Biodiversity publications archive

Biodiversity and Fire: The effects and effectiveness of fire management

Proceedings of the conference held 8-9 October 1994, Footscray, Melbourne
Biodiversity Series, Paper No. 8

Biodiversity Unit
Department of the Environment, Sport and Territories, 1996

Discussion: Papers 1-3

Phil Cheney

Q: You posed the question about fuel discontinuity and the joining up of various fuel elements as the fire intensity increased and pointed out the difficulty in knowing when that happens and how to measure fuel in that situation. I was thinking we have very many different fuel types and sorts of discontinuities, let alone those around Sydney. How do we find out empirically when this sort of condition occurs – does it have implications for the way we do fire research and the way fire research is funded?

Phil Cheney: I'll leave the funding question out! I think the implication is that it could be well nigh impossible to define mathematically the fuel characteristics that are driving the fuel front. We can take some good guesses at it and make some sub-divisions, but to be honest it may be well nigh impossible. And so we end up identifying a fuel type and saying; "OK; this is a uniform fuel type, it grows in a certain way, it has certain characteristics" and the fire is going to then spread in a certain pattern in that and determine that pattern. And I think that the effect on prescribed burning; whereas if fuel load is not directly related to fire spread, what it is related to, maybe – and it could be a surrogate for a whole lot of other things – that are fuel continuity, and as fuels grow up after fire, the height of the fuel bed, height of the fine fuels, greenness of the fine fuels, all of which may be represented by the fuel load of the leaf litter; they are co-related. But fuel load itself may not be a fire spread determinant.

Q: Why I added on the costing was that, if we've got to determine the conditions under which that fire/fuel continuity is established by empirical means, it's going to be very costly, because we have got to get out there whenever there's a severe fire and work it out, what those conditions are. We can't just do an experiment.

Phil Cheney: I think we have to do sufficient experiments to enhance our understanding. I mean we could take the American approach and define fuel types and fudge a number of factors so that our predictions appear to go about how they spread, and I guess to an extent we do that and we fudge a number of factors and say "OK.; you don't really know how it goes but if you change this, it appears ... ". That can lead us to some pretty gross errors in extrapolation at times, and it has done.

Q: Is there any integration of experience relating the intensity of the control burn to the fire phenomena afterwards as a wildfire? And, that 6mm fuel size, how good is that?

Phil Chaney: Six mm in fact turned out to be pretty good; when we look at the time it takes to burn out different dimensions of fuel particles and at the distribution of particle size, then 6mm for forest is a fairly practical cut out. In fact its going to be smaller, say for grasslands and shrublands ; it might be 2-3mm – larger fuels will burn out behind the flame front. But I don't think I've answered your question ...

Q: It was about prescribed burning; if people burn more intensely – do they do a better job?

Phil Cheney: The thing that burning more intensely does in most cases, is remove more bark from the trees. This affects the spotting process and I think we probably have to pay more care to prescriptions with low intensity fires to select those conditions where, if bark removal is an aim of the manager, then he writes his prescriptions to ensure he burns under conditions when the bark burns. Now that doesn't have to be high intensity fire, and if he is smart, he won't make it a high intensity fire because it will be very difficult to control for the very reason of the factor he's trying to get rid of. The large fuel component doesn't matter; farmers in Western Australia wanted white ash burns to take out just about everything. But looking at the size classes there is no reason for that unless one is dealing with vast amounts of slash fuels; then you might want to up the size class that you want to remove.

Q: I liked the discontinuous fuel example you gave of the birch and spruce forests in North America, but to me it illustrated just how little ground fuel you need to support a running crown fire. What are your comments on that?

Phil Cheney: You have to look at not only the spatial discontinuity that we have, but the vertical discontinuity – which occurs in our forests. In North American wet and cold forests there is a very compact ground litter understorey, but then a ladder of fuel (lichen, hanging branches) along the trunk of the trees leading into the canopy. Hence only a very mild fire there stays on the ground. In Australia, the self pruning habit of most eucalypts results in a greater fuel discontinuity between surface fuels and the crowns, so you need a more intense fire at ground level to preheat the canopy and cause a crown fire. In pine plantations we are trying to work out prescriptions with respect to the density of planting and height of the pruning to prevent crown fires. I think the important point is that when you consider a prescribed burning prescription it has to be related to that particular fuel type. You can't make the same assumption or prescription for every fuel type.

Kevin Tolhurst

Q: With wiregrass, did you take into account the amount of dead material in the biomass, after burning, as it accumulated?

Kevin Tolhurst: That is taken into account in the hazard rating. But what I've done here, to try to put it on to a time scale in fact is projecting the growth into a rating, so there's an age component that isn't taken into account there and certainly wire grass which is old has more dead material and is a higher fire hazard than younger wire grass of same height. So, no, I did not fully account for this although it is certainly incorporated into the hazard guide.

Q: Firstly, I would like to challenge your statement that the Wombat research can be extrapolated to other lowland mixed species forests. In our area (near Kinglake National Park) the same forest type has superficially the same sparse shrub layer before fire as in Wombat. However, following a fire, as also observed by Dianne Simmons nearby, the shrub layer regenerated very quickly, in our area within only two years to form dense masses of shrub layer with pea and wattle species post fire. It then senesced, but at ten years is still far more than before the fire. Thus shrub fuels were not decreased by fire for ten years, they were increased. Before generalising, you should consider the shrub species present locally. Secondly, your research has not yet quantified the degree to which bark or shrub fuel adds to fire risk. If you did this you might observe a more gradual increase over time rather than the apparent jumps shown in some of your graphs.

Kevin Tolhurst: The second question first – on the quantification of the fuel; the system that we're fitting it into – the bark hazard and elevated fuel guides – have discrete jumps, its not continuous. The problem is not so much of quantifying the fuel so much as putting it into a category. The transition between categories is not a sharp line. I would concede that where it goes 4, 4, 4, it would perhaps start at 2 and move through to 8 ... but still the demonstration is still reasonably valid.

With regard to extrapolating to other sites, it has to be done with a certain amount of intelligence – the shrub layer in the Wombat forest is not very dense, and if you move into other areas where it is, the shrub component is important and I think some of the regeneration responses and rate of growth responses observed in Wombat are still relevant but would have to be viewed in context. What I said in the abstract is that Wombat is relevant to a number of foothill forest in Victoria – 20-25% of Vic Forests, not all forests in Victoria.

Q: How does your work on stringybark forests extrapolate to gum forests. Do the conclusions you reach stand up to the empirical evidence of the effectiveness of fuel reduction burning in controlling wildfires?

Kevin Tolhurst: The reason for concentrating on stringybark here is that the forest in fact had 60% stringybark; the rest are gum species. In the hazard guide, gum barks are a lesser hazard; so I was taking the worst case by looking at stringybarks. There is no problem in incorporating gum if there is only gum there. There is no point in incorporating that at this stage as it is not the dominating factor in terms of bark. What was the second question?

Q: In general how do your conclusions relate the effectiveness of fuel reduction burning to the prevention of wildfires?

Kevin Tolhurst: That's not an objective of the study that we are doing. But I think that there's a bit of a myth around, and its slowly being debunked, that fuel reduction burning is there to stop wildfires. One of the main things a fuel reduction burning does is reduce the intensity of a fire, and also its rate of spread, so the size of the fire is in fact reduced. There is a lot of anecdotal evidence of fuel reduction burning reducing the intensity and spread of fire and therefore helping the control of the fire. The only cases you can find of fires being stopped are in one or two year old fuel reduction burning areas; but that shouldn't be the test of fuel reduction burning. the slides I showed you of the fuels eight years after the fuel reduction burning were very different in structure, and fire behaviour and the ability to control fire in those forests would maintained much longer than even eight years, even though the litter level had returned to long-unburnt levels within four years. So the structure had been changed for quite a long period of time. From a fire-fighter's point of view that gives great benefit. But it hasn't been quantified.

Q: I've been doing some work over at Macedon where the big fires in '83 were so hot that there was hardly a log left on the ground. Have you looked at what happened to fuel levels after that extremely hot wildfire?

Kevin Tolhurst: No, my work is really looking at the effects of fuel reduction burning, not of wildfire, so I haven't been involved in looking at the Ash Wednesday areas. Max Coulter, who is here, has collected some data after the Ash Wednesday fires. Fuel has been a little slower to recover because instead of the 4 t/ha after a fuel reduction burning, next to no litter was left after the fire and the leaves were burnt off the trees, so it took longer than the four years. The effect that it has on heavy fuel and shrub fuel and so on, hasn't been quantified.

Dianne Simmons

Q: About the question of the pre-European level of fire ... ethnographic evidence suggests that fire was moderate intensity and frequent, but adapted to the particular environment to which it was being applied ... not a blanket regime, although an early settler suggested a three to four year frequency seemed to be what was going on in dry sclerophyll forests. We must also recognise some of the understories we are looking at now are not pristine; because since European settlement these have not been undergoing the Koorie regime.

Dianne Simmons: At the Round the Bend Co-operative there are very good fire records for the area over 25 years. But it was difficult to relate the community floristics to this fire history ... perhaps because it is already so changed and the community response is not measurable any more. It was much more of a random effect.

With regard to pre-European fire regimes I am not sure we can know what they were. Instead of trying to mimic pre-European regimes, we should think about what our current objectives for the flora and fauna are and how we can use fire management to achieve these; that we sit down and make decisions on the basis of what we think it might have been, what we would like it to be, what's appropriate for species now. We should manage to a preferred condition rather than mimicking any past regime.

Q: How did you sample for bare ground, particularly after the moderate intensity fire. I would suspect that these forests would have a lot of patches of bare ground naturally. Did you decide to look at the patch size?

Dianne Simmons: I was going to talk about some of the design constraints and problems in a lot of the fire work. We sampled randomly selected points along long transects and then made estimates of bare ground in the quadrats we were sampling. It is a great problem of design that in a patchy environment, then how do you measure some component which has great patchiness; Malcolm Gill's recent paper discusses that problem to some extent. But we have other data from a range of sites; this was just one small part of the data set we have been working on. In many of these forests in long unburnt sites there was not much bare ground, but ground covered by litter and I think the moderate intensity site did stand out as having more bare ground, not just from the low intensity site but from the surrounding areas in general.

Q: I'm interested in your sampling design. In terms of your quadrats – what size were they and what other things did you measure e.g. slope, aspect , soil type etc?

Dianne Simmons: With sites like this, where it is opportunistic - it is a follow on from when two fires occurred of different intensities – then you have to be careful about what you say as it could be a reflection of past site history. But we are slotting it into information from a range of other sites where we've done the burning and they were all low intensity burn sites. We are looking at the moderate intensity burn as a variation on the other data sets that we have on low intensity burns. In terms of picking the actual quadrat size and what things to measure: our quadrat size was picked as a reflection of the species area curve which tuned out to be 30 square meters (not 30 metres square). We used long transects and picked out, from random number tables, sites along it to place those quadrats and measured species presence and absence, and their cover on a broad log case scale, and also used this to measure the litter cover and bare ground. The litter load was measured using 1/2 x 1/2 metre quadrats, again along a transect in the centre where we were doing the floristic survey. For the particular study described, all sample sites were on a NW aspect.

I accept that there are many problems of design. There is a lot of emphasis these days on statistically valid design, and no-one can argue with that. But one of the problems of this great emphasis is that many people now stratify their sampling to the point where we often get statistical validity but not necessarily ecological validity e.g by burning a larger number of sites that are too small. We have to be statistically valid but we sometimes get so carried away that we do some ecologically invalid things.

Q: Do the species area curves change with low intensity fire versus moderate intensity?

Dianne Simmons: I have become quite interested in comparing species area curves of sites with different histories. I think there is probably some very interesting stories in doing that. It is my feeling that species area curves will change with fire history and this is consistent with other results in the literature.