Big holes, Janet Jackson, and fire severity

Article reviewed: Stand-replacing patches within a ‘mixed severity’ fire regime: quantitative characterization using recent fires in a long-established natural fire area

By B.M. Collins and S.L. Stephens. Published in the journal Landscape Ecology and available for download.  

The plot line: This study used an area within Yosemite National Park where wildfires had been allowed to burn over the past ~30 years. They looked for patterns in how often fires created large gaps (holes in the canopy) versus small ones. In other words, they measured how often fires killed a lot of trees versus just a few. They also attempted (with pretty good success) to explain the reasons why some gaps were large (fire more severe) while others were small. They found that, while most gaps were small (less than about 5 acres), there were also a few very large gaps that were created by fire- up to 230 acres! Overall, the portion of the burned areas that actually created gaps (as opposed to the fires remaining on the surface and not killing lots of trees), was about 15% over a 30-year time period. For previous fire occurrence to reduce the chance of another high severity fire occurring, the fire had to occur recently (within about 30years). This is what I call the Janet Jackson effect… “what have you done for me lately?” They conclude that, while the high severity fires that create gaps were not the dominant type of fire behavior that occurred in this case, they had a significant contribution to the mix of fire severity that occurred.  

Relevant quote: “While high-severity fire represents a fairly low proportion of the total burned area (15%) stand replacing patches should be considered an important component shaping these forests.”

Relevance to landowners and stakeholders:

Studies like these that attempt to measure how disturbances shape forests are important because debates about forest management often come down to debates about what types of disturbances are “more natural” than others. If a certain treatment “mimics” a natural disturbance then it might be considered better. For example, doing clearcuts or allowing high severity fires to occur may be preferred because they are thought to be a more natural type of disturbance. On the other hand, doing light thins or only allowing low severity fires may be thought of as more natural. Mimicking a disturbance regime might be the primary objective of management, as was discussed in this post about using disturbances as a guide for management.

With respect to fire, most people think that the “most natural” regime for the Sierra Nevadas is one referred to as mixed-severity. As the authors point out, this term is difficult to define. Very broadly defined, it simply means that when fires occur, they are diverse in terms of having some areas where lots of trees are killed but also having areas where no or very few trees are killed.

The study points out the confusion of this term, however, when it is defined more precisely. From the results, one could conclude that these fires were not of mixed severity at all because most of the canopy gaps were relatively small (i.e. it was a low severity regime). On the other hand, the portion of total area in canopy gaps was dominated by a few very large gaps (i.e. a high severity regime). Putting this fire regime into the context of other types of regimes that we see in different forest types around the world, however, I think that it is safe to say that the researchers found these fires to be of mixed severity.   

Relevance to managers:

Figure 4 in this paper is very useful. Perhaps not as a broad guide for management across the Sierras, but more as a demonstration of how one could think of disturbances as a guide for management:

The graph shows that, in this case, most of the gaps that were created by the fires were relatively small. In general, there was a downward trend in the frequency of larger gaps. One could related this to management, for example, by allocating forests to either even-aged or uneven aged management in order to also achieve a downward trend in gap size. I think the total patch size area would be tremendously variable from fire to fire, so that part of the graph is less relevant. And if they had looked for smaller gap sizes (their minimum was 1.2 acres), the dots on the graph may have ended up looking more U-shaped.

Taking it another step, one could even set a rotation age based on this graph. Over a 30-year time period, 15% of the total area that burned was converted to gaps (i.e. regenerated to new trees). If one were to mimic this conversion rate into the future, it would lead to an approximate 200 year rotation age. Again, I don’t think this is useful as a broad guide until more studies like this are done, but the method may be useful for those how have an objective of mimicking what they think is a natural disturbance regime.

Critique (I always have one, no matter how good the article is):

I really like what these researchers did, and I’ve been waiting for something like this to be done for mixed conifer forests. Similar studies have been done in other forest types, but it is more difficult with mixed-severity fire regimes so there are some limitations.  

In terms of being useful for management, it would have been much better to have a smaller minimum mapping unit that 1.2 acres. What we consider regeneration of a distinct cohort can occur at much smaller scales. Ideally, we would go down all the way to the scale of a single canopy tree dying. It makes perfect sense why they used 1.2 acres- it was because of technological limitations of remote sensing data. But I think it would have been useful to do some kind of sensitivity analysis. In other words, how would the results have changed if the MMU was smaller? Bigger?

While this study uses an area that is probably as good as we can find when it comes to areas where fire has been allowed to burn, the reality is that it still has not been very long. This area has only had two fires, and as this study points out, fires are tremendously variable so more will be needed in terms of having broad implications. The authors know this and point it out, but it is worth noting as a limitation- we’ll get better information as more time passes and more fires burn in these areas. It will take a while for us to overcome the 60+ years of fire suppression, a period of time that were the dark ages of fire ecology where we learned nothing! Fiat flamma!