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 11-16

Barbara Wilson

Q: I'm wondering if you could comment on the advisability of conducting a fire management regime for one species that may be endangered regionally, even nationally endangered, when that fire management may in fact, be to the detriment of many other species.

Barbara Wilson: My approach to that is to sit down and work out a decision pathway. Particularly if you have critical habitat for a particular species, then you have to take that on board, especially if it is not going to eliminate other species from the area – I understand [that is the case here]. But you must have a decision pathway.

Q: You showed earlier an overhead which showed the '83 fire and said it wasn't stopped by some previous fires – can you tell us a bit more about that?

Barbara Wilson: Those two fires in 1980 and 1981 were escape fires which were quite large and hot – I have aerial photographs of them. They burnt right down into gullies and melaleuca swamps. However in that particular case they were not substantial enough to have major effect on the 1983 fires.

Q: What about that area to the north – why wasn't it burnt?

Barbara Wilson: That is private land belonging to International Harvester, last burnt in '78, and the reason it did not burn in 1983 is due to their protection efforts. They have a big slashed break, a road, their own fire-fighting equipment and they stood there and fought the fire and spot fires.

Q: You showed us a graph of succession in your area and didn't really have time to tell us which species was which. I was wondering whether you could go over that again?

Barbara Wilson: The early species were Mus and New Holland mouse. The mid species were A. stuartii, and we put Sminthopsis in that classification because it's all over place – it's very patchy. Late species included the rat species and swamp antechinus Antechninus minimus. This graph was done when we predicted swamp antechinus would only come back at a late stage – it was eliminated after the fires and it only started to recolonise last year in 1993/4. Brown bandicoot have gone down in number of sites and densities – but there might also be predator effects.

Q: Where did the Antechninus minimus come from – if its taken so long to get back?

Barbara Wilson: It is recolonising right on the coastal areas and the sand-dunes, around Aireys Inlet – there's steep cliffs under the lighthouse – it's coming around that coast, and we know that was burnt. It's now got around Urquhart's Bluff and hopefully will recolonize inland up through the gullies.

Hugh Possingham

Q: More a comment than a question. Years ago, before I'd heard of PVA, I used a simple spreadsheet with ground parrots to do the same thing – and came up with much the same answer ... [part missing due to tape change] ... a large number of catchments – wildfire there can be very large, and could easily have created an artefact – in that a wild fire in the situation where the population was highly skewed in age structure, and with that time gap before they start to breed again – it's actually very vulnerable to extinction. It's essentially very similar to your results.

Hugh Possingham: Two points: my answer to having only 2 (?unburnt) patches – that's without having wildfire on the system. So that will be interesting to try that over the next two weeks, and see what happens. On the other point – you stopped me selling my PVA package, my 'Alex' package because, yes, it can all be done on a spread sheet. You can achieve a lot, just with simple thought models on spreadsheets and a lot of the models that I give to people called 'Alex' to do PVA – they're not as sophisticated, as most people might think.

Q: Is there a change in the patch size that goes with the change in the fire-free intervals that you choose? Have you actually seen a co-relation?

Hugh Possingham: No, I haven't looked at that – I haven't run simulations where I vary both the number of patches and the time interval. Does that 18-year time interval give the optimum? – that's the important question. And I will definitely do it – when I do my written paper – that will be in there.

Q: With your initial point of maximising the diversity of fire regimes being put through – it's probably a good interim solution – because that's what many people said this morning. But it also seemed from questions this morning and from the talks, that land managers in general, just wanted a one number solution to a particular ecosystem, be it grasslands: three years, heathland: seven years, forests: 20 years and so on. The term fire regime is essentially being populously slotted into a one number solution. And what you have implicitly said and some of the others have, is that this is essentially the wrong approach.

Hugh Possingham: I agree entirely, I think that most of the other speakers were getting at that and maybe didn't want to say. It's a simple thing to say. We should be erratic and maybe even vaguely irrational about our fire management – we want some places to burn a few times and others not at all. That right, from the management point of view just putting everything, firstly into variable ideas, into the context that you may have to think about probabilities – you might burn this or you may even have contingencies – given what's happened in the past and our observations. That just makes the management problem more difficult, but we should be able to promote that approach.

Q: It's all very well saying we should be somewhat "irrational" in our approach but the fire plans for national parks have the most rigorous burning regimes written in there. As David Cheal pointed out, it's extremely difficult trying to change them.

Hugh Possingham: Intriguingly in SA national parks there is a no burn policy – in every single national park, pretty well – someone may correct me – but there may be a little control burning. There we have the other extreme. A lot of the ecologists are saying "may be you should have some burns – let's have a look". So they have done the first trial fire. Clearly there is variation between agency policy and States – in a way this gives us some erratic variation in fire regimes.

Q: I wondered that, way in the future, when we have enormous data bases we could interrogate, that give us very precise information about all of the species and all of the circumstances in which they were occurring – would we ever go ahead and make decisions? But the reality is that we are dealing with natural systems which react to a whole lot of variables, and we don't really have a great handle on it – and there is a tendency within the human species to take a gardening attitude to things and snip the edges and tidy things up so it looks nice in a human perspective. And may be we should step back a bit and let the thing go – uncontrolled stochastically - evolution.

Hugh Possingham: I agree when there is limited information. But sometimes, although I don't like one species management – it is putting everything you do into one species – nonetheless, there are circumstances where it is worthwhile to study just one or two species carefully and try and conserve them in a particular place – perhaps just for cultural reasons – they're soft and cuddly. Sometimes these are high profile species and one can be open to criticism that we put too much emphasis on them. It's got to be a balance, depending where you are.

Gordon Friend

Q: I think we've canvassed it, in that if you have a positive response after burning, it may be more apparent than real – because the ants may have less area to forage over after burning – because all the litter is gone and so on. If that's true – if that possibility affects other groups – then what you observe is not a change of abundance, but an increased trapping success?

Gordon Friend: I think this is a big problem particularly with pit trapping studies because it does rely on the activity of the animal, that the increased activity is so immediate, that it probably is an increased trappability effect. When you remove [litter] the invertebrates can move around more readily and are more likely to fall into the pits. You're putting a numeric response on it, but they were always there, but they just weren't as trappable. You have to be very careful putting a label on them. The same could apply to spiders. It's one of those inherent problems in methodology.

Q: One of the things that concerns me about the study – and it's common to a lot of forestry research projects is that you haven't actually got any replications of your treatments – you've got sampling replications of 16 sites within the area, and you are using that as pseudo-replication. Do you want to make any comment about the variability from one site to another, regardless of fire history and what effect that might have on your results?

Gordon Friend: As I said the locality effects were very important. Those two grids 8 and 9 were structurally and floristically similar to the other grids, but they were sometimes completely different. To have replication of treatments is essential. I am doing that now with the Kauri and Jarrah forest studies. I have two or three blocks and put at least two grids in each. This is a big problem with all of the studies on top of the fact that they don't have much pre-fire information. So, I agree with that, it is very important.

Q: To continue the sampling discussion ... do you think that the lack of information about the fires causes as much noise as the variation within sites as Kevin was talking about. And if so does that mean we have to stand back a bit from space and time substitution studies?

Gordon Friend: That is what I would argue. I wouldn't bother doing any more of those retrospective studies on invertebrates. You need to look at a site in the long-term and then do an experimental fire. I would want two – three years' good data before fire and take reasonable measurements of the fire. Intensity, weather conditions and every thing can influence that fire – for instance, on grids 1 and 2 it was quite patchy, yet other burns done in the same month were quite different – then along comes this wildfire which was different again, and so we had great variation there – all in one month.

Q: I was interested that you had a wildfire site. In terms of some of the experiments we saw yesterday – how valid are the inferences we are trying to draw from those sorts of sites, compared to the wildfire situation? We heard that the response of some of these groups change with increasing fire frequency in a control burn situation. I was wondering what would happen when you control burn in a place that has had a wildfire, for example as a follow-up for property protection. There seems to be so much variability between your sites – I was wondering just how valid the results coming out from just the control burn sites are?

Gordon Friend: It depends on whether you want to look at the effects of control burns or wildfires. I would have thought intuitively that there would have been a huge difference. I was quite shocked to observe no differences apparent, after that wildfire. It was the same with vertebrates. There were Honey Possums back in the wildfire burnt area after 6 months. And they were back in a totally open area. I was quite surprised to find similar results, but I still think they should still be studied separately, as they are quite different events.

Alan York

Q: My question is about the apparent increase in some species of ants. I was surprised that some were more common only two to four years after burning. Ant colonies take several years to establish from the queen. So the observed ant colonies must have already been on site or nearby prior to the fire. So are the areas already regularly burnt, and adaptive species already established in it – or are they moving in from nearby areas?

Alan York: Any sampling process is fraught with difficulties of knowing whether you have got the whole community or just parts of it. I have done some other studies which support what you said. They were there, but they were in very low numbers – surviving in patches of habitat – and population sizes that are that small are at risk of local extinction. And as the habitat becomes more available, they become more abundant. So yes in a sense, that is part of the question. Most ants actually just go underground during the fire, although there are those that get consumed in the litter and those that nest in trees, that do disappear, and do recolonise. But most of them, and its the same with many other invertebrates – it's not so much recolonising – they survive. The litter is that thick there is some of it, and a lot of them take refuge in that litter and in the soil.

Q: You emphasised the importance of scale in your presentation, and one aspect of that is how quickly can these specialist ants recolonise – there's a space/time relationship I guess – how fast can in fact ants move in? If it's a small area, perhaps they will recolonise the whole quickly, but if its a large area, they may be moving in from an edge – it may take a long time. I just wondered what the situation there is?

Alan York: I made the point that in the frequent burning study, these were just 1 ha patches interspersed in other forest environments. The species lost from the patches were in the surrounding areas, and as far as we could assess, there were no inhibiting factors preventing them from potentially moving back into burnt patches. We did catch alates on the burnt sites, but there were no established colonies, so I came to the conclusion that the habitat was so unsuitable for colonists or "propagules" to establish. The actual rate of colonisation is difficult to determine. If you are dealing with vast burnt areas it could be a long time. But for some ants, it's no problem at all, they can move in quite quickly and termites are the same – you have sudden flushes of alates moving in response to weather conditions. But the actual rate of colonisation – I'm not sure.

Q: Thinking about the time frame predicament – the time frame you are looking at is getting a bit longer than some that others looked at. I ask you to speculate – What do you think about the possibility that there are species that require areas unburnt for 50 years, that may have disappeared completely a long time ago, or may have persisted in some other area – could it be that species that need extremely long periods aren't there?

Alan York: Yes, its always a question that crops up in these studies. We go into an experimental area as it is today, and quite often we are not aware of the historical background. It may be the history of fire is such that is has selected those species which are tolerant to fire – and there may be little patches somewhere – and I know this certainly occurs with other groups of organisms – where there are specialists still living happily in undisturbed areas, whatever undisturbed means, for a long time. So I suspect that's the case, and I think the areas we would be looking at in this case is large decomposing logs in forest systems in the gullies – because they're the things which have probably persisted for very long time periods and to a certain extent are immune to short term disturbances. So I expect that's where we will find the very interesting ones.

Q: I wanted to ask you about your specialist predator group. I noticed that you said they were late in the succession because they were specialists feeding on Collembola. But there are of course Collembola soon after the fire, so that there needs to be some other factor that prevented them coming in sooner.

Alan York: It was useful to go to the level of species or genus with many of these because we have some information on what they do – and that was the patchy information I had – that they feed on Collembola – but what ever else they do I have no idea. Collembola were present right through but may be different species and perhaps they predate on particular groups of Collembola. That may be the information we need to know.

Q: I wonder if you would like to speculate further as to how likely it is that regular fire on a short-term basis, that is very frequent fire, has reacted with fragmentation and habitat loss to give a very significant biodiversity loss, already in these sorts of environments.

Alan York: The Myall Lakes area, which is the first example I gave, is a large extent of forest, relatively undisturbed except for some logging in 1950s – a national park now – but it has a fairly high fire frequency and it is very rich. By contrast the state forest study areas have a long history of disturbance; they were felled in the 50s, have undergone timber stand improvement and are heavily disturbed and frequently burnt. And yet the richness was the same in those – the same groups of ants doing the same things. Either these groups are tolerant, or we have lost all the intolerant groups in the past ... we don't know if there are other groups in there somewhere, that we haven't yet found. It's hard to tell; there's been so few studies done.

Robin Coy

Q: One of the things that strikes me about your results is, that as you only measured one year after in a community that may not be burnt again for another 200 years – how do you see about drawing conclusions about the effects of fire intensity on a couple of patches of forest there compared with the effect at 50 or 100 years? Twelve months is very short compared with the life-span of that community.

Robin Coy: Certainly. This is a brief study of the immediate effects of fire to see which groups may be most seriously affected. Obviously funding precludes a long-term on-going study, and retrospective studies are fraught with problems with site differences. Annual differences are a problem too – so this was a very brief study to look at the immediate impact of fire in one area at Powelltown. I agree with you that I can't make any long term predictions of the effects of fire in this environment. It is a thing which needs to be studied.

Q: The point I want to make is that if you went back to those two areas 50 years after that fire, and tried to pick the differences between those two fire intensities, you would get perhaps quite a different picture – after it had a chance to get those few wobbles over in the early stage of development. You can't talk about those wobbles in the early stages.

Robin Coy: I agree with you that they may actually mean nothing ... but the disappearance of a whole Order – that is an important factor that has to be looked at – even in the short term. You are talking about problems with the shrimps in the system – and surely that's an important food source for some animals. They are also important for litter decomposition, and regulation perhaps of fungal and microbial populations. You are looking at early successional stages which quite often relate to later successional stages. Unfortunately both sites have since been cleared. And that's the problem with the Central Highlands – we don't have the opportunity to go back in 50 years to most sites – there is very little of that Central Highlands district that hasn't been severely altered in some way in the last 20 years.

Q: I want to ask you a bit more about the amphipods. Have you got them identified? How many species were there? The reason I ask is that there is some suggestion, certainly in Tasmania, that some amphipod distributions are affected by salt concentration, and you only get some of them in certain environments.

Robin Coy: Obviously they are all Tartisids(??). I'm still waiting on identifications – I'm sure you know the problems that we face. I believe there are between 14 and 17 species in first years collection on the two sites – however I could well have put males as one group and females as another. They were all moist forest specialists and obviously very different from the tolerant coastal species. That is perhaps why they were so sensitive to the opening and drying of the forest.

Q: It's a bit of a shame that you haven't control data. I'm just wondering – the area you were working in – was that hit hard by the '83/4 drought or not?

Robin Coy: Yes, that was. Obviously we had a drought year, from July, right through to January-February. So may be moisture was already a limiting factor – but it is likely that, even in a drought year, the litter is still moist, unlike agricultural land which suffers far more severely. The water saturation levels and soil moisture conditions don't alter that much below the forest even in a drought year.

Ross Hall

No Questions [no time]