January 8, 2010

Drop it like it’s hot: burning and thinning influences on microclimate

Article reviewed:  Spatial variability in microclimate in a mixed-conifer forest before and after thinning and burning treatments

By S. Ma, A. Concilio, B. Oakley, M. North, and J. Chen, In Press in the journal, Forest Ecology and Management.

The plot line: In a very patchy forest in the southern Sierra Nevada, the researchers measured various microclimate indicators (temperature, light intensity, humidity, soil moisture, wind) both before and following thinning, burning, and thin + burning treatments. As would be expected, any treatment that changes the forest canopy (i.e. removes trees) also changes microclimate. In this case, thinning tended to make things hotter, drier, and wetter in the soil. Burning intensified the effect of the thinning by quite a bit, especially temperature, soil moisture, and wind speed. Surprisingly (to me, anyway), the burn only treatment was the only treatment that reduced surface temperature and soil temperature.

Relevant quote: When a canopy is opened with thinning treatments, the near-ground atmosphere in the forest becomes drier, but soils become wetter” [based on these results, the same thing could be said of burning treatments; in the case of soils becoming wet, even more so].

Relevance to landowners and stakeholders:

The micro-climate beneath forests change when you alter the canopy. This should be intuitive to anyone who has walked from beneath a dense forest canopy, then out into the open without a hat on. If a landowner has the option of choosing (or a stakeholder has the option of supporting) either doing nothing, thinning, or burning, each option will likely have a different type of influence on microclimate. But change in micro-climate is not necessarily a bad thing, per se. All of these treatments (except doing nothing), can be effective at reducing wildfire severity, but if the impacts on micro-climate are so extreme as to cause other undesired effects that are worse than the benefit of reducing wildfire severity, then there is a problem. Whether the changes caused by the changes in this study are truly “extreme” is a good question that the researchers did not answer.

Relevance to managers:

There were two results that I thought were particularly interesting for managers:

1. The burn only treatment was the only treatment that decreased surface temperature and below-ground temperature. This is counter-intuitive since an albedo effect (increasing radiation absorption) would increase surface temperature. It makes me think that the burn only treatment was very light. Indeed, according to the table given, it reduced basal area by exactly zero in all stands (an error, perhaps?). In general, it was the effect of the burning after thinning that seemed to have the biggest affect on micro-climate (although the authors imply that the the heavier thinning was most "extreme"). Adding the burn treatment to the light thin often had more impact in causing microclimate changes than doing the heavier thin instead of the light thin (this was the case for temperature, humidity, vapor pressure deficit, soil moisture, radiation, and wind; not the case for surface or soil temp). The relevance to management is that burning is a blunt tool. Sometimes it can have very little effect, sometimes a big effect. Not that this is a bad thing. Spatial variability, after all, is a reasonable objective to have. Thinning, of course, can also be variable, but the intensity of treatment is much more under the control of managers. A blunt tool, it is not.

2. Soil moisture increased with treatment intensity (especially the thin + burn), DESPITE synchronous increases in humidity and vapor pressure deficit. This is most likely due (in part, anyway) to a reduction of trees that are pulling water out of the ground. Less trees pulling water in the ground = more water staying in the ground = higher soil moisture. This may be counter-intuitive, but is commonly found following even very intense treatments. Also interesting was the fact that burning tended to increase soil moisture even more than thinning. Burned soils sometimes repel water, but these seemed to be retaining it just fine. This is may be because of the low intensity of the fire and the coarse structure of the soil.

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

They suggest that the heavy thinning treatment caused “extreme” changes in microclimate, but from the data it appears that the act of adding a burning treatment was what caused even greater extremes. Further, they never describe how the heavy thinning treatment was different than the low thinning treatment.

The use of repeated measures ANOVA should have been described in more detail. There is a potential problem in using time as a within-subject factor. It assumes circularity, which means variance before the treatment was the same as it was following the treatment. Given the potential for this circularity assumption to be violated (especially given this study design), this should have been addressed. This could have inflated the risk of a Type I error. Either show that it was not a problem, or use a different test (options include a multivariate analysis or collapsing the two response variables (before and after) into one that expressed relative change).