Our research focuses on the population dynamics of plants and how they are influenced by impacts of natural disturbances and global environmental change. We are particularly interested in the interactive effects of fire, grazing and drought in grasslands and woodlands in southern Australia, and how climate change, fragmentation and shrub encroachment affect ecosystems.

Monday, 12 December 2011

The Victorian Bush - its 'original and natural' condition

The Victorian Bush is a book by ex-forester Ron Hateley that encourages a re-examination of popular beliefs about the historical role of fire in Victoria and its use in landscape management. It uses historical accounts of early surveyors and settlers, amongst others, to examine the nature of the Victorian landscape at the time of european settlement, and how the ecology of the bush may have changed in the interim.

Hateley challenges the widely held view that aboriginals dramatically changed the nature of Victorian forests through ‘firestick farming’, i.e. regular and extensive burning of the landscape as a means of ensuring a plentiful supply of food. This is diameterically opposed to the views presented in another recent book utilising historical ecology by Bill Gammage (The Biggest Estate on Earth - How Aborigines Made Australia; Allen & Unwin, 2011) who proposes that aboriginal burning was responsible for transforming large tracts of land into an ‘open and park-like’ state. It's the most interesting aspect of his book - examining the weight of evidence for understanding the role of fire in southern Australia.

While Hateley acknowledges the widespread use of fire by aboriginals, he suggests that its influence in Victoria has been vastly exaggerated. He challenges the oft-repeated claims of proponents of frequent fire that have extrapolated evidence from northern Australia to assume that aboriginal burning practices in the savannas were also employed in the forests of south-eastern Australia. The primary evidence suggests that the tall wet sclerophyll forests and temperate rainforests of south-eastern Australia were characterised by a thick understorey and not subject to regular firing by humans prior to European settlement. Phil Zylstra's research, based on dendrochronology and charcoal and pollen deposits, supports Hateley's view here; fire frequency in the Australian Alps and the wet sclerophyll forests of south-eastern Australia, for example, has substantially increased (not decreased) since European settlement.

This book encourages us to think about issues such as fire management and landscape reconstruction, and to question simplified interpretations of nature. I'd recommend it to anyone with an interest in vegetation dynamics, landscape history and ecological management. The Victorian Bush – its ‘original and natural’ condition is available from Polybractea Press www.polybracteapress.com.au

Friday, 25 November 2011

What makes a plant species fire-resistant?

I've just returned from the annual meeting of the Ecological Society of Australia in Hobart. This is the peak ecological conference in Australia for ecologists and it's always a chance to meet up with old friends, hear about some great new ecology, and to have your thinking challenged by leaders in the field.

Malcolm Gill
This year, Dr Malcolm Gill was awarded the ESA Gold Medal for his long and outstanding contribution to the field of fire ecology. You may be aware of Malcom's work (see here) - I first became aware of him in my Honours year when I read his important review on fire in Australia - Fire and the Australian flora: a review. Australian Forestry 38, 4-25. It's still a seminal review for its time that should be read by all new post-grads working in this field.

Malcolm highlighted, in his own dry way, just how much the field of fire ecology has come since then. In particular, he made the compelling case that plant species in fire-prone ecosystems are adapted to fire regimes and we should place our work in this context. While it is relatively easy to characterize individual fire events and plant community responses to single fires (often limited to the timescales studied by PhD students), these need to be couched in terms of fire return intervals, as well as components of the fires themselves (severity, extent, patchiness). Importantly, he stressed that adaptive characteristics such as thick bark, serotiny, soil seed banks do not guarantee species persistence.

This is something that I have been thinking about lately - that the response to fire of a species can be a population attribute, not a species attribute per se. Let me give you an example.

In 2003, the Victoria alps - from the foothills to the alpine peaks - was burnt in a very large bushfire during a particularly severe drought. The subalpine woodlands and forests, dominated by Eucalyptus pauciflora (Snow Gum), were extensively burned. Snow Gum is a thin-barked species and the above-ground stems seem very sensitive to fire of any intensity. Hence, stems die in just about any fire but individuals recover by vigorous resprouting from a lignotuber. Hence, fire would apparently have very little impact on Snow Gum populations.

By 2009, I started to wonder whether this was true. It dawned on me that in the subalpine woodlands I work in, individual Snow Gums trees had responded quite differently to the same fire event. And this difference was mediated by tree size (or more correctly, girth).

Snow Gum with small girths - fire-resistant?
(Photo: John Morgan)
Snow Gum with big girths - fire-sensitive?
(Photo: John Morgan)

A quick walk through the forest hinted that resprouting from lignotubers (both the number of stems produced, and their vigour) was most pronounced in plants with small girth. By contrast, very large plants had often succumbed to very low intensity fires. We've started to collect data on this response and I hope to soon have some numbers to support these observations. There clearly is a relationship between girth and the number of dormant buds waiting to be released after fire and I'm sure there are both anatomical and physiological reasons for the response. I'm looking forward to unravelling the mechanisms that underpin these observations.

Interestingly, Malcolm talked about similar findings for tropical savanna trees in Darwin. Hence, it's clear (to me) that population responses must be incorporated into our thinking about vegetation response to fire events. Population dynamics are largely ignored when we assign fire response traits to plant species (i.e. resistant versus sensitive), perhaps to our peril.

Thursday, 3 November 2011

What limits recovery of semi-arid oldfields? Seed addition experiments reveal the importance of seed versus microsite limitation

Many native grasslands in southern Australia have been utilised for agriculture. Grazing and cropping are two widespread landuses that have pushed these ecosystems to the brink of extinction. But what happens if you remove agricultural disturbances? Can native ecosystems recover? I've been leading a research team for some time that has been asking: does the function, structure and composition of oldfields return after the cessation of cultivation?

The recovery of native communities after cultivation may be constrained by two key ecological factors: (a) the failure of species to reach a site due to poor dispersal or (b) their failure to survive once there. Seed addition is a common method to test for seed versus microsite limitation. Most studies, however, do not follow populations beyond seedling establishment (to see if long-term persistence occurs), nor do they measure seed dispersal (to see if seed movements really are limiting recovery).

Recently, my grad-student Andrew Scott and I set out to determine the constraints on the recovery of semi-arid grasslands in northern Victoria after cultivation. We examined dispersal across native grassland / oldfield boundaries (using Astroturf to catch seeds at increasing distances from boundaries) and also investigated the relative importance of seed and microsite limitation across multiple life-history stages and generations by adding seed of two grassland forbs that are absent from the early successional recovery after cultivation.

Perhaps unsurprisingly, seed trapping over two seasons showed little movement of native seeds into old fields; most species had extremely localized dispersal of only a few metres. Consequently, similarity between the seed rain and standing vegetation was moderate to high. Hence, seed dispersal is a key constraint to the recovery of oldfields. This might be overcome by the deliberate introduction of propagules.

Goodenia - one of the species we used in seed addition
experiments in northern plains grasslands to test for
seed versus microsite limitation.
(Photo: Andrew Scott)
Seed addition into oldfields showed that two annual species (Goodenia, Rhodanthe) were able to establish in all, and flower in most, oldfields in the first year. Seedling establishment increased with sowing density, consistent with seed limitation. However, the relative importance of microsite limitation increased over the life-spans of the species. Density-dependence reduced the number of flowering plants, resulting in a large decline in seedling density in the following generation. This decline continued so that the initial positive effects of sowing density on seedling numbers disappeared by the fourth generation and hence, the long-term persistence of populations is uncertain. This highlights that seed additions might overcome dispersal barriers, but it does not guarantee that sustainable populations will then develop.

Thus, by monitoring seed dispersal and following experimental populations beyond seedling establishment, a rare achievement in the seed addition research field, we showed that dispersal limits species distributions, but microsite plays an important role in limiting population growth and persistence. Perhaps what this points to is that multiple seed re-introductions (i.e. seeds added across years) will be necessary to recover oldfields.

This work will appear in Oecologia soon.

Monday, 24 October 2011

Science and land management

I'm a plant ecologist partly because it allows me to understand the natural world. But I'm also an ecologist because I want to see the management of that natural world done with the best possible scientific knowledge. This not only promotes the wise and sustainable use of natural resources, but ensures that appropriate management occurs in those areas where nature conservation is the primary aim.

Some of the best understanding of natural ecosystems in Australia can be found in the alpine grasslands, wetlands and heathlands of the Eastern Highlands. Scientific research has a long history in alpine Australia (dating back almost a century) so you'd think land management decisions would be informed by this research, i.e. the level of uncertainity about management actions is much lower there than elsewhere in the landscape where much less scientific research has been conducted.

The high mountain grasslands and wetlands of Victoria are most
extensive on the Bogong High Plains. This is where most of the scientific research
on impacts of disturbance have been conducted.
(Photo: John Morgan)
Cattle grazing impacts on range condition of high mountain ecosystems is one research area that has been particularly prominent. 

Maisie's Plots - started in 1946 by Maisie Fawcett from the School of Botany at Melbourne University - has been a pioneering study examining the effects of cattle grazing on alpine ecosystems. It shows clearly that bare ground is highest in grazed areas, which is not good news for erosion, and that bare ground favours regeneration of shrubs > forbs. The plots continue to be monitored till today, giving new insights on vegetation change in reponse to infrequent fire and climate change.

Maisie's Plots in 2005.
On the left, the area grazed by cattle for 100 ys.
On the right, the area unavailable to cattle since the 1940s.
Note the contrast in vegetation - composition and abundance is
very different inside the fence
(Photo: John Morgan)

More recently, Dick Williams and others tested the hypothesis that "alpine grazing reduces blazing" - in response to the idea that cattle are necessary to reduce fuel loads in alpine areas and prevent mega-fires. There wasn't much evidence that cattle grazing impacted on the extent and severity of landscape fires in 2003. Rather, the patterns of burning were mostly a consequence of fuel type. These studies (and many others), give managers, policy makers, and society a very solid basis on which to make informed decisions about the management of alpine lands. And it was such science that swayed governments to phase out cattle grazing from high mountain catchments over the last 30+ yrs - because of their detrimental effects of wetlands, vegetation cover, soil erosion potential and biodiversity values. Such rigorous assessment should be the standard for all land management activities.

But just last week, the Mountain Cattlemen's Association of Victoria trotted out their oft-stated claim that ''arguably suspect science'' surrounded the decision to remove cattle from the high mountains of Victoria. This is a baseless accusation given the science that has occurred there over six decades has followed normal scientific practices (development of hypotheses, replication, controls, rigorous measurements, etc), as well as being peer-reviewed before publication. Such statements need to be held to account - because without challenge, they can gain some level of credibility and undermine / erode confidence in the scientific process.

 What's often not understood is that scientists are inherently conservative beasts - before proclaiming "effects" of X on Y, they have tested their ideas through replicated observation and experiment, considered alternative hypotheses, and  used statistical inference to interpret their findings. They do this because they want to make sure that they don't proclaim significant effects when none occur - a false positive. This is called a Type 1 error in statistical terms.

Statements like those of the MCAV highlights a mis-understanding of how science works and, more generally, its role in society. Science is an evidence-based discipline, and allows degrees of 'certainty' to be proclaimed - e.g. we are 95% sure that adding nitrogen to native grasslands in Australia promotes invasion by exotic species. Hence, it's generally not a good idea to spread nitrogen around if you want to conserve diverse native grasslands. There will be exceptions to this (in our field, we talk about how effects may be 'contingent' on site factors), but this should not undermine the general confidence we have in making these predictions.

However, in society, people are often inclined to believe what they want, with or without evidence. This is because religion, politics and upbringing all affect the way that people view their world. It makes the likelihood of a fruitful scientific debate on sustainable land management difficult. And how might we argue the case for acting on climate change in such an environment? Discussing the environmental impacts of an increasing population size in Australia (predicted to hit 35 million by 2050) is another topic that springs to mind where science should have an integral input into the debate.

More than 300 years after the Enlightenment ushered in the Age of Reason, superstition and belief still vie with rationality; the scientific method remains ill understood and we are as likely to believe what our peers say as our scientists. In debates we pick sides and become entrenched in positions, rather than weighing up the evidence. So how do we change this idea that our science is peripheral to good management when (I'd argue) it should be front & centre?

Certainly not by giving up! One of the best things we can do, as scientists, is to point out errors of fact and logic that may be promoted in the mainstream media cycle. And, crucially, to accept that conservation biology is not purely a scientific endeavour. [Read John Lawton's excellent article on "The science and non-science of conservation biology" in Oikos 79: 3-5 for an interesting take on this]. It involves society and we need to much better engage with that society (i.e. the users and decision-makers of the land) if we are to convince them of the methods (and merits) of informed land use. It won't be as simple as that, but I've found that if you take the time to explain your science to the average man on the street (or land manager), he's usually really quite receptive. In many cases, it's about making science (and nature more generally) relevant to people who might not really understand how and why we do what we do.

Thursday, 20 October 2011

Open-source Ecology takes Root

A nice little commentary piece on the Nutrient Network Project I am involved in has just been published in Science. You can find it here. It highlights how collaboration, using very simple experiments, can help answer some of the most important questions in ecology. Hopefully, it also highlights that collaboration is exciting, and can allow ecologists (both junior and senior) to contribute to really important science, even if they haven't pulled in huge research grants.

Alpine grasslands at Falls Creek: one of the NutNet sites exaiming
top-down versus bottom-up controls on species diversity in grasslands across the globe.
(Photo: John Morgan)

Tuesday, 18 October 2011

Ecological divides

Fencelines provide excellent opportunities to do comparative ecology.

Across southern Australia, livestock grazing has been so widespread and intense that it's transformed the natural grassy ecosystems, sometimes beyond recognition. Early settlers were drawn to the vast grassy plains of the lowlands, as well as the drought relief that the alpine high plains provided. For ecosystems that evolved in the absence of large, hooved, congregating animals, it's no wonder that changes in the native vegetation were recorded within five or so years of occupation by europeans. In many cases, we can only dream of what the original ecosystems might have looked like. Early paintings provide some insights, particularly about the structure of the vegetation and perhaps some of the dominant tree species. For example, the image below shows that the Yanakie Isthmus at Wilsons Promontory in the 1870s was undoubtedly a grassy woodland at the time of settlement, and extensive at that. Today, the Isthmus is covered in shrubs and hardly a blade of grass can be found, a consequence of 100 yrs of stock grazing and fire exclusion.

From Lookout Hill, towards Mt Latrobe.The trees here are probably She Oaks (Allocasuarina verticillata).
Painted by John Black Henderson, about 1870. (Image provided by Jim Whelan)
Fenceline comparisons, however, provide much more detailed information about the effects of historical and current regimes on the composition of native vegetation. In some cases, the fences went up in the very early days of settlement (around cemeteries, along railway lines, and to demarcate travelling stock routes), preventing the ecological transformation that occurred throughout the remainder of the landscape that was to be grazed. It's no wonder they have been used extensively in ecology as "natural experiments".

To illustrate the importance of fenceline comparisons, I've been just looking at the distribution of C4 grasses on the riverine plains of northern Victoria with a view to thinking about how to restore and manage these endangered ecosystems. Currently, grasslands here are dominated by C3 grasses such as Wallaby Grass and Spear Grass, with a wide variety of annual and perennial intertussock herbs. C4 grasses are very rare (despite their implied importance in the pre-european flora). The current structural and compositional values are being managed by status quo management - using sheep grazing - based on the idea that the vegetation we see today is a product of its recent grazing history and hence, this is the best way to manage it to maintain those values. But what elements might be negatively affected by such a regime?

At Terrick Terrick National Park, there's a little gem of an unused laneway that runs for perhaps 1.5 km. It's no more than 20 m wide and it adjoins large paddocks with native grasslands that have been grazed for a century, as well as cropped at some stage (as evidenced by the cultivation lines). Such a laneway offers an amazing insight into what grasslands might have looked like in the absence of grazing.

Land use on the Northern Plains of Victoria. This native grassland shows the typical
signs of having been 'used' before it was conserved. There are cropping lines, obvious effects
 of water points on sheep movements. And, at the top of the picture, a laneway - different
in colour from much of the vegetation. Is this because it supports a different flora?
(Photo: Google Earth image)

Two things strike me about the fenceline comparison here:
1) in the ungrazed area (left), C4 grasses that are incredibly uncommon in the grazed grasslands - such as Enteropogon (Spider Grass) - are super abundant (and co-dominate with other C4 grasses). Strikingly, C3 grasses are almost absent from the ungrazed area (yet are ubiquitous in the grazed area). Hence, functionally, the laneway grassland is a summer-active one, not a winter-active grassland as now dominates most of the landscape. This is a profound change.
2) unsurprisingly, in the ungrazed laneway, there are lots of sub-shrubs and herbs which are absent from the grazed areas. Maireana and Atriplex, in particular, only exist at high abundance where grazing has been minimal.

These are not new insights. Changes in species composition due to grazing have been recorded many times before in Australia. This is often due to differences in palatability. But, it does point to some key questions.

Does grazing promote C3 grasses that have a regeneration niche dependent on disturbance? It is thought that the C4 grasses regenerate better where the ground is covered in plant litter, but this is removed in grazed areas, potentially favouring Spear Grasses. Perhaps this is why there are no Spear Grasses in the laneway. This needs to be experimentally tested.

Can status quo management ever allow the system to recover poorly represented components of the original flora? The laneway is so different that it is clear that ongoing grazing by sheep maintains different vegetation states across a fenceline. It does this because it prevents the C4 components (and other subshrubs & herbs) from transitioning back into the system. If every grassland is managed in the same way on the northertn plains, then we are conserving a subset of the original flora. And a subset of the ecosystem services and functions performed by such a flora.

It would be incredibly useful to assess the plant trait distributions of species found in the laneway and adjoining grazed grasslands. Does grazing maintain species richness, but at the expense of functional diversity? Are we causing a functional homogenization of the grasslands in our pursuit of conserving exisiting values?

I've highlighted just how important fencing experiments can be to show that vegetation patterns are affected by management history. Providing a mechanistic understanding of the changes that occur, however, has been less well demonstrated. I hope to fill in these gaps soon. Till next time....

Sunday, 2 October 2011

Colder plants in a warmer world?

I'm keen on cross-country skiing. I guess it's an extension of one of my other great loves: bushwalking. I can think of almost nothing better than strapping on some skis, packing the tent and shuffling out across the high plains of the Australian alps to camp in the snow on some distant peak.

I particlarly love spring skiing - it's warm (so I get to ski in shorts!), the days are long, the snow is hard in the morning - which makes for some fast downhill runs, and there usually aren't as many people around. I was hoping to get one final ski in this season, but like many years (perhaps seven out of the last ten), it's been another disappointing year for snow in Australia. That doesn't stop the ski resorts here from charging top dollar, but that's another story.......

The SnowCams at Mt Buller don't lie! Even at one of Victoria's highest
peaks (>1800m), there ain't much white stuff and hasn't been all winter

What we are seeing now is probably a window into the future.

About a decade ago, Kevin Hennessey and collegues at the CSIRO modelled the potential distribution of snowcover at 2050 and showed that,  under 'high warming' scenarios, very few mountains in Australia would sustain snowcover for >90 days. I think we won't have to wait till 2050 for that to become a reality. This will likely have lots of consequences in the Australian Alps such as (a) there being less water released slowly in the spring thaw (with resultant greater stream surges with potential for downslope flooding) and (b) lower albido.

One of the biological consequences that hasn't really got much attention yet is how less snowcover will affect those plants (and animals) that rely on snow for protection from the cold. Generally speaking, snow is a magnificent insulator against the cold. In the sub-nivean space (the interface between snow and vegetation), it rarely drops below zero degrees. Hence, plants there are not exposed to extreme cold (frost) and can even continue to grow over winter in some cases. But with less snow, there will be less of an insulating blanket, and this will invariably melt earlier each year, exposing plants to extremes of cold they perhaps are incapable of resisting. Might alpine plants in a warmer world actually be more susceptible to the cold?

Susanna Venn, Janice Lord and I have been investigating the frost sensitivity of alpine plants with this question in mind. We've been assessing the early spring frost tolerance of a range of alpine plants to determine just how they might react to earlier snowmelt.

For those (mostly woody) species that occur on wind-exposed ridges where the snow rarely settles, it is clear they are very cold tolerant early in the season. This is because they are rarely insulated from winter cold. The Alpine Star Bush (Asterolasia trymaloides), for instance, has an LD50 of -17.6 deg C in the weeks after the spring thaw. Lethal temperatures were considered those at which 50% damage occurred to the photosynthetic apparatus in leaf samples. Pimelea axiflora is even more cold tolerant - it's LD50 is -18.7 deg C.

Brachyscome nivalis - less cold tolerant than many alpine shrubs
(Phto: John Morgan)
By contrast, species (mostly herbs) that are normally deep under snow for the entire winter and hence protected from the extremes of cold, and which don't melt out till mid-summer, exhibit much less cold tolerance early in the season. Brachyscome nivalis, for instance, has an LD50 of just -7.0 deg C when it was removed from snow early in the spring. This hints that it might be sensitive to early melting if that meltout coincides with very cold spring temperatures. To avoid frost damage, the species would likely need to remain in areas of late melting snow (as it currently does), areas which are predicted to decline in extent over the coming decades.

So, while there will be opportunities for extended growing seasons in the high mountains of Australia, it is likely that frost sensitivity may constrain some species. This needs to be ascertained for a far larger number of species than we have examined so far.

Thursday, 22 September 2011

Getting over the Hump

For decades, ecologists have toiled to nail down principles explaining why some habitats have many more plant and animal species than others. Much of this debate is focused on the idea that the number of species is determined by the productivity of the habitat. Shouldn't a patch of prairie contain a different number of species than an arid steppe or an alpine tundra?

Exactly how biodiversity relates to productivity, however, has been a highly controversial issue in ecology over the past few decades. Initially, for example, many researchers believed that the relationship between species richness and net primary productivity (often expressed as the number of grams of carbon produced within a square meter of an ecosystem over a year) could be visualized as a 'hump-shaped' curve, with richness first rising (as stress is alleviated) and then declining with increasing productivity (due to increasing effects of competition).

I've just been involved in a global study that tries to shed light on the hump-shaped relationship between species richness and productivity in grasslands across the globe. Our paper in Science has just been published and the results were, to put it mildly, rather surprising.
Fig. 1
Locations of the 48 sites that provided data for this study

In a multiscale assessment of 48 herbaceous plant communities on five continents (including four Australian sites), we demonstrate that the modal productivity-diversity pattern is quite rare in nature, rather than the dominant relationship. This is really interesting given that so much attention had been devoted to this theory.

Our study shows no clear relationship between productivity and the number of plant species in small study plots when you look at data collected from across a range of grassy vegetations from across the globe. These range in temperature, rainfall and geological and evolutionary history. Indeed, the evidence for the predicted 'humped-back' reponse is rather weak within most of the sites studied regardless of these big differences in history.

Within-site relationships between productivity, measured as peak live biomass (dry weight)
and species richness. The inset shows the frequencies of relationships that were
nonsignificant (NS, thin dashed lines), positive or negative linear (thick dashed lines),
and concave-up (+) or -down (–) (solid curves). The marginal histograms show the frequency
 of species richness and peak live biomass across all sites.

The findings suggest that ecological understanding may advance more rapidly if ecologists focus on exploring a range of topics that are germane to the productivity-diversity relationship in a changing world (e.g. how will loss of species affect productivity, stability and resilience of ecosystems to change), rather than continuing the search for a dominant pattern ("the Holy Grail"). Our work not only sheds light on this classic question, it also demonstrates the power of a network approach. Working in a network with ecologists who share your passion and interest in theoretical ecology is a great way to facilitate insights into the functioning of nature, insights that aren't possible in a focus on individual ecosystems.
Here is a link to the paper:

And an accompanying Perspectives piece

Wednesday, 14 September 2011

Offspring for the next generation

One of the interesting questions our Lab asks is: how is it that many plant species can coexist together at high density? In the diverse woodlands of western Victoria, for instance, Brandon Schamp, Jodi Price and I are trying to understand just how diversity is partioned in space (due to microsite variation - or 'niche availability') and time (how fluctauting resources such as rainfall affect coexistence). These are two really valid approaches to studying patterns of diversity at small scales, but I've recently started to think about this problem in another way.

Diversity in woodlands in southern Australia can be very high
at small scales: up to 45 species per square metre (Photo: John Morgan)

  Within the crowded natural plant populations of species that we work with, the traditional prediction is that most of the offspring from which future generations are drawn will be contributed by the relatively few individuals belonging to the larger size classes. Yet, the extent to which this is true should depend on whether the inevitably more numerous, but relatively small suppressed plants within the population manage not only to survive suppression, but also to reproduce before death. If they can do this, then species coexistence should be ensured because R (the intrinsic rate of growth) >0 for most of the coexisting species.

Hence, resident species are successful not because they are relatively large, but because they produce numerous descendants from numerous (often small) offspring. Lonnie Aarssen has coined the term 'reproductive economy' to describe this phenomenon. So, if species can survive to reproduction despite the clear disadvantages of growing at high density, there is an opportunity for that species to maintain itself in that space - and if many species can do this, then there is a simple mechanism to explain how species can coexist.

We don't have much data on reproductive success and size of individuals in a population. Last year, we harvested all flowering individuals of four herbaceous species in grassy woodlands in plots that were about 100 m2 in area. We then dried and weighed these individuals and plotted the frequency of plants by weight. The data for one species (Arthropodium strictum) is shown below (and yes, the sample size is only 87 - but we are collecting more data this year).

The number of flowering plants by their size in one population.
Note: most reproductive plants in this population are small.

What appears to be clear from this data is that the reproductive size profile for this species is conspicously right-skewed. Most flowering plants are small, but these numerous plants are likely to contribute lots of seed to this population. This is better than leaving no progeny for the next generation. Hence, if small plants of many species can successfully reproduce at high density, this hints that species might be able to persist. Rather than being the weaklings, their collective reproductive output vastly exceeds that of their larger (but fewer) conspecific neighbours.

We intend to collect data for a large range of species this spring and see if most individuals are (a) small but (b) successful at flowering and if so, to quantify just how many seeds enter a population and how many are contributed by the smallest of the individuals. Stay tuned.

Wednesday, 17 August 2011

Assessing habitat condition – an old approach to an old problem

The drawing above, depicting a one yard by one yard quadrat, was published in 1936 (R.T. Patton Ecological studies in Victoria; Part IV – Basalt Plains Association. Proc. Roy. Soc. Vict., 48, 172–190). It is one of the first descriptions of the diversity in native grasslands near Melbourne. Historically, botanists tended not to focus on the area west of Melbourne. Von Meuller (the famed 19th century botanist) seldom visited the area. It was not until 1916 (Sutton) that the flora was reviewed and then only the Keilor Plains. By the time of John Stuwe’s review of native grasslands in western Victoria in 1986, he was only dealing with 0.15% of the original area of Themeda grasslands.

Such historical information is of immense importance when trying to determine how the composition and structure of native grasslands has changed in the almost complete absence of floristic information, and to identify remnants that most approximate the historic condition.

Despite the small samples size (n=1), we can gain many importance pieces of information from this 90 x 90 cm quadrat:
Native species richness = 12 (3 grasses – all perennial, 9 forbs – 1 annual, 8 perennial)
Exotic species richness = 1 (the annual grass Aira)
Themeda tussock density = 13 (comprising many small tussocks)

So how does a current day native grassland dominated by Themeda compares to one historic example we have?

Quadrat depicting current species density at Mt Cottrell
(Drawing by Hannah Forrester)

Recently, some students of mine re-drew grassland quadrats in a conservation reserve that is less than 3 km from Patton's original site. The grassland is dominated by native species . A typical example of the current day diversity is shown above.

Native species richness = 3 (the dominant perennial grass + 2 forbs)
Exotic species richness = 4 (1 annual grass, 1 annual legume, 1 Iridaceae, 1 perennial forb)
Themeda tussock density = 31 (comprising many small tussocks)

The grassland at Mt Cottrell looks like a high quality remnant
- but comparison with historical data suggests otherwise
(Photo: John Morgan)
It's clear that, despite the grassland looking like a native grassland from a structural perspective, profound changes have occurred. Invasions of non-native species are one obvious change - but they are largely sub-ordinate species and do not necessarily strongly influence the function of the grassland. Rather, it is the almost complete loss of interstitial native species that has been most profound. Where there might have been 11 (largely perennial) species in the inter-tussock spaces, our survey showed just two. And these occurred at incredibly low density.

It's likely that the loss of native species has had two principle drivers - the loss of palatable species during the period of stock grazing that preceeded the declaration of the reserve probably accounts for the absence of many typical grassland forbs unaccustomed to on-going and heavy herbivory. And some loss might have also occurred due to the lack of frequent burning. Themeda assumes monodominance at the site, and appears to strongly compete for light and space with interstitial species. Many temperate grasslands are burnt irregularly after conservation, leading to rapid declines in diversity, so similar processes might be at play here.

While the comparison we have made is very simple, and perhaps old fashioned, the historical data gives us something to aim for when considering the restoration objectives for the site. It tells us what sort of native species might be returned and at what densities. This is an incredible challenge, and not for the faint hearted. But at least our objectives are evidence-based and this is likely to lead to restoration that has a clearly defined end-point. It tells us that the re-establishemnet of interstitial forbs, not weed control, is our priority, and that to do this, reducing the dominance of Themeda is paramount.

Sunday, 7 August 2011

Global plant trait data now more accessible

Theophrastus - one of the first botanists!

The use of plant functional traits to describe patterns in ecology has a long history. Indeed, it was Theophrastus who, in about 300 BC, first categorised plants by their growth form: tree, shrub, herb. Ian Wright's paper on the Worldwide Leaf Economics Spectrum remains a classic for me because it showed how using a large dataset (in his case, >2500 species) could search powerfully for patterns in leaf traits and how these were shaped by climate drivers. Such studies are well-beyond the scope of the average researcher, but by collating and sharing data, such insights become possible.

The PFT field is flourishing - so well that a new database has just been launced to act as a repository for trait data. The unimaginatively named TRY Database looks like a real winner for ecologists interested in examining large-scale patterns in nature using the traits of plants that underpin their response to environment and disturbance. It's not a publicly accessible database - you'll have to provide datasets to get access - but I think it could act as an important repository for trait data that remains buried in Honours and PhD theses, and the recesses of Excel files.

Location of sites across the globe for which plant
trait data has been submitted to TRY

This database aims to gather datasets that cover a variety of biomes, geographic areas, and traits. Already, the database comprises about 2,400,000 trait entries for more than 64,000 plant species and about 1000 different traits. My own research group is busy collating the data we have assembled over the years - an impressive dataset of >500 species from the temperate and alpine grassy ecosystems of southern Australia. In the spirit of co-operation, we hope to submit our data to TRY so that others can use our hard-won data.

For further information on TRY, a couple of recent papers have just been published:

Kattge et al. (2011) TRY - a global database of plant traits. Global Change Biology 17, 2905-2935

Kattge et al. (2011) A generic structure for plant trait databases. Methods in Ecology and Evolution 2, 202-213

Monday, 25 July 2011

Monodominance in C4 grasslands

A typical view in southern Australia. Here,
Kangaroo Grass dominates the structure
and cover of a native grassland.
(Photo: John Morgan)
One of the things that strikes you about the C4 grasslands in southern Australia is the complete dominance by Kangaroo Grass (Themeda triandra). The concept of monodominance is actully rather rare in grasslands. Only a minority of the world's grasses (around 600 out of 11,00 speces) are documented as being ecologically dominant. These dominant species, however, seem to share a common(ish) evolutionary history.

In an interesting paper on the origins of C4 grasslands by Erika Edwards and collegues, dominant grasses appear to be phylogenetically clustered, suggesting that certain clades of grasses are more prone than others to evolve traits that promote ecological dominance. But what might these traits be?

The answer to this question is not as straightforward as we might presume. While we accept the fact that Kangaroo Grass dominates the grasslands of southern Australia, there is still some uncertainty about why it does so. There's likely to be a few reasons. Some of them are evolutionary, while others are more ecological. Here, I outline a couple that I think are likely to be important.

Kangaroo Grass is a C4 species. "C4 photosynthesis" refers to a suite of biochemical and anatomical traits that increase photosynthetic efficiency in high light and high temperature environments. While C4 enhances the efficiency of photosynthesis, C4 plants only have an advantage over C3 plants in certain conditions - namely, high temperatures and low rainfall. Hence, C4 grasses are conspicuously absent from the world's cooler regions. This may, in part, explain the dominance of C4 grasses such as Kangaroo Grass in southern Australia.

Kangaroo Grass, in Grime's CSR plant strategy scheme, would comfortably be considered a competitive species. Such species are able to outcompete other plants by most efficiently tapping into available resources. Competitors do this through a combination of favorable characteristics, including rapid growth rate, high productivity (growth in height, lateral spread, and root mass), and high capacity for phenotypic plasticity. This last feature allows competitors to be highly flexible in morphology and adjust the allocation of resources throughout the various parts of the plant as needed over the course of the growing season.

Kangaroo Grass might be thought of as a pyrogladiator. It is fire-adapted, resprouting strongly from basal meristems with very little fire-induced mortality. By contrast, our Lab has shown that C3 grasses, such as Wallaby Grass and Spear Grass, can experience substantial levels of tussock mortality after fire (perhaps because of the fire event itself), further weakening their position in C4-dominated grasslands. Kangaroo Grass, by contrast, quickly accummulates biomass between fires, probably because the C4 pathway supports high photosynthetic rates and nitrogen use efficiencies, especially in the high-light environments after fire. The high water-use efficiency afforded by C4 metabolism probably also provides a competitive edge.
Beth Forrestel, from Yale University, admires
a C4-dominated grassland at Vite Vite on the
western plains of Victoria.  (Photo: John Morgan)

The concept of monodominance is not just of academic interest.

Dominant species shape communities and drive ecosystem processes. Our research has shown that healthy swards resist weed invasion. Hence, they should also be of key interest to restoration ecologists wanting to restore resilient ecosystems (see my last Blog as an example of this). Finding ways of returning dominant species (across large scales) might therefore be just as important as returning rare species to ecosystems.

And understanding how dominant species respond to climate change is a challenge we are only just starting to tackle.

Thursday, 7 July 2011

Grassy White Box Woodland Restoration

Native grasslands and woodlands in Australia have been transformed since European settlement. Because they occur on the fertile soils (by Australian standards), and are dominated by palatable grasses, they were amongst the first ecosystems settled, and amongst the most intensively utilised. As a result, much of the original ecosystem has been lost - to cropping, to grazing, and to pasture improvement. Probably less than 15% of woodlands remain in eastern Australia, and grasslands occupy much less than 5% of their original range.

White Box woodland in a bush cemetery
(Photo: S. Prober)
But some remnants do survive - in areas that have escaped heavy utilisation such as bush cemeteries, travelling stock routes, town commons and railway lines. And they survive in relatively weed-free states with high native plant diversity. They provide a key insight into how degraded remnants might be restored.

Ian Lunt from Charles Sturt University and Suzanne Prober from CSIRO have been working for a long time now on the conservation and restoration of white box woodlands in southern Australia. Given the perilous state of these systems, their scientific studies are at the cutting edge of practical conservation biology.

They have looked at why small remnants have maintained their diversity - and come to the conclusion that soil nutrients plays a key role. Where nutrients are high, exotic plants are favoured and these tend to outcompete the small native species that have evolved to survive on scant resources. Where nutrients are low, native species thrive because the exotic species basically have too few resources to survive.

Soil nutrients increase for a couple of reasons - the most obvious one is that they are applied by farmers to increase productivity. Less well known, however, is that when deep-rooted, long-lived perennial native grasses such as Kangaroo Grass are lost from ecosystems, lots of nutrients (including nitrate) are released into the soil and the elevated levels favour annual grasses. Annual grasses, of course, are short-lived - so they use soil nitrate to grow and flower, but because they die each year, that nitrate gets released back into the soil to be used again in the following year by even more annuals. And so the cycle continues.

Therefore, the groundlayer of grazed and degraded remnants rarely recovers well after fencing and livestock exclusion because these sites often have high soil nitrate levels that favour the exotics. So, how to overcome this problem.

Ian and Suzanne have found that it is absolutely imperative that soil nutrients be reduced if native species are to be re-established, but this is easier said than done. Indeed, until their work begun, few conservation biologists had really thought about this problem in Australia. They have trialled a number of techniques in small experimental plots - well replicated of course! Their treatments included i) re-establishing deep-rooted perennial native grasses (to lock up nutrients), ii) burning (this leads to some loss of nitrogen in smoke), and iii) adding sugar (to reduce nitrogen availablility due to microbial activity).

The results have been nothing short of stunning, and give hope that grassy woodlands and grasslands can be restored. It's an example of how really good science can inform practical conservation outcomes. Their work has just featured on the ABC's science program Catalyst - which I've included here so you can see what this long-term study has been able to achieve.


Tuesday, 5 July 2011

2010 ISI Impact Factors are now out

You can't ignore the fact that Impact Factors have had a huge effect on publishing trends and the choices authors make about where to publish. This is somewaht unfortunate - I now hear scientists talking in the corridors of universities, or worse, at conferences, about where they published their most recent paper, not what they are publishing on!

Regardless of how much weight you put on Impact Factors (see this damning review as evidence that some don't rate IFs at all), all new PhD students and Post-Docs have to play a game of publishing in (perceived) high impact journals if they are to get that next job.

So what are Impact Factors and how are they calculated?

In a given year, the impact factor of a journal is the average number of citations received per paper published in that journal during the two preceding years. For example, if a journal has an impact factor of 3 in 2009, then its papers published in 2007 and 2008 received 3 citations each on average. The 2009 impact factor of a journal would be calculated as follows:
A = the number of times articles published in 2007 and 2008 were cited by indexed journals during 2009
B = the total number of "citable items" published by that journal in 2007 and 2008. ("Citable items" are research papers; not editorials, book reviews or Letters-to-the-Editor)
2009 impact factor = A/B.
I've included the 2010 Impact Factors for journals in conservation and plant ecology (and compared their "performance" to their 2009 rating). The big winners were Ecology Letters, Trends in Ecology and Evolution and Frontiers in Ecology and the Environment. Interestingly, the IFs of most journals rose over the last year.

Applied Vegetation Science: 1.802 (2010) versus 1.349 (2009)
Austral Ecology: 1.820 versus 1.578
Australian Journal of Botany: 1.681 versus 1.868
Biodiversity and Conservation: 2.146 versus 2.066
Biological Conservation: 3.498 versus 3.167
Conservation Biology: 4.894 versus 4.666
Diversity and Distributions: 4.248 versus 4.224
Ecography: 4.417 versus 4.385
Ecological Applications: 4.276 versus 3.672
Ecology: 5.073 versus 4.411
Ecology Letters: 15.253 versus 10.318
Frontiers in Ecology and Environment: 8.820 versus 6.922
Functional Ecology: 4.645 versus 4.546
Global Change Biology: 6.346 versus 5.561
Global Ecology and Biogeography: 5.273 versus 5.913
Journal of Applied Ecology: 4.970 versus 4.197
Journal of Biogeography: 4.273 versus 4.087
Journal of Vegetation Science: 2.457 versus 2.376
Molecular Ecology: 6.457 versus 5.96
Nature: 36.101 versus 34.480
Oecologia: 3.517 versus 3.192
PNAS: 9.771 versus 9.432
Polar Biology: 1.445 versus 0.582
Science: 31.364 versus 29.747
Trends in Ecology and Evolution: 14.448 versus 11.564