How to burn a grassland

A Themeda-dominated grassy woodland in
western Victoria.
Photo: John Morgan

Fire in temperate grasslands and grassy woodlands in southern Victoria is something close to my heart. I spent 4 yrs during my PhD trying to understand what role it might play. I focused on the intertussock flora primarily - the biodiversity. In short, fire in productive grasslands:
 * removes biomass of the dominant grasses such as Kangaroo Grass (Themeda triandra). This, importantly, improves light penetration and allows less competitive species (the intertussock forbs) to coexist with the competitive C4 grasses that contribute most of the biomass in grasslands.
* does not cue germination of (the majority of) intertussock species because most forbs have transient soil seed banks. Rather, these species primarily recover by vegetative resprouting from stolons, rhizomes, tubers and bulbs.
* promotes flowering of forbs, geophytes and grasses in the spring after burning. This pulse of flowering leads to lots of seed, which leads to delayed seedling regeneration, i.e. 12 months after an autumn fire.

More recently, I've become interested in the role that fire plays in maintaining mono-dominance in grasslands. While Kangaroo Grass dominates grasslands and some woodlands, in the long absence of fire, we have found it tends to smother itself and the tillers die. New species then move in, including exotic perennial grasses. Clearly fire removes dead thatch and allows high numbers of vigorous tillers to be maintained in each tussock.

But I think fire also is important in another key way.

I've been following mortality after fire of C4 grasses like Kangaroo Grass and comparing it to other native grasses like the C3 genera Austrostipa (Spear Grass) and Rytidosperma (Wallaby Grass). These genera are components of grasslands, but rather uncommon when Themeda dominates. This might be because they are inferior competitors to Kangaroo Grass or, and it's a hypothesis at this stage, it's because they are less effective at resprouting after fire than Kangaroo Grass. In part this may be because their tillers are rather loosely arranged and not well-protected from the heat of a fire. Hence, with every fire, if mortality is high in these species, but low in Kangaroo Grass, then fire would re-enforce dominance by Themeda.

To test this idea, I have been setting fire to grasslands and grassy woodlands these last few days. It's the height of summer here, and the C3 grasses are largely dormant, but Kanagroo Grass is growing and green(ish). I'm going to follow mortality after fire of the C3 and C4 grasses that I have introduced into these areas, and will report on the findings soon.

But I thought I'd show you how we go about burning a grassland/woodland. I've had the local rural fire brigades helping burn a site for me in western Victoria - they are more than keen to do this because it reduces fuel.  Actually, they burn this grassland almost annually, so the fire intensity is generally very low (100 - 300 kW/m) because of low fuel loads.

1. First, a mineral fire break is ploughed on one side to create a fuel-free barrier on the edge of grassland. This edge is then wet down by a fire tanker. It is critical that the fire edge can be contained, otherwise private landholders are put at risk.

Wetting down the fuel edge
Dunkeld, Jan 2013

2. On the other edge of the grassland, the downwind edge, the fire is lit using a drip torch. This is allowed to burn slowly for several metres into the wind to create an edge that has no fuel and hence, will allow us to control the downwind fire.

First, the fire is lit on the downwind side to create a burnt out break

3. Then, a second fire line is lit, half way across the grassland. This is normally lit when the first fire line is 100 - 200 m in front. This burns quickly downwind into the slowly moving fire moving upwind from the edge.

A second fireline is lit, half way across the grassland

4. A final fire line is lit on the edge . It burns quickly downwind but extinguishes as soon as it hits the burnt grassland.

A third fireline is lit from the mineral earth break

While the fire burns the grassland  thoroughly (because Kangaroo Grass is very flammable), it is clear that fire has done little more than burn the canopy. A few minutes after the fire has passed, you can see green tiller bases remain. And the soil crusts that carpet the intertussock spaces appear little singed. Hardly any heat is transferred to the soil, and even small saplings are only partly singed. This is probably a function of the low fuel loads here and low flame heights (<50 cm).

Low flame height and rapid consumption of the vegetation are typical

This photo was taken 20 mins after it was burnt. Green vegetation remains
at the base of this Kangaroo Grass tussock.

I'll be interested to see if C3 grass mortality is high here, despite the relatively 'cool' fire. I'll need to see if the C3 grasses re-sprout and then die in the summer drought. Or whether they fail to resprout altogether because fire has killed them. One thing I do expect to see is rapid regrowth by Kangaroo Grass as this is its primary growing season, and it is very deep-rooted, allowing it to tap into deep soil moisture. Perhaps it is this rapid regrowth (and ultilisation of soil moisture) that elevates C3 grass mortality?

Thanks to Anthony Watt, CFA, and the Cavendish, Hensley Park, Karabeal, Mirranatwa, Grange & Hamilton brigades

Keeping up-to-date in ecological research

It's hard to keep up-to-date with all the new ecology that is being conducted across the globe. Not only is there precious little time to read scientific papers, scan journals for interesting work, and do research, but the sheer volume of material being published means that many interesting research papers will slip by. This is a shame, because it forces us to narrow our search focus when we do read papers.

So, I have started to use Blogs to keep up to date with ecological developments. I started my own Blog because I wanted to communicate about the research we do in my Lab, and to comment on ecology more generally. It's been almost two years since I started, and I've had 15,000 page views. So, I guess someone finds this stuff interesting! Assuming that others share the same philosophy, I started to scan for Blogs that summarised new papers in the general area of Ecology and Evolutionary Biology and came across The EBB and Flow: blogging all things ecology and evolutionary biology. This is a really neat Blog. It's big on reporting about papers that focus on ecological theory (which we should use as the basis for all our investigations) - under the 'Research Focus' tab. The most recent Blog on a new paper by Jonathon Levine about how coexistence theory can help inform community assembly makes a difficult/challenging topic palatable.

EBB and Flow also has excellent sections on 'Conservation Focus', 'Academic Life' (which PhD students will find really informative, 'Career Corner' (ditto), 'Meeting Dispatches', 'Recent Papers of Interest' and the excellent 'Researcher Spotlight' that send the reader off into weird and wonderful (and more importantly, relevant) ecological territory. It's well worth bookmarking!

Happy New Year!
JOHN

Ecology Apps.....for Christmas?

I've just upgraded my mobile phone and got myself a smartphone. My Grad Students would perhaps say it's about time I got into the year 2012 (which is not so great given it'll be 2013 in a little over a week). It's not that I don't like technology; rather, I'm a slow adopter. My current phone works fine, and I do hate the idea of consumerism.

But, I've bitten the bullet. One of the reasons was that Nick Bell, a Summer Research Student working on our long-term ecological plots examining alpine vegetation change, recently showed me how useful smartphones can be for the average field ecologist. While he was talking to me about technology stuff that seemed to be based on the English language, he showed me two simple Apps that I might find useful.

Nick talking to John
(thanks to Calvin & Hobbes)

One was an app to help measure tree heights (Smart Measure). Having recently done this with Nick in the field the old fashioned way - tape measure, compasses, pythagoras' calculations - this seemed really simple and effective. Then, he showed me how you could ghost an image on the screen while trying to re-take the same image. This would be awesome for my re-photography work that I am currently undertaking, comparing photos from the 1930s with the current day.




Measuring tree height is easy. Right?
Source: http://www.nativetreesociety.org/measure/tree_measuring_guidelines.htm

I went away and started to do some research. Not on getting a smartphone, but on the apps that I could use in ecology. Very quickly, I was thinking "hmmm, there's a lot of technology available that is cheap, accessible, and that I've totally overlooked".

One of the sites I found was Emilio Bruna's excellent 'Mobile Ecology' webpage. This page lists a bunch of apps for research, teaching and outreach (with a US emphasis, so I guess I'll have to look for the Australian equivalent). If you're interested in such things, it is well worth a look. My favourite was the app called RInstructor - this looks like something that will really facilitate my learning and use of R, particularly how to customise plots.

So, with Christmas coming up, I might just have to treat myself and upload (or is that download) a couple of these apps to see whether my investment in a new phone was actually worth it! In any case, I can see great potential in this technology to help me facilitate my data collection and efficiency in the field.

Merry Christmas, and best of wishes for the New Year. I look forward to writing more about plant ecology in southern Australia in 2013.  JOHN

Fire in south-east Oz - 'new' things to observe

I've just returned from the annual meeting of the Ecological Society of Australia, this year held in my home town of Melbourne. As usual, there was a great mix of student presentations, challenging plenary talks, and ample time for discussion during the breaks.

One thing that struck me was the number of talks I saw on fire. There was a whole symposium dedicated to fire, which was very well-attended. I also saw talks scattered through the programme that dealt with the impacts of individual fires, the regeneration biology of key species, the importance of fire return interval, why fire severity matters, etc, etc.

This is not surprising given that south-eastern Australia is one of the most fire-prone areas on the globe. It is a recurrent feature in many ecosystems, and is projected to increase in frequency because of climate warming over the coming century. Additionally, it is also a contentious issue - the management of bushfire threat in SE Oz revolves around policies that mitigate risk by hazard reduction burning.

Two important issues arise - how often should it occur in an ecosystem (both as a tolerable event and as a desirable disturbance to cue important ecological processes such as seed germination). Second, what type of fire event is likely to have positive or negative impacts on these outcomes.

I've been musing about these questions for the last few weeks, but from a different perspective. And it returns me to one of my favourite pasttimes: observations in ecology.

While fire is a recurrent event throughout much of SE Oz, there are indeed very long fire-free intervals for some parts of the landscape. We intuitively know this. In the rainforests of eastern Victoria, fire is probably fairly rare - perhaps in the order of hundreds of years apart. But what is probably less appreciated is that in the drier, less productive parts of central Victoria, perhaps fire is also equally as rare.

How can I make such a statement?  What evidence do I have for one of the most fire-prone areas of the world having very long fire-free intervals. Well, that is where observation comes in. In the dry, eucalypt-dominated woodlands of north-central Victoria, around Rushworth to Wangaratta, there is a clue on the upper slopes and rocky outcrops. It's a pretty obvious clue when you think about it.

Here, there is a plant that is born to burn (or so it would seem). Grasstrees, in the genus Xanthorrhoea, are arborescent monocots, developing tall stems that allow them to grow to great heights. We know they grow very slowly. Let's focus on X. glauca. For the first 50 yrs, plants consist of leaves but no stem. The stem then emerges and height growth is about 10-25 mm per annum. My old PhD student Peter Curtis measured growth rates over 10 years, so this seems a pretty reasonable estimate. So, a plant 3 m tall might be anywhere up to 350 years old. Each year, the leaves die but are held appressed to the stem rather than shed. In unburnt grasstrees, this 'skirt' can extend all the way along the 'trunk' to the ground. When a fire occurs, the 'skirt' is a ready-made fuel source to rapidly burn the plant - a quick fire where most heat is carried rapidly away from the meristem by convection. Plants generally survive burning, although Peter's work does show that mortality of grasstrees can be quite substantial in the decade after fire. But that is another story..........



Burnt Grasstrees, Rushworth State Forest, October 2012
(Photo: Michele Kohout)

I've been wandering around the bush, observing something recently. In the last few years, many tall grasstrees have been burnt as part of new 'targets' set by the State Government - hazard reduction burns. This, in itself, is notable because many areas of bush being burnt clearly have not been burnt for centuries. How do I know this?

Well, old photos tell us so! Here are some images of grasstrees in the Warby Ranges - taken in the early 1980s by the well-known fire ecologist David Cheal - that are notable for (a) their enormous height (up to 8 m perhaps) and (b) the fact that they have grass skirts that extend all the way to the ground. This suggest that this part of the landscape has not been burnt for upwards of 600 years. Hence, we are not talking about fire suppression since european settlement. And it is not just one or two individuals, but entire slopes, suggesting that fire is very uncommon here. On recent inspection, it is hard now find examples of these great unburnt plants. Low intensity fire has seen to that.



Grasstree in the Warby Ranges, 1983. Note the scale (approx. 1.75 m)
and dead leaves all the way to the ground. The trees in the background
have not been burnt. Rather, they are dying from intense drought.
(Photo: David Cheal)

Grasstrees, all with skirts that indicate long intervals between fires, across a slope at
the Warby Ranges, 1983. Plants are approx. 2.5 m tall.
(Photo: David Cheal)

Long unburnt Grasstrees, Rushworth State Forest, 2012. Scale approx. 40 cm.
(Photo: Michele Kohout)
 

The lesson here, if there is any, is to observe patterns in nature. The simple observation that grasstrees in some parts of the range (low productivity, rocky) have not experienced frequent fire is probably very important, particularly when thinking about fire return intervals. While low intensity fires may seem ecologically benign, this needs to be put in the context of the fire history of the site.

It also challenges our notion that dry forests burn frequently simply because they occur in a part of the continent where fire is known to occur regularly!

The basics of repeat photography

The photo below, taken in 1979 by La Trobe University botanist Neville Scarlett, is of one of the best remnants of native grassland near Melbourne called Truganina Cemetery. I know this site really well as I did some of my Honours thesis research there, on the recruitment ecology of the Button Wrinklewort.

What is great about old photos is they convey information about the historic state of systems, well before many of us were involved in management and research, and they potentially allow current condition to be compared to prior states. Clearly, this grassland had a high cover of forbs in 1979. The silver coloured vegetation is probably the Lemon Beauty Head (Calocephalus citreus), the yellow flowing daisy is Chrysocephalum apiculatum, while the white sprays are mostly Bracyscome dentata in the background, with Stackhousia (Creamy Candles) in the foreground. While tussock grasses (Themeda triandra) are obvious, they certainly don't dominate. You get the impression this was much closer to a herbfield than a grassland.



Truganina Cemetery, spring 1979
(Photo: Neville Scarlett)


I thought it would be great to take an up-to-date shot of the same location to see how things have changed over the last 30 years. Truganina Cemetery is well-known as an important grassland remnant (it has two nationally endangered species present) and it is managed by irregular burning, while vehicle access is prohibited. Armed with a tripod and digital camera, I went out a few weeks ago to capture the same shot. Below is my effort to replicate Neville's original shot.

Truganina Cemetery, spring 2012
(Photo: John Morgan)

Two things are obvious. Firstly, and probably most obviously, it is clear that much of the colouring afforded by forbs in 1979 has dramatically declined, indicating their cover and abundance has likely diminished by an order of magnitude or more. All four species that I mentioned as obvious in the original shot are still present, so what we are looking at is a dramatic change in species evenness rather than local extinction. We could visualise these changes by constructing a rank-abundance curve, plotting the percent cover of species from most common to least. This would hint that dominance has shifted completely from a multitude of forb species to one native grass, and that forbs are mostly now at very low cover.

Secondly, I found it really hard to re-take the orginal shot. In this case, I don't think it mattered hugely because the magnitude of the change is so great. But, it got me thinking about how I might go about doing this better next time. Using old photos to assess environmental change is becoming increasingly popular in environmental science. Part of this, I suspect, is that old photographs are now being scanned and stored in archives such as the excellent one called TROVE (curated by the National Library of Australia). Hence, how might one go about doing this accurately?

The field is called "repeat photography" or "re-photography". It seems that a pretty common method is indeed just walking out to the same spot and taking the same photo. But then you want to match the conditions as well as possible. There's a lot going on with seasonal variation and time of day in terms of shadow length. Visual parralax is the key to the basic technique; you can get pretty close to the same location by comparing rocks and details that are partially obscured by objects in the foreground... but this does not give you exactly the same picture, just something that is clearly the same viewpoint of the same landscape. Part of the problem here is that an oblique perspective creates a continuously varying scale and this makes it hard to link landscape features in the photograph to absolute spatial co-ordinates.

A good starter is Methods in Re-photography: A Research Method for Amateur Researchers. This has some really excellent background on the scientific relevance and utility of the technique, and a very quick'n'dirty guide to the fundamentals. Apparently the bible on the method is called Repeat Photography: Methods and Applications in the Natural Sciences,edited by Webb et al..

It is pretty clear that matching photos can be a powerful means of comparing local changes in vegetation, particularly when photographs extend well back in time. Aerial photographs are great for the same reason, but most don't extend beyond the 1940s in Australia, so repeat photography has great merit to address small-scale, long-term change. I'll leave you with the nice example below by Roush et al. (2007) that compares the alpine treeline ecotone in Glacier National Park in 1930 with that in 2003. Note the massive invasion of trees across the entire landscape!

 

Roush, W., Munroe, J.S. and Fagre, D.B. (2007) Development of a spatial analysis method using ground-based repeat photography to detect changes in the alpine treeline ecotone, Glacier National Park, Montana, U.S.A. Arctic, Antarctic and Alpine Research 39, 297-308.