Habitat loss.....the signature conservation problem of the 21st century

Extinction from habitat loss has been suggested to be the signature conservation problem of the twenty-first century. Indeed, the Millennium Ecosystem Assessment predicts that near-term extinction rates could be 1,000 to 10,000 times higher than background rates. This alarming statistic seems to have got lost in the current focus on climate change, important as that will likely prove to be.

Images from Google Earth highlight just how much
 woodland clearing has occurred 
in western Victoria for cropping and grazing

One key reason species will go locally extinct due to habitat loss is because fragmentation leads to small populations. A central tenet of conservation biology is that small population size makes them much more vulnerable to inbreeding, recruitment failure, and climate extremes.

In our Lab, we've been testing the idea that small populations are more vulnerable to local extinction than large populations. Despite the simplicity of the hypothesis, there is not a lot of good evidence around in the literature to support or reject this idea (at least for plants in Australian ecosystems).

To address this question, we've employed re-visitation studies. This involves using data on species abundance collected at some previous point in time from well-defined places (lets call this the historical dataset, time1), and revisiting the very same places some time later to determine whether species have persisted (lets call this the contemporary dataset, time2). If you have data on abundance at t1, you can start to understand whether the smallest populations at this time are less likely to be present at t2 (as theory would suggest).

We used data collected in 1975 by a superb amateur botanist called Cliff Beauglehole as our t1 for a number of remnant grasslands and grassy woodlands in western Victoria.

Scaly Buttons is one of the common species
in grasslands and woodlands that has declined
over the last three decades
(Photo: John Morgan)

 Beuglehole recorded species lists for these sites (many of which were very diverse at the time) and allocated an 'abundance' measure for each species at each site that approximates a log scale - a few dozen, up to 100, in the 100s, in the 1000s.

We re-visited these same sites (t2) and recorded all species we could find; importantly, we estimated their population size in the same way as Beuglehole. When comparing the data on abundance at t1 versus t2, remembering that small populations should be more vulnerable to local extinction, we found some very interesting results which we have reported in the Journal of Ecology.

As predicted, the liklihood of local extinction over a  31 yr period was highest (34%) for those species whose population was initially very small. But initially small populations did not always decline - indeed, a small number (9%) actually increased and became abundant (in the 100s) or even very abundant (in the 1000s).

Changes in population abundances of native plants species
from 1975 to 2006 in grassy woodlands in western Victoria
(From: Sutton & Morgan Journal of Ecology)

But the story doesn't end there. We uncovered something in our re-visitation study that we had not predicted, nor even expected.

Some of the plant populations that were initially considered abundant or very abundant in woodland remnants had become locally extinct. Hence, having a large population did not guarantee persistence over the three decade period between observations. There could be a number of reasons for this, but we think the main one relates to habitat quality.

Over the 30 yrs between surveys, it is very likely that the habitat has deteriorated for some species - edge effects, weed invasions (by exotic Iridaceae, a curse in grassy woodlands in western Victoria) and inappropriate disturbance regimes have probably all caused a decline in habitat. In particular, the lack of fire is probably a key driver of change. Fire-sensitive shrubs, as well as dominant tussock grasses, have increased at the expense of poor competitors in the groundlayer.

Our study hints that while there is a strong need to find simple, general rules in conservation ecology, these general rules need to be put to the test. Not all small populations are vulnerable to short-term extinction. Indeed, 'common' species deserve our attention as well, as others have suggested.

And it is the common species that drive ecosystem function and provide much of the habitat for fauna. Simple re-vistation studies like ours are an excellent way to monitor trends in remnant vegetation, provided that old datasets are archived in such a way that they can be re-used in the future.

Abundance at 'home' predicts abundance 'away'

One of the best things about teaming up with collaborators, either in your own university, across different institutions, or even across continents, is that it allows you to probe some of the big questions in ecology.

And one of the biggest (and most important) questions that many of us are trying to answer is: why do plants invade new systems? And which species do so? And what are the traits that underpin this?

Recently, I teamed up with a 36-strong global research team that took aim at the reasons for plant invasions in herbaceous ecosystems. Using data collected as part of the Nutnet Project, a research network looking at the importance of top-down and botton-up processes on grassland diversity, we tried to understand why some invaders are common in new habitats, while others are not. Predicting the success of invading species has always relied on the assumption that these plants are more abundant in their new settings than they are in their native communities because they behave in a special way.

Cirsium vulgare - one of the 26 species that
we had data on for the 'home' and 'away' range
(Photo: Royal Botanic Gardens Sydney)

We are not so sure this assumption is correct.

Indeed, we think we have found something pretty cool. Because our study was a global one, we were lucky enough to have data on plant abundance of 26 species in their 'home' range - this allowed us to see which species are common and rare in their natural habitats. For the same 26 species, we had data when they had invaded new habitats (usually grasslands on new continents), which we called the 'away' range.

We discovered two really important things: 1) increases in species abundance in the new range are, in fact, unusual. So, those species that were uncommon in their 'home' range were often uncommon in their 'away' range. 2) We saw quite strong evidence that species common at 'home' were also common 'away' - this might be a really useful way to predict future invader potential.

Hence, success of a plant in its native range can possibly be used to predict its spread at introduced sites – a criterion which currently is not included in biosecurity screening programs.

We've just published this work in Ecology Letters. Check it out.

Observations as the starting point for understanding the ecology of species and ecosystems

Mountain Ash - the tallest flowering plant in the world;
some trees have been measured  as being up to 150 m tall, but sadly
logging and fire has removed these magnificent forest giants
(Photo: John Morgan)

David Ashton was a plant ecologist who is best known for his observations and experiments on the regeneration ecology of the world's tallest flowering plant (Eucalyptus regnans, Mountain Ash). He dedicated his entire working life (from his PhD studies right up till his death in 2005) to the dynamics of the Ash forests - a period spanning over 50 yrs. Most of what we know about this species, and the community it towered over, is due to Dave's committed study of this ecosystem - you can read a summary of this work in The Big Ash Forest - changes in one lifetime. Rarely has a scientific paper's title reflected the subject matter so well.

Dave was a mentor of mine and I once asked him what he thought made a good ecologist. In his gentle, penetrating way he responded:

"a tape measure, a spade, a good set of eyes, and an inquiring mind".

While the tape measure and spade are still very important components of the ecologists toolbox, and the toolbox might now include things like a GPS, a laptop computer, a digitial camera and dataloggers, etc, I think it is valuable to reflect on the last two points, for without them, all the technology doesn't really have a purpose.

All good ecology starts with observations. And careful ones at that. Having made some observations, it then requires the ecologist with a sharp mind to interpret what they have seen. Indeed, I often think of ecology as involving detective work, putting all the pieces of the puzzle together.

I'll give a recent example to illustrate the point.

I've just returned from the Otway Ranges where Phil Keane, Pete Green, David Cameron and I (along with a bunch of undergraduate students) have been observing the ecology of cool temperate rainforests. CTR are dominated by Nothofagus cunninghamii (Myrtle Beech), a tree with a Gondwanan origin. The genus Nothofagus occurs in New Zealand, in South America and, formerly, in Antarctica when it was ice-free, and provides evidence that the southern continents were once connected before continental drift led them to seperate and move far apart. Today, Nothofagus occurs in the Otways in the deepest of the moist gullies where fire is very rare. This part of the story is well-known.



Myrtle Beech in the Otways - dead trees are clearly visible
(Photo: John Morgan)

 One of the major causes of mortality of Nothofagus is the Myrtle Wilt, an air-borne fungal disease caused by the naturally occurring native pathgenic fungus Chalra australis. The fungus infects trees that have been wounded in the outer bark - perhaps by windstorm and falling branches. Following infection, 'wilting' of the canopy occurs; this is easy to observe because leaves first turn yellow, the tree dies, and then brown leaves shed over the following 18 months. This part of the story is well-known.

Or is it? While observing Nothofagus that were recently infected (evidenced by the yellowing of leaves) or had been infected some time ago (evidenced by brown leaves to completely defoliated canopies), we observed something that throws a spanner into this ecological story.

On one clearly infected tree, we noticed that nearer the base, new shoots with green leaves were growing quite vigorously. This is not supposed to happen. As David put it:

"Myrtle Wilt is a death sentance for a Myrtle Beech tree".

We then made a second and a third observation to the same effect. Hence, this is not an isolated occurrence. Perhaps Myrtle Wilt is not always fatal as it was once thought to be?

At another site in the rainforest, we observed very large trees that had been defoliated some time ago, presumably because of the fungus. There again was vigorous regrowth from the base, casting further uncertainty on the well-known "fact" that Myrtle Wilt always kills Myrtle Beech. Our observations have confused us - we are observing a phenomenon that is not supposed to occur. Perhaps the story is more complicated than was originally thought. Perhaps the disease is not always lethal. Perhaps there are circumstances where the tree can survive infection.

Of course we have a lot more work to do before we can definatively re-write the story here. But it demonstrates two things. By making observations of nature (and these are very simple observations) - do Nothofagus die or regrow after Myrtle Wilt - and thinking about what we are observing (i.e., the implications of these observations in light of current understanding), we have gained some rather profound insight into the ecology of this rainforest. This means sometimes questioning the accepted wisdom on the topic.

And we haven't yet got our technology out of its toolbox.

Secondly, it highlights that long-term observations of natural systems are crucial to understanding them and making correct decisions about their wise use and management. Dave Ashton knew this - he repeatedly tested ideas about the regeneration and succession of Ash forests over a lifetime. For most of us, we don't have a lifetime to answer the pressing questions such as climate change impacts on biodiversity. But we can gain such insights if we cleverly use revisitation studies and historical datasets (such as herbarium records & quadrat data), use chronosequences (space-for-time substitution), and establish permanent plots for monitoring with specific questions in mind.

I look forward to us further unravelling the Myrtle Wilt story, a story that was once thought written. In the meantime, we need to go off and make a few more observations!

Sleeper weeds in the alps



Alpine grasslands on the Snowy Range
 are typically weed-free (Photo: John Morgan)



The Australian Alps were once thought rather resistant to weed invasion because the cold climate would act as a strong selection against many non-native species. It is certainly true that mountain grasslands and herbfields are much less invaded than their lowland counterparts. In the lowland grasslands of south-eastern Australia, a suite of Eurasian grasses that evolved in human-dominated ecosystems have been particularly successful at invading native ecosystems. These grasses have been far less successful at higher altitudes where there has been no agriculture.

Recent studies confirm, however, that many weeds do occupy the treeless mountains in Australia. McDougall et al. 2005 found over 125 species had become naturalised in alpine vegetation and it is highly likely that many species currently found at lower altitudes could become future invaders (Alexander et al. 2011).

It is within this context that managers of alpine vegetation need to make decisions about what weeds to control, where and why. Some emerging weeds (such as Hawkweeds and Willows), as well as some species well-established in alpine areas (Broom), are currently the target of annual, resource-limited control programs in NSW and Victoria. But what about future invaders? Who might they be? Where might they invade? And which ones should be controlled?

Recently, I undertook some monitoring in the Victorian Alps with the specific aim of recording locations of new weed incursions. I have been undertaking weed surveys in remote alpine areas for over a decade now - typically, I visit areas that are free of serious invasions and walk ad-hoc transects over several days that cover a variety of aspects, vegetation types and altitudes. These long-term observations hint that some species currently not considered serious invaders are 'on the move'.

One such plant is Ox-eye Daisy (Leucanthemum vulgare). From the daisy family, this species has always had a presence in subalpine woodland near Dinner Plain, but it was scattered and apparently not very competitive. In 2011, however, it now appears to have dramatically increased both its range and its abundance - hence, it constitutes an example of a 'sleeper weed' - a species that has been in the area for a long time at low density before rapid population growth.



Ox-eye Daisy

Large areas of subalpine grassland and woodland near Mt Hotham are covered in Ox-eye, and satellite populations are spotting all over the place. This seems to be occurring in the absence of soil or vegetation  disturbance. Worryingly, the same is happening on the snowplains near Kiandra in Kosciuszko National Park. Whether Ox-eye has negative impacts on native plant biodiversity awaits further study, but it highlights that alpine areas are likey to come under increasing pressure from invaders, particularly as climates warm.


Alpine summits currently support few invasive
species - but for how long? (Photo: John Morgan)

Early detection of exotic species would appear to be one of the best defences we have against their invasion and subsequent dominance - hence, on-going monitoring of alpine vegetation would seem a good investment to me and the challenge will be to develop monitoring strategies to detect these intially rare (but potentially important) incursions.



Is conservation science informing conservation?

As a plant ecologist with a strong interest in conservation, I'm often looking at the sort of science and actions that we are undertaking to deal with the great challenges that face native biota of southern Australia. Something recently caught my eye.

A few weeks ago, I read in The Age that Eastern-Barred Bandicoots are being released into a predator proof exclosure at Woodlands Historic Park near the Melbourne Airport (bandicoots head back to the wild). Re-introductions are critical for the persistence of many species that have had their habitat transformed by agriculture (as has happened to the grassland and woodland habitats of the EBB), or where feral predators have obliterated the last of the wild populations.

The introduction of bandicoots into the grassy woodlands (itself an endangered ecosystem) at Woodlands is not new - there have been various attempts to establish a population at this site since the 1980s - and as far as I am aware, there has been only limited success. The causes for failure, unfortunately, have not been well understood, but it would appear that failure has not always been due to predation by feral animals such as foxes. Weight-loss after release (Long et al) and drought (Winnard & Coulsen) have contributed to the failure observed, yet in 2011, more animals are being released behind an expensive (to build and maintain) fence.

Predator-proof fencing at the Arid Recovery Centre

On further investigation, I found that the exclosure erected  is 300 ha in size, yet the home range of the EBB varies between 13 to 20 hectares for males and 2 to 3 hectares for females. I'm not sure how much overlap there is in home ranges of the males, but it is possible that the exclosure might only permit a maximum of 50 or so males, and 100 females perhaps. Is this going to be enough to maintain a viable population in the long-term, as well as maintain the habitat (in this case, grassy vegetation cover) that is necessary for this species? If we are going to invest in the conservation of threatened species to ensure their survival in the wild in the long-term, I'd argue that we are obliged to do it properly so that repeated failures are not the 'norm' of these activities.

There's lots of literature on minimum viable population size and island biogeography that would suggest that small, island populations are at most risk of local extinction due to stochastic, demographic and genetic processes. I note that the predator exclosure at Arid Recovery Centre in Roxy Downs is 60 km2, an area much more likely to maintain wild populations of threatened animals.  Larger fenced areas protect larger (more viable) populations, but are harder to maintain, and more expensive to build and this is the difficult trade-off facing all conservation actions. But, if we want to achieve the outcomes we want (species survival), then we have to factor this into our decision making.

Native grasslands on the Werribee Plains likely
to become  part of the new Western Grassland Reserve
(Photo: John Morgan)

I suggest that there is much to be learnt about optimal exclusion fence size. The new Western Grassland Reserve to be created between Werribee and the You Yangs (as part of the offsets for loss of native grasslands due to urban expansion) provide an ideal opportunity to test the importance of (a) the need for predator proof fencing and (b) the size of exclosures. 

Imagine a large fenced area (say ten times bigger than that at Woodlands) in which animals could be released and their population dynamics followed over the coming years (as well as that of the habitat). This could be compared to areas outside the fence where 'normal' fox control activities are undertaken - indeed, in East Gippsland, the Southern Ark project doesn't rely on fences to preserve native animals but rather, intensive feral animal control. And at Mooramong in western Victoria, EBB released there in the 1990s survive (perhaps in the largest numbers in Victoria) without the need for an exclosure (Winnard & Coulsen). After 5 or 10 years, a comparison of the 'state' of the population in large fenced areas, small fenced areas, and areas where traditional control strategies were undertaken would put us in a much better place to understand how to do this optimally in future. This is the essence of the adapative management approach - the 'learning-by-doing' strategy - it's often emphasised and here's an opportunity to put it into practice. At best, we can learn from our failures - something that I am not sure has always happened in the past with EBBs.