macroresilience

resilience, not stability

Forest Fire Suppression and Macroeconomic Stabilisation

with 24 comments

In an earlier post, I compared Minsky’s Financial Instability Hypothesis with Buzz Holling’s work on ecological resilience and briefly touched upon the consequences of wildfire suppression as an example of the resilience-stability tradeoff. This post expands upon the lessons we can learn from the history of fire suppression and its impact on the forest ecosystem in the United States and draws some parallels between the theory and history of forest fire management and macroeconomic management.

Origins of Stabilisation as the Primary Policy Objective and Initial Ease of Implementation

The impetus for both fire suppression and macroeconomic stabilisation came from a crisis. In economics, this crisis was the Great Depression which highlighted the need for stabilising fiscal and monetary policy during a crisis. Out of all the initiatives, the most crucial from a systems viewpoint was the expansion of lender-of-last-resort operations and bank bailouts which tried to eliminate all disturbances at their source. In Minsky’s words: “The need for lender-of-Iast-resort operations will often occur before income falls steeply and before the well nigh automatic income and financial stabilizing effects of Big Government come into play.” (Stabilizing an Unstable Economy pg 46)

SImilarly, the battle for complete fire suppression was won after the Great Idaho Fires of 1910. “The Great Idaho Fires of August 1910 were a defining event for fire policy and management, indeed for the policy and management of all natural resources in the United States. Often called the Big Blowup, the complex of fires consumed 3 million acres of valuable timber in northern Idaho and western Montana…..The battle cry of foresters and philosophers that year was simple and compelling: fires are evil, and they must be banished from the earth. The federal Weeks Act, which had been stalled in Congress for years, passed in February 1911. This law drastically expanded the Forest Service and established cooperative federal-state programs in fire control. It marked the beginning of federal fire-suppression efforts and effectively brought an end to light burning practices across most of the country. The prompt suppression of wildland fires by government agencies became a national paradigm and a national policy” (Sara Jensen and Guy McPherson). In 1935, the Forest Service implemented the ‘10 AM policy’, a goal to extinguish every new fire by 10 AM the day after it was reported.

In both cases, the trauma of a catastrophic disaster triggered a new policy that would try to stamp out all disturbances at the source, no matter how small. This policy also had the benefit of initially being easy to implement and cheap. In the case of wildfires, “the 10 am policy, which guided Forest Service wildfire suppression until the mid 1970s, made sense in the short term, as wildfires are much easier and cheaper to suppress when they are small. Consider that, on average, 98.9% of wildfires on public land in the US are suppressed before they exceed 120 ha, but fires larger than that account for 97.5% of all suppression costs” (Donovan and Brown). As Minsky notes, macroeconomic stability was helped significantly by the deleveraged nature of the American economy from the end of WW2 till the 1960s. Even in interventions by the Federal Reserve in the late 60s and 70s, the amount of resources needed to shore up the system was limited.

Consequences of Stabilisation

Wildfire suppression in forests that are otherwise adapted to regular, low-intensity fires (e.g. understory fire regimes) causes the forest to become more fragile and susceptible to a catastrophic fire. As Holling and Meffe note, “fire suppression in systems that would frequently experience low-intensity fires results in the systems becoming severely affected by the huge fires that finally erupt; that is, the systems are not resilient to the major fires that occur with large fuel loads and may fundamentally change state after the fire”. This increased fragility arises from a few distinct patterns and mechanisms:

Increased Fuel Load: Just like channelisation of a river results in increased silt load within the river banks, the absence of fires leads to a fuel buildup thus making the eventual fire that much more severe. In Minskyian terms, this is analogous to the buildup of leverage and ‘Ponzi finance’ within the economic system.

Change in Species Composition: Species compositions inevitably shift towards less fire resistant trees when fires are suppressed (Allen et al 2002). In an economic system, it is not simply that ‘Ponzi finance’ players thrive but that more prudently financed actors get outcompeted in the cycle. This has critical implications for the ability of the system to recover after the fire. This is an important problem in the financial sector where as Richard Fisher observed, “more prudent and better-managed banks have been denied the market share that would have been theirs if mismanaged big banks had been allowed to go out of business”.

Reduction in Diversity: As I mentioned here, “In an environment free of disturbances, diversity of competing strategies must reduce dramatically as the optimal strategy will outcompete all others. In fact, disturbances are a key reason why competitive exclusion is rarely observed in ecosystems”. Contrary to popular opinion, the post-disturbance environment is incredibly productive and diverse. Even after a fire as severe as the Yellowstone fires of 1988, the regeneration of the system was swift and effective as the ecosystem was historically adapted to such severe fires.

Increased Connectivity: This is the least appreciated impact of eliminating all disturbances in a complex adaptive system. Disturbances perform a critical role by breaking connections within a network. Frequent forest fires result in a “patchy” modularised forest where no one fire can cause catastrophic damage. As Thomas Bonnicksen notes: “Fire seldom spread over vast areas in historic forests because meadows, and patches of young trees and open patches of old trees were difficult to burn and forced fires to drop to the ground…..Unlike the popular idealized image of historic forests, which depicts old trees spread like a blanket over the landscape, a real historic forest was patchy. It looked more like a quilt than a blanket. It was a mosaic of patches. Each patch consisted of a group of trees of about the same age, some young patches, some old patches, or meadows depending on how many years passed since fire created a new opening where they could grow. The variety of patches in historic forests helped to contain hot fires. Most patches of young trees, and old trees with little underneath did not burn well and served as firebreaks. Still, chance led to fires skipping some patches. So, fuel built up and the next fire burned a few of them while doing little harm to the rest of the forest”. Suppressing forest fires converts the forest into one connected whole, at risk of complete destruction from the eventual fire that cannot be suppressed.

In the absence of disturbances, connectivity builds up within the network, both within and between scales. Increased within-scale connectivity increases the severity but between-scale connectivity increases the probability of a disturbance at a lower level propagating up to higher levels and causing systemic collapse. Fire suppression in forests adapted to frequent undergrowth fires can cause an accumulation of ladder fuels which connect the undergrowth to the crown of the forest. The eventual undergrowth ignition then risks a crown fire by a process known as “torching”. Unlike understory fires, crown fires can spread across firebreaks such as rivers by a process known as “spotting” where the wind carries burning embers through the air – the fire can spread in this manner even without direct connectivity. Such fires can easily cause systemic collapse and a state from which natural forces cannot regenerate the forest. In this manner, stabilisation can cause changes which cause a fundamental change in the nature of the system rather than simply an increased severity of disturbances. For example, “extensive stand-replacing fires are in many cases resulting in “type conversions” from ponderosa pine forest to other physiognomic types (for example, grassland or shrubland) that may be persistent for centuries or perhaps even millennia” (Allen 2007).

Long-Run Increase in Cost of Stabilisation and Area Burned: The initial low cost of suppression is short-lived and the cumulative effect of the fragilisation of the system has led to rapidly increasing costs of wildfire suppression and levels of area burned in the last three decades (Donovan and Brown 2007).

Dilemmas in the Management of a Stabilised System

In my post on river flood management, I claimed that managing a stabilised and fragile system is “akin to choosing between the frying pan and the fire”. This has been the case in many forests around the United States for the last few decades and is the condition into which the economies of the developed world are heading into. Once the forest ecosystem has become fragile, the resultant large fire exacerbates the problem thus triggering a vicious cycle. As Thomas Bonnicksen observed, “monster fires create even bigger monsters. Huge blocks of seedlings that grow on burned areas become older and thicker at the same time. When it burns again, fire spreads farther and creates an even bigger block of fuel for the next fire. This cycle of monster fires has begun”. The system enters an “unending cycle of monster fires and blackened landscapes”.

Minsky of course understood this end-state very well: “The success of a high-private-investment strategy depends upon the continued growth of relative needs to validate private investment. It also requires that policy be directed to maintain and increase the quasi-rents earned by capital – i.e.,rentier and entrepreneurial income. But such high and increasing quasi-rents are particularly conducive to speculation, especially as these profits are presumably guaranteed by policy. The result is experimentation with liability structures that not only hypothecate increasing proportions of cash receipts but that also depend upon continuous refinancing of asset positions. A high-investment, high-profit strategy for full employment – even with the underpinning of an active fiscal policy and an aware Federal Reserve system – leads to an increasingly unstable financial system, and an increasingly unstable economic performance. Within a short span of time, the policy problem cycles among preventing a deep depression, getting a stagnant economy moving again, reining in an inflation, and offsetting a credit squeeze or crunch….As high investment and high profits depend upon and induce speculation with respect to liability structures, the expansions become increasingly difficult to control; the choice seems to become whether to accomodate to an increasing inflation or to induce a debt-deflation process that can lead to a serious depression”. (John Maynard Keynes pg163–164)

The evolution of the system means that turning back the clock to a previous era of stability is not an option. As Minsky observed in the context of our financial system, “the apparent stability and robustness of the financial system of the 1950s and early 1960s can now be viewed as an accident of history, which was due to the financial residue of World War 2 following fast upon a great depression”. Re-regulation is not enough because it cannot undo the damage done by decades of financial “innovation” in a manner that does not risk systemic collapse.

At the same time, simply allowing an excessively stabilised system to burn itself out is a recipe for disaster. For example, on the role that controlled burns could play in restoring America’s forests to a resilient state, Thomas Bonnicksen observed: “Prescribed fire would come closer than any tool toward mimicking the effects of the historic Indian and lightning fires that shaped most of America’s native forests. However, there are good reasons why it is declining in use rather than expanding. Most importantly, the fuel problem is so severe that we can no longer depend on prescribed fire to repair the damage caused by over a century of fire exclusion. Prescribed fire is ineffective and unsafe in such forests. It is ineffective because any fire that is hot enough to kill trees over three inches in diameter, which is too small to eliminate most fire hazards, has a high probability of becoming uncontrollable”. The same logic applies to a fragile economic system.

Update: corrected date of Idaho fires from 2010 to 1910 in para 3 thanks to Dean.

Bookmark and Share

Written by Ashwin Parameswaran

June 8th, 2011 at 11:35 am

24 Responses to 'Forest Fire Suppression and Macroeconomic Stabilisation'

Subscribe to comments with RSS or TrackBack to 'Forest Fire Suppression and Macroeconomic Stabilisation'.

  1. Looking forward to reading your next post. Thank you for such an interesting and readable article.

    Francisco

    8 Jun 11 at 3:25 pm

  2. Hey – fantastic blog! Brilliant to see someone else addressing things from the 10,000 foot POV.

    Regarding monocultures though, which is what you are talking about – I think they are unfairly despised.

    All life strives towards monoculture (and so does physics – gravity) and in many cases it is successful. A successful monoculture (like vertebrates or monotheism or money) blends into the background so we forget its there.

    Economic evolution of the human race must likewise tend towards monoculture, and indeed it has done. If the current monoculture fails then presumably it would be supplanted by one better adapted.

    Here’s my own attempt:

    http://liminalhack.wordpress.com/2011/01/10/monoculture/

    scepticus

    9 Jun 11 at 3:13 pm

  3. [...] Please see this blog: macroresilience, and in particular the first post (overlong though it is for a first post!), and most recent post. [...]

  4. Please correct date errors in paragraph three—-off by 100 years.

    dean

    9 Jun 11 at 4:17 pm

  5. Scepticus – thanks! Your post is interesting and I agree with a lot of it. It’s a topic I have a very long unfinished post on. But let me try explaining my views in a few lines.

    I agree that the existence of idealised diversity in the real world is exaggerated. Most adaptive systems have a dominant homogeneous strategy that works best. Whether that system is resilient or not depends on to what extent this dominant “monoculture” is subject to constant threat from new entrants on the fringe. The system is fragile when upon its collapse, there is no alternative to take up the slack. Hence my obsession with maintaining entry of new firms into currently cronyist industries (especially in my posts in the category cronyism on the side).

    I’ll hopefully write a post on this topic of diversity in real-world complex adaptive systems very soon.

    Ashwin

    9 Jun 11 at 4:25 pm

  6. Dean – done. Thank you!

    Ashwin

    9 Jun 11 at 4:38 pm

  7. Ashwin, how would you categorise the rise of google, apple, facebook et al if not new entrants versus the incumbents of wintel?

    The shadow banking system was I suggest an attempt to challenge the hegemony of central banks, but it failed.

    The next attempt may not fail, but is probably unlikely to be enacted by the same actors.

    However the attempt, and the resulting brushfire may lay fertile ground for the next try.

    scepticus

    9 Jun 11 at 5:45 pm

  8. Scepticus – I don’t argue that the tech industry has any real cronyism problem. Monocultures are fleeting and constantly under threat. But this is not the case in finance and healthcare certainly and probably in a few other industries I don’t have personal knowledge of.

    More than 30% of American workers need a license compared to 5% in the 50s and 13% in the UK. The trend even in the UK is towards more entry barriers, not less. http://www.economist.com/node/18678963

    In banking for example, we’ve had the same investment and commercial banks for the last 100 years – many of them would have gone bankrupt long ago if it was not for the persistent support provided to them by the central bank. The last 30 years especially has seen creditor bailouts post the LatAm crisis in the early 80s, Mexican crisis in mid 90s, LTCM and finally the current crisis. No wonder we have a fragile cronyist financial system!

    My fear is that the next brushfire turns into a crown fire and a systemic collapse. Although ultimately a collapse may not be a bad thing from the perspective of systemic change.

    Ashwin

    10 Jun 11 at 4:03 am

  9. Ashwin, I think the basic problem with the finance sector is that without some kind of backstop, and in normal economic times, maturity transformation just isn’t that attractive a business proposition in the long term.

    If its positioned as a short term business opportunity only, then it just attracts spivs and chancers.

    Unfortunately we don’t have the institutions that are needed in order to recognise this fact and manage things accordingly. Indeed, the one criticism I have of your work here is that it seems to imply that the regulator is outside the system being ‘stabilised’. The whole notion of environmental stabilisation relies on an outside actor considered separate from the environment. It would be more useful I think to consider the entire system, regulators, bankers, depositors, government and all, as a single evolving entity, or perhaps some kind of petri dish culture.

    If you do that then I think it changes the narrative considerably, and the red flags end up getting planted in different places.

    The evolution of any phenotype is governed by external constraints and selection events, but also by internal constraints such as gene regulation, biological modularity, laws of physical form and so on. Therefore I see the stabilisation approaches of regulators as more of a gene regulation type function than one of environmental manipulation, which like I said, changes the narrative considerably.

    scepticus

    11 Jun 11 at 3:47 am

  10. Scepticus – on maturity transformation, I wrote a post a while ago http://www.macroresilience.com/2010/10/21/questioning-the-benefits-of-maturity-transformation/ on why we no longer need it. Short story is that we have a sufficient supply of capital willing to lend/commit for long periods (pension funds, insurers) – banks transforming deposits into long-term loans actually reduce the supply of loans available for PFs and insurers, apart from introducing instability.

    Your point on making regulators endogenous is fair – in my posts on cronyism, I have made the point on why regulators stabilise drawing on Mancur Olson’s work. And I have a rough framework explaining post-WW2 evolution of our economy with everything endogenous but still unfinished and very very long! Hopefully will finish it soon. But Minsky’s books capture a lot of the story.

    Ashwin

    11 Jun 11 at 4:58 am

  11. I read the post on maturity transformation and I agree with the premise (and that holding up the spectre of long term lending collapse is scaremongering). However there is another important aspect to the story I don’t think you touched on. A liquid long term security is not really long term lending from the lender/saver’s perspective.

    Thus treasury bonds that purport to be a 25 year deal but in fact change hands every 3 months don’t really qualify as long term lending. You cite Kocherlakota in some of your posts – perhaps you have seen his ‘The Societal Benefit of nominal bonds’ on this subject?

    So I think we have seen a market led short circuiting of long duration lending in excess of the actual need for it, which I think is empirical evidence in favour of your proposition regarding maturity transformation. Both securitisation and QE are examples of the removal of essentially ‘fictional’ long term lending, and casts both of these much maligned recent developments in a more positive ‘evolutionary’ light.

    scepticus

    11 Jun 11 at 5:27 am

  12. Fine summary article of fire suppression problems. Being Idahoan myself and having listened to F.S. presentations on this subject and having seen for myself some few remains of the old tribally managed forests (the “American Indians” used to burn for game starting in mid summer), I would say the old way was preferable from the horticulturist viewpoint. However, if you look at Charlie Russell paintings of western Montana, you see a preponderance of smoke filled skies. The old timers I knew said that such was indeed how they looked before WWI–all summer long. Not sure my lungs could handle that any more.

    There is an even more macro view to consider: if climate change study has revealed anything, it is that the earth’s crust has been sequestering carbon for a very long time in the form of fossil fuels. So much so that the CO2 content of the atmosphere has dropped to an historic low–until very recently with the advent of the industrial revolution. CO2 is not so low that plant life is being asphyxiated but low enough to seriously reduce the total amount and possibly change the fire dynamics. Maybe less conifer is needed and more grass and brush for optimal plant conditions given the reduced amount of CO2. Who knows? Arguments cut both ways.

    The northern boreal and equatorial rain forests are widely reported to be net O2/CO2 neutral. That leaves mainly the oceans as source/sink for CO2. Life processes in the oceans are also reported to be net O2/CO2 neutral which leaves only mean temperature of the oceans as the other governing factor for changing atmospheric CO2. But temperature is predominately a function of nuclear processes of the sun and the earth’s core. Only one thing is certain of complex dynamic systems in nature: that there must be operative at any one time a net negative feedback mechanism or the system couldn’t exist and no human could exist within it to say so. It seems to me, as an engineer, that the first object of any study of such systems would be to discover the stabilizing mechanism/s– as many if not most climatologists seem to be doing in spite of the media furor over “climate change”, “eco fragility” and the subsequent politicizing of the science and the positive feedback mechanism of government grant monies.

    Obviously, neither human economic systems nor human political ones need operate the way physical ones do. Humans have the peculiar habit of trying to specify physical/social/cultural realities that don’t exit, may exist, or have no possibility of existing. Thus “economic” systems may be specified by human fiat and may operate without benefit of negative feedback at least for a relatively brief and exciting time. Thankfully, the physical order is not so arbitrary.

    The only reason I can see why some people would ever attempt to establish an economic system by fiat would be to create advantageous pricing mechanisms benefiting those with the power to establish some feature of the system. Any stable “natural” economic system would have negative feedback mechanisms in place to keep it stable. Therefore, if any one set of economic actors desires to alter the system to their advantage without submitting to the existing stabilizing feedback (usually viewed as troublesome and “regressive”), they must set up positive feedbacks (said to be “progressive”), such as central banking, government debt, credit expansions (and other Ponzi) where surpluses can be siphoned off in defeat of credit worthiness and other negative feedback implicit in auction interest rates and commodity pricing. Small perturbations wouldn’t suffice for the “progressives” for any length of time if the system were stable/adaptive enough and too large perturbations could destroy it. I think this conclusion is in keeping with that of Macro Res.

    Even Macro Res attempts to specify ideal or at least better economies. Blessedly, however, the site also tries to compare natural dynamic systems with human ones such that “better” might at least be possible. Three thumbs up and all that!! Love your thought processes.

    mccat

    24 Jun 11 at 1:45 pm

  13. Someone is channeling Ashwin: The following comment, posted at Chrismartenson.com on June 10, 2010, was written after you had written your earlier posts comparing financial instability with ecological resilience, but before you had written this post specifically on forest fires:

    Keynesian vs. Austrian economics made simple – Good analogy?

    If you have been confused about the differences between the Keynesian and Austrian schools of economics, may I humbly propose an analogy that may help your understanding. If you don’t think this is accurate or helpful then please let me know. I’m new to this myself.

    Fire policy — In the early 1900s the US Forest Service implemented a new policy to aggressively suppress all forest fires. Citizens didn’t like to see trees burn and the lumber companies saw profits going up in smoke. Over the years they obtained equipment, trained men, and worked out elaborate plans. With a lot of time and money they reduced fire damage considerably.

    Over the years the forest management experts began to notice problems. Without periodic small fires the undergrowth became thick. It ignited easier, acted as additional fuel making the fire hotter, and sent it up into the crowns of the mature trees. This spread the fire faster and farther, killing more of the forest.

    With further study they found that in the pasts forest experienced periodic small fires which cleared the undergrowth and burned up the deadwood. They usually did not reach the crowns and burned themselves out without doing great damage. They triggered the tree seeds to sprout and start a healthy new generation. The mature trees also grew better. The old policy of trying to manage nature was counter productive. The policy was changed to allow the natural cycle to return.

    Keynesian school – All recessions are bad and must be suppressed by government actions. This protects established businesses and jobs. The methods are elaborate and costly, but a benefit to the public overall.

    Austrian school – When markets stray too far from reality they must be purged by adversity. This clears unneeded or failing enterprises so capital is not allocated wastefully, and new businesses can emerge. Periodic small recessions are the price of a healthy economy.

    Recent situation — Unfortunately, after decades of total suppression many forests were overgrown tinder boxes. At Yellowstone National park the “hands off” policy let a fire get completely out of control before they used aggressive suppression. 36% of the park burned and much of this area is still black 20 years later. At Los Alamos a “controlled burn” to clear undergrowth ran away from them and destroyed 400 houses.

    Yellowstone – Bear Stearns, Lehman Brothers, current economic predicament.

    Los Alamos – Recognizing the problem and the perils of trying to make up for past mistakes.

    I have often found a good analogy helpful to understand or explain a concept. It is easy to push one too far, but if used properly it can be a good tool. Is this one accurate or helpful? What do you think?

    http://www.chrismartenson.com/forum/keynesian-vs-austrian-economics-made-simple-good-analogy/40845

    Michael Strong

    4 Oct 11 at 11:49 am

  14. Michael – Thanks. To be honest, I’m surprised that more people don’t make this analogy. And it’s not just an analogy – the underlying drivers that generate the stability-resilience tradeoff are found in many other complex adaptive systems. And they’re not just restricted to ecological systems – electricity networks, geological systems and of course many human systems – pest control in agriculture, medical drug usage. I could go on and on.

    There’s not even a shortage of mathematical models – just that they don’t quite provide us with the neat and clean analytical answers that equilibrium models do.

    Ashwin

    4 Oct 11 at 1:15 pm

  15. [...] entire long arc of post-WW2 macroeconomic policy in the developed world can be described as a flawed exercise in macroeconomic stabilisation. But there is no better example of this principle than the Euro [...]

  16. [...] paradoxically lead to decreased resilience in a complex system. Ashwin mentions the example of how forest fire suppression has led to catastrophic fire danger and less fire resilient forests. Stabilization of a financial [...]

  17. [...] to increased systemic fragility. But there is a lot more to it. In comparing economic crises to forest fires and river floods, I have highlighted the common patterns to the process of system fragilisation [...]

  18. Excellent comparison. As a believer in Austrian economics and former Wildland Firefighter, it was particularly interesting. People have no idea how much worse fires are now, than the were historically because we have stopped nature from doing it natural cleansing process. Environmental, and economic intervention need to stop immediately.

    Bob

    18 Dec 11 at 2:54 am

  19. Bob – Thanks. As I mentioned in a comment above, the patterns recur in many other complex systems. As I tried to analyse in this post http://www.macroresilience.com/2011/12/14/the-pathology-of-stabilisation-in-complex-adaptive-systems/ , the roots of this tendency to stabilise are very deep.

    Ashwin

    18 Dec 11 at 12:50 pm

  20. [...] esposizione al rischio di incendi di portata devastante delle foreste statunitensi è fornito qui, insieme ad un interessante parallelo con lo sviluppo delle politiche macroeconomiche di [...]

  21. Sometimes I think that (e.g. monetary) policy should be partly stochastic: if you base policy decisions on random numbers, with certain parameters varying over various locations, over time, in a random way. The decision would set an average, and a spread, but a random process would make the final decision. That would force businesses and banks to prepare for a certain expected bandwidth of uncertainty. The system would be more resilient.

    .. and then if a crisis hits, you could reduce the randomness of your policy, and use that as a tool to stabilise the situation. In that way, you would deliberately introduce a small inefficiency into the system, to keep it strong, and to allow the possibility to reduce risk at a point in time when you need it.

    Nichol

    17 Jun 12 at 1:54 pm

  22. [...] spark might be catastrophic. Stability breeds instability. (See e.g. Parameswaran here and here. Also, David Merkel.) We must deliberately set financial forest fires to prevent accumulations of [...]

  23. [...] spark would be catastrophic. Stability breeds instability. (See e.g. Parameswaran here and here. Also, David Merkel.) We must deliberately set financial forest fires to prevent accumulations of [...]

  24. A new method of fire suppression is currently under development. It is currently in the applied research phase of development. We can control the flames of a fire with electromagnetic signals only.

    Control includes both fire extinguishment and flame manipulation. The bending of flames by electric fields was first demonstrated in the 1870′s. Manipulation of flames could permit the creation of “escape corridors” in flame-filled environments, and achieve the spatial localization / confinement required to prevent spreading of fire to other combustible materials.

    Control the flames to control the fire.

    FRANCIS YEE

    17 Nov 12 at 2:21 am

Leave a Reply