September 21, 2012

Forest scientists bet on the trifecta SWEEP in the Sierra Nevada

Article reviewed: Forests and water in the Sierra Nevada: Sierra Nevada Watershed Ecosystem Enhancement Project (SWEEP)

By R.C. Bales, J.J. Battles, Y. Chen, M.H. Conklin, E. Holst, K.L. O’Hara, P. Saksa, and W. Stewart

The plot line: [Note that this is a “white paper” (self-published), so I am straying from my typical format of reviewing only peer reviewed articles. Given the relevance for management and the quality of this particular paper, it seems worth making an exception]. This group of forest scientists quite aggressively makes the argument that forests in the Sierra Nevada can be managed for improving both the quantity and quality of water to benefit the commonwealth of California, and that there should be monetary incentives for the landowners who do such management. Their case is built upon the notion that water is of very high value and that several studies done in other similar forests clearly document that lower density forests (i.e. recently harvested) do increase water yield and potentially increase snow pack persistence. They make the case for large-scale studies that can be used in the future to help foresters and landowners meet the triad objectives of water, fire severity reduction, and species restoration (the trifecta SWEEP).

Relevant quote: “The perspective that forest management for water supply is not worth the trouble is ingrained in both upstream and downstream resource managers. The SWEEP team contends that forest management for water supply is worth the trouble…”

Relevance to landowners and stakeholders:

Forest landowners pay attention any time a scientist or economist suggests that they should be paid more for the “ecosystem services” that they provide to society. Forests support wildlife, clean air, and natural beauty that people from the city enjoy. So why shouldn’t the folks that own these forests get paid for it? There is of course a way in which landowners can be paid for protecting their forests. That is, through a conservation easement. But what these scientists are suggesting is something quite different than a conservation easement. Instead of a forest landowner getting paid to do nothing with an easement, they are suggesting that they get paid to do something! Doesn’t that sound more feasible as an economic model?

We are of course a ways away from this actually happening, but this team of scientists is trying to conduct research that will help such a system to develop. Rigorous experiments will have to be done in order to measure with accuracy how much more water can actually come from a forest managed for water quantity and quality (when I say quality, I am referring mainly to the timing of snow melt- if snow melts later in the spring/summer, then it is of higher quality in terms of value).

You can find the arguments for why forest management could be managed for water in any forest ecology text book. A simplified version of it is this:

1. All plants have leaves.
2. Leaves do photosynthesis, which pulls water from the soil and transpires some of it into the air
3. Leaves intercept snow and rain, some of which evaporates directly back into the atmosphere
4. The fewer leaves that are present, the less water will be sent into the air, and the more water will leave the site and go into reservoirs or hydro-electric facilities.

Relevance to managers:

I think the relevant quote above says it all for managers. These authors are right- water can no longer be ignored. I have personally heard other scientists and mangers state that forest management simply cannot make a significant difference when it comes to water yield or the timing of runoff. But the large amount of evidence presented in this paper suggests the contrary. And it is no secret that water is becoming a more valuable resource every year, so even small increases in yield can be meaningful. It is only a matter of time before markets force us managers to more explicitly manage forests with water as the objective. If the research these and other scientists propose comes to fruition, then we’ll be more ready for the challenge.

The UC Center for Forestry has been managing for what we call a “water efficient forest” for the past decade. It is at a slightly lower elevation than what these authors say will be optimal for increasing water yield and runoff timing, but they also provide some logic in this paper that suggests these lower elevations could increase yield as well. The easy part was thinning the forest down to a level where one could reasonably expect an increase in throughfall and runoff. In this particular case, we have harvested to a density at about 50% of the maximum that we observe on nearby stands. The density then fluctuates between about 50 and 75% of maximum over time in between harvests. Based on the estimates from this paper, this level of density reduction might result in somewhere between a 9 and 18% increase in water yield and it should mean snow persisting for a little while longer (although for low elevation forests, it is likely more about water yield than the timing of snow melt).

In my experience, the easy part in managing for water has been conducting the commercial thins. After all, it is a productive forest so we can generate revenue from the thins by harvesting commercial sized trees. We have been able to, concurrently or immediately following harvests, reduce the small tree cover and surface fuels to make the forest resistant to high severity fires. According to this paper, this action has resulted in what should be a structure that yields more water (somewhere between 9 and 18% increase). The challenge over time has been in managing the understory vegetation in order to prevent it from developing a significant amount of leaf area that would defeat the purpose of increasing water yield. It is challenging because this means conducting treatments that are not paid for with a commercial harvest. Theoretically, one could save the revenue from the commercial thin and apply it to understory treatments in between thins. This is indeed what we have done in this particular case (in the form of mastication and broadcast burn treatments), but without a financial incentive to do these treatments, I can see the situation occurring where these follow-up treatments simply aren’t done. So the authors make a good point in this paper that a water efficient forest needs to be maintained over time. It is not a one-and-done situation.

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

They say that many of the upper watershed forests are zoned as wilderness areas, the implication being that these areas cannot be managed for increased water yield. I would argue that these areas can also be managed for water quality with fire being the mechanism for maintaining low density. Without fire in these areas, they will burn with higher severity fire that could input massive amounts of sediment into downstream watercourses, thus countering any positive effect of water quantity and quality treatments that are done in non-wilderness areas.

They make an excellent point that, if runoff is delayed because of forest management activity, then hydro-electric energy production can occur later in the summer, when demands are high. I think they missed out on a point to make about the further potential for these treatments to benefit energy production during the summer. If the treatments are done in the summer and involves a biomass harvest of small trees and tops/limbs, then this would also potentially result in energy production during a time when it is most needed. Perhaps this is too speculative, but it is interesting to think about the potential for biomass harvests to by synergistic with water yield treatments from an energy production perspective.

They focus on forests between 5,000 and 12,000 feet elevation as having the most potential for increasing water yield and runoff timing, because they are productive and warm (above freezing). 12,000 feet… really? Any time I’ve been at 12,000 feet in the Sierra Nevada, I have not noticed many trees. At 12,000 feet, I’m catching my breath and enjoying the view because there aren’t many trees, if any at all. And lots of the winter period is cold at this elevation. Given their logic, it seems like this elevation should be shifted downward, perhaps between 4000 and 9000 feet. 4000-5000 foot elevation forests may not be dominated by snow, so the potential to delay runoff timing is less. But based on their logic and points scattered throughout the paper, forests in this elevation could increase yield substantially. The paper could use some clarity in reconciling all of the different factors of water yield and runoff timing in order to justify the 5 to 12,000 foot elevation target.

They report an average basal area in one of their targeted study areas of 400ft2/acre, with an average canopy cover of only 51% and an average canopy height of only 60 feet in a forest dominated by 100 year old trees. These numbers are not adding up in my head. That basal area seems very high for a forest that does not appear to be highly productive (trees growing 60 feet in 100 years). On the other hand, the LAI they report is also exceptionally high. A high LAI is the only way that I can visualize a forest like this having such a high basal area, so perhaps the numbers are good. But their statement about this forest being typical of much of the northern Sierra Nevada is a stretch- especially considering the 5000 to 12000 foot elevation range that they are talking about.

Adapting to climate change: Forests will try, but they can’t do it on their own

Article reviewed: Forest responses to climate change in the northwestern United States: Ecophysiological foundations for adaptive management

By D.J. Chmura, P.D. Anderson, G.T. Howe, C.A. Harrington, J.E. Halofsky, D.L. Peterson, D.C. Shaw, and J.B. St. Clair Published in the journal, Forest Ecology and Management (Vol. 261: 1121-1142).

The plot line: This is a review of the likely and potential effects that climate change will have on the physiology of trees in the western US. The authors discuss how these effects might influence forests at larger scales and also discuss the degree to which forests might be able to adapt to a changing climate. They focus on a changing snowpack and drought stress as important stresses that may lead to changing fire regimes and forest pest interactions. While significant impacts appear certain, they also note the tremendous uncertainty in predicting the details of how impacts will play out. They conclude that forests will not be able to adapt without management intervention. The recommended management actions that may help vulnerable forests adapt to climate change include density management, planting, and assisted migration.  

Relevant quote: “Overall, density management should be the most effective [silvicultural] approach because of its ability to lessen drought stress, fire risk, and predisposition to insects and disease.”

Relevance to landowners and stakeholders:

If forest landowners are anything like me, they go through ups and downs when it comes to worrying about how climate change might influence their forest. For forest managers, it is arguably their responsibility to think in long time frames so it is therefore their responsibility to think about how climate change might influence the forests they manage. But landowners may not have that same incentive to think longer-term. I admit that sometimes my time frame only extends to the time at which I think I am going to sell the land or when I will no longer be able to physically work on it. This tends to make me rather blasé when it comes to worrying about climate change effects. But even for those like me that suffer this periodic short-sightedness, this review reminds readers that it is wise to address climate change impacts now. The uncertainty and complexity of how climate change will affect forests are frankly overwhelming. This review includes how climate change might influence factors of how forests grow:

• Carbon dioxide concentration (going to go up)
• Temperature (going to go up)
• Precipitation (not sure where it’s going)
• Drought (going to be more common and longer)
• Wildfire (going to be more frequent and severe, but might go down after a while)
• Insects and diseases (going to emerge in new locations and intensities)

Those are just 6 factors that we know are going to change (in uncertain ways), but there are probably more. Sometimes we can consider one factor individually and make a scientific guess about how it will affect forests. But the reality is that these factors will be interacting with each other to affect forests in completely uncertain ways. We really have no clue what the exact effects will be or how long they will take to occur. But we do know they will be a big deal socially, economically, and ecologically. As I’ve reviewed in previous posts, active adaptive management is really the only realistic management response to such a foreboding reality.

Relevance to managers:

True to the title of the paper, the review focused on the foundations for adaptive management so there are not many actual management recommendations. I think these are the primary foundations which can be drawn upon from this review with respect to constructing adaptive management plans:

• Inter-breeding populations are the scale at which plants can adapt, so management decisions are ideally done at a fairly local level
• The regeneration phase of trees is the most vulnerable to the impacts of climate change
• The abiotic changes that will most likely either directly or indirectly influence forests are drought stress, a shrinking snowpack, and an earlier timing of snow melt (I am thinking mostly of dry montane forests here)
• We have already seen climate change interact with existing pests to result in unpredicted epidemics (i.e. mountain pine beetles). Expect more of the same.

The authors very briefly suggest the following as possible management responses:

• Density management. Thinning forests makes individual trees more resistant to drought stress
• Planting. Because the regeneration phase is most vulnerable to failure
• Assisted migration. It was confusing, but I believe their emphasis was on within-species range migration
• Forest stand triage. Foresters should think of the different seral stages and structures that they manage for, and then consider which of these might be most vulnerable to climate change. For example, forests that have reserves where density is very high and fuel is also very high could be the most vulnerable. Because of the vulnerability of the seedling stage to changes in climate, young stands (or those in an understory re-initiation phase) might also be especially vulnerable.  

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

The management recommendations were not as thorough as I was hoping. They provided very detailed reviews of how climate change might influence forests differently in different parts of the western states. But management recommendations were not given with anywhere near the same level of detail. Assisted migration, molecular and genetic breeding, and gene conservation were mentioned as possible strategies. Given that many folks are very skeptical of these types of intervention (in my experience, some people think assisted migration is a capital offense), it would have been useful to provide some examples or perhaps bounds on how they should be used given the range of plausible ecophysiological responses to climate change.  

Article Reviewed: Do mountain pine beetle outbreaks change the probability of active crown fire in lodgepole pine forests?

This review provided by the Battles lab of UC Berkeley

By M. Simard, W.H. Romme, J.M. Griffin, and M.G. Turner. Online preprint. Ecological Monographs.  Availability: http://esa.org/papers/

Plot line: Bark beetle outbreaks have caused extensive mortality of pine forests across western North America. In the aftermath, these forests with many dead and dying pines are widely considered to be at extreme risk of catastrophic wildfire. The goal of this study was to evaluate this conventional wisdom. Specifically they focused on the potential for the combined effects of these two disturbances, beetle kill and fire¸ to alter the structure and function of the Greater Yellowstone Ecosystem. Their approach was to quantify the impact of beetle-induced mortality on fire behavior in lodgepole pine forests.  They measured forests fuels in an extensive network of plots that spanned a gradient in time since bark-beetle mortality. In other words, they used a chronosequence approach. This chronosequence included currently undisturbed stands (no bark-beetle mortality) to stands where bark beetles killed all the pines in the canopy 36 years ago. They then used this empirical data to inform fire behavior models.  These models were used to simulate the impact of beetle damage on fire behavior.

The results of the simulations convincingly showed that mountain pine beetle outbreaks in the lodgepole pine forests of Greater Yellowstone do not increase the risk of active crown fires, the most destructive type of wildfire. Instead, their results suggest that beetle kill may actually decrease the risk of crown fire. Their explanation is that the main fuel for crown fires is tree needles. Once killed by mountain pine beetles, these needles are retained in the canopy for only brief period (1-2 years post outbreak).   Thus the exposure to the risk of catastrophic crown fires is short. In the longer term (25-35 years post outbreak), the accelerated growth of understory trees may increase the potential for passive crown fires. However these fires are less intense and spread more slowly than active crown fires.  In general, the authors concluded that weather conditions may have a greater influence on wildfire behavior than fuel characteristics affected by bark-beetle mortality.

Relevant Quote: “Our results suggest that mountain pine beetle outbreaks in Greater Yellowstone may reduce the probability of active crown fire in the short term by thinning lodgepole pine canopies.”

Relevance to landowners/stakeholders

The public perception of the ongoing bark-beetle outbreak in western North America is that the resulting forest of dead red and grey trees is an environmental disaster.  Chief among the concerns is the risk of catastrophic wildfires like those experienced in 2007 in southern California (primarily San Diego county).  This study, one of the few informed by evidence, suggests that the perceived risk of wildfire is exaggerated for the lodgepole pine forests in the Yellowstone area. However this study does not address other forest health concerns related to the widespread tree mortality nor does it address the future dynamics of these forests.

Relevance to managers

The abundance of dead and dying in trees in the western North America causes great concern among private and public forest managers.  These managers struggle with the question of what to do about the “aftermath” forests.  One option is salvage harvesting. The arguments for and against such an intervention are complex but a common reason to intervene is to reduce the risk of catastrophic wildfires. The results from this research cast doubt on this rationale. Based on their data and simulations, beetle kill does not increase the risk or hazard of wildfire in these forests.

Critique and/or limitations (there’s always something, no matter how good the article is) for the pedants:

This research relies on two well-established methodologies in forest science – chronosequences to acquire temporal data and simulation models to predict ecosystem behavior.  Chronosequences assume that the only difference between research sites is time since last disturbance (in this case, the disturbance is the bark beetle outbreak). The authors do a good job supporting this assumption but still no space-for-time substitution is perfect.  Fire science must rely on models since direct tests at the appropriate scale are impossible. Nobody is going to set an experimental fire at a landscape level. The fire behavior simulator used in this paper is a well-established approach. The physics of fire are well understood. However our ability to accurately represent the three dimensional distribution of fuels in a forest is limited. For example, the surface fuel loads and their interaction with fire are summarized in fuel models. While the authors did measure the surface fuel loads, they did not explain how they translated these results into synoptic fuel models. Nor do they specify which fuel models were used in their simulations. Also a key to their argument is the potential for crown fires in the aftermath forests but the current fire models (like the NEXUS model used in the paper) struggle to capture the complex dynamics of crown fires. This challenge is particularly relevant to question posed in the title since mountain pine beetle outbreaks kill trees while they are still standing. There is a well-documented progression from recently dead trees where all the leaves die and turn red but remain on the tree (red stage) to the later stage where these dead leaves have fallen off the tree. Leafless standing dead trees are referred to as the grey stage. The transition from live green tree to red stage to grey stage also represents a change in canopy fuel loads. The authors account for these losses in foliage by assigning discount rates but these assignments are made in an ad hoc manner without support or evidence to justify the discounts.  Since predicting crown fire behavior is essential to assessing the hazard related to beetle-induced mortality, it seems that a sensitivity analysis is needed to demonstrate that the results are not particularly dependent on the discount rate assignments.