations of dieback or recovery in your area, please leave a comment below.
Further reading
Allen CD, Macalady AK, et al. (2010). A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management 259(4), 660-684. [You can download a full text copy of the article for free from this link].
Ashton DH & Spalding DK (2001) The ecology of a stressful site: Mount Towrong, Central Victoria 1967-1997. Australian Forestry 64, 143-150. [You can download a full text copy of the article for free from this link].
Ian Lunt Ecology 
Very worrying story.
I hope Cris Brack from ANU and his team’s work on E. triandra and C. maculata at the National Arboretum Canberra will be able to shed more light on this as well.
Hello Max, thanks for commenting. I’ll look up Cris’s papers and will see if I can work them into a future blog. A lot of research is being done in SW WA too, where there has been very extreme dieback due to a big, long-term decline in annual rainfall. Best wishes Ian
Hi Ian
I was working in WA in the latter half of the 1970s, and documented the death of shrubs due to drought in the Eneabba region (1981,Hnatiuk, R.J. & Hopkins,A.J.M. Western Australian species – rich kwongan (sclerophyllous shrubland) affected by drought. Aust.J.Botany 28: 573-585). The long drying had just started and was having a significant impact on the species-rich shrublands there. I don’t know how they have fared in the following 30+ years as moisture has declined since then. In the long term, we know that the SW dried out greatly several thousand years ago, with wind-blown dunes right to the south coast. The species rich shrublands must have contracted to refugia before spreading out again. What will happen this time, with most of the land under cultivation will be a rather more complex ecological system.
Roger Hnatiuk
Canberra
Hello Roger, thank you very much for writing in. I must have read your paper long ago, and shall look it up and read it again. The drying out of SW WA is much worse than has happened in SE Australia, and perhaps provides a portent of the future for many areas. ‘We ain’t seen nothing yet’ I guess. Thanks again, best wishes Ian
Hello Ian, I found your post very interesting, as it is a phenomenon that is occurring at such a large scale.
We have been studying the episode of Eucalyptus moluccana dieback in Western Sydney and found that dieback was more severe at sites dominated by Noisy Miners, and less severe at sites with small insectivorous birds. Other studies show that Noisy Miners exclude smaller birds from forest/woodland through competitive aggression. So putting these together, it appears that the top-down effect of depredation of insect pests by small insectivorous birds has played a role in maintaining tree health at times of insect attack. What drove the irruption of the insects (psyllids) in the first place is not something we looked at, but is likely to be related to climate based on the evidence you and others have presented.
Hello Michelle, thanks very much for writing in, it’s great to hear of your work and findings. It’ll be fascinating to see how many cases of dieback and insect attack Noisy Miners contribute to, and how many cases act without (or regardless of) Noisy Miners. I guess it’ll always be important to strive to maintain healthy populations of small insectivorous birds to help minimize insect damage. Thanks again, best wishes Ian
HI Ian, We had another example of drought induced dieback in Heathy Woodland Vegetation around the Yarram area. during the dry years in the noughties. There was widespread death of mature Saw Banksias (Banksia serrata) which is probably the dominant canopy species in this vegetation type locally. Not sure how widespread across Sth Gippsland this dieback occurred i.e. whether or not it happened in similar vegetation at the Holey Plains SP or in Wilsons Prom. All in all probably not a good result for Pygmy Possums and other nectar feeders.
In terms of insects Fireblight Beetle is having a dramatic effect on vegetation with thick regrowth of Black Wattle from the 2009 fires being seasonally completely defoliated. See this post for more info http://wp.me/p26y09-fF It is also having an impact on Wattles on sites established by direct seeding.
Hello TB, thanks for sharing your observations. I hadn’t heard about the Saw Banksia dieback. I wonder (like you) if it was more widespread? I hope not, as the old Saw Banksias provide lots of food and nectar for birds and animals. Thanks for the link to the post on Fireblight Beetles. I’d encourage other readers to follow the link and read the posts on FB beetles as they are really interesting. Keep up your great work in Gippsland. Thanks again Ian
Hi Ian, Interesting article and subsequent discussion and comments. The observations made by yourself and others have been mirrored here in central Victoria around Bendigo. The effects on both trees and the shrub and herb layers have been in some localities quite marked. Interestingly around my locality the majority of sudden deaths occurred after several days of extreme temperatures suggesting as noted by yourself and others that there are a number of factors at play. One of these that doesnt seem to have been given too much attention is the current state of our natural remnants. These are not necessarily that natural with over 85% subjected to at least one if not multiple clearing events over the years since Europeans arrived. This tends to result in remnants with much greater stands of small and dense trees than would be “natural”. Because trees like stringybarks tend to occupy the upper slopes and ridges these are probably particularly vulnerable to challenging climatic patterns. While the deaths of some trees seem negative it is also possible that the native systems are to some extent re-aligning in response to these earlier clearing events. As you say these systems are complex. What is of most concern is that some understory elements that are long lived have sufferred dieback that I believe is not at all usual. Hope the comments I have made are legible as I have not been able to see most of what I have typed due to issues with my browser. Regards Steve Williams
Hi Steve, thanks very much for your comment. We don’t know much at all about how past disturbances affect dieback or other dynamics, and your point is a good one. I’d imagine that disturbed sites would be most prone to dieback if the earlier disturbances had altered the soils and reduced their capacity to hold water, and this could accentuate dieback of dense regrowth stands too. It shall be interesting to see how these stands change in future years. Thanks again and best wishes Ian
Hi Ian, While there are usually significant impacts on litter and mulm layers associated with clearing the most important factor is that instead of the 200 odd tree stems per hectare (pre-european) we now have 1000+ stems in many of our remnants. The competition for water and other resources must be significantly higher. If the 200 stem represents a more natural system then we should be expecting significant dieback as these forest stands age. I suspect the exact processes are not well understood. Regards Steve
Hi again Steve, yes, you make a good point. Sorry for my delay in replying. We can also get high stem densities after natural disturbances like fires. Droughts may be the period when dense regeneration most rapidly thins out, due to the severe restrictions on water availability. It will be fascinating to see how the dense stands thin out in the future. Some dense stands have persisted through severe droughts in the past and, while thinning out, remain dense. See for example an earlier post on dense trees: https://ianluntecology.com/2013/05/22/dense-old-trees-bitter-and-twisted-charismatic-megaflora/ Thanks again and best wishes, Ian
Hi Ian, Love your blog. There is also this more recent paper on forests around the globe, heat, drought and hydraulic embolisms published in Nature: http://www.nature.com/nature/journal/v491/n7426/full/nature11688.html. Worrying stuff, as is this local study from Qld on Eucalypts, mentioned on the ABC today: http://www.abc.net.au/news/2014-01-13/eucalypts-suffering-from-changing-climate-study/5197684.
I looked at many of the remaining Banksia marginata (tree form) in May 2013. It wasn’t pretty. Most were in severe decline or dead – including both old and planted trees. At Mooramong, I looked at 19 planted riparian trees that are likely to be over 10 years old – 16 were dead, most apparently failing in the previous summer. In other stony rises at Dundonnel, virtually all replantings had failed (3 left from 35 trees originally planted in 88/89, with 6 recent deaths). Isolated old trees were dead or near dead at Mooramong, Skipton, Dundonell and Lake Wongan. Given the fate of the majority old trees and much of the revegetation, I reckon retaining the genetic diversity in the remaining trees will require some dedicated effort including ex-situ species management (via seed orcharding, preferably irrigated with attendant pollination vectors) and consideration may need to be given to introducing genetic material from further afield. Use of seed from domestic planted trees (from the old trees) will also be desirable. Broadening the sites of revegetation perhaps include cooler, wetter sites on the plains should be considered.
The other species that took an obvious hit after the big wet for me was Acacia melanoxylon. Many, many dead young trees in roadside reveg across the state. It is possible that deaths of some Banksia marginata and Acacia melanoxylon are attributable to the fact that many were planted stock (which never performs as well as sown plants) and planting was poorly matched to the site. Young Banksias may also be inbred /genetically weak because they were grown from a very narrow parent population. At the same time, the old trees were arguably senescent anyway (nothing lives forever). Interesting and important research questions for a budding researcher, and I hope someone can take on the challenge.
Regards, Scott
Hi Scott, thanks very much for the links to other articles, and for your observations. The decline of old Banksias in particular is very worrying. At least the Acacia species have soil seed banks and may have a chance of regenerating in later years. By contrast once the Banksias die, there is no soil seed bank for them to regenerate from. A number of readers provided information on their activities to propagate tree-form Banksia marginata in another blog post that was devoted to Banksias – https://ianluntecology.com/2013/10/13/forgotten-woodlands-future-landscapes/ If you haven’t seen it, you may find it interesting. Thanks again Ian
Hi Ian
Another example of tree canopy dieback presumably related to drought can be found in reported in the regional press here http://www.portnews.com.au/story/1906796/mountains-going-a-strange-shade-of-patchwork/ . The forest at Middle Brother Mountain is opposite the Pacific Hwy in northern NSW so it resulted in plenty of inquiries to the local NPWS office. I suggested to them the very dry spring period where temperatures and rainfall normally increase in parallel may be of significance, in late 2013 the usual spring rainfall did not appear as temperatures increased and swathes of E. pilularis mainly on dry ridge slopes exhibited dieback. Further north I monitored canopy dieback in Nothofagus rainforest (it looked like frost damage) with around 30% of the annual spring growth lost due to drought. A very unusual occurrence for an environment with an MAR of 2200 mm and elevation of ~1200 m. Two months later I observed the first wildfires in northern NSW Nothofagus rainforest since 1956. My group already had a network of long term rainforest plots in place and we plan to report on the post-fire changes this year. While this area of northern NSW represents a very different climate scenario to the central Victorian woodlands, the rainfall deviation is probably of equal significance considering rainfall is concentrated in the summer months and RH [relative humidity] is normally >80%. Thanks Ross
Hello Ross, thanks very much for your observations and the link to the newspaper article. That’s a lot further north than any of the observations from other readers. I hope you get better falls of rain this year. Best wishes Ian