Improving the future of box-ironbark forests with targeted learning
Value-of-information analysis reveals the expected benefit of reducing uncertainty to a decision maker. We performed this analysis on management of box-ironbark forests.
With three management alternatives (limited harvest/ firewood removal, ecological thinning, and no management), managing the system optimally (for 150 years) with the original information would, on average, increase the amount of forest in a desirable state from 19% to 35% (a 16‐percentage point increase).
Resolving all uncertainty would, on average, increase the final percentage to 42% (a 23‐percentage point increase). However, only resolving the uncertainty for a single parameter was worth almost two‐thirds the value of resolving all uncertainty.
Victoria’s box-ironbark forests are home to many unique plants and animals. These include nationally threatened swift parrots, regent honeyeaters and dozens of EPBC-listed plants. Much of the box-ironbark forest was cleared or degraded during the 19th Century’s gold rush (and its aftermath). In the early years of the 21st Century much of what remained of this ecological community, over 200,000 hectares, was protected in new national parks and conservation reserves; contributing to a significant expansion of the protected area network in the state of Victoria.
Now, under the custodianship of Parks Victoria, a key objective of these public lands is to ensure the persistence of functioning ecological communities and species populations. Managers knew that achieving this goal required mature box-ironbark forests containing large, old trees and a diverse understorey of shrubs and herbs. But they were concerned that a majority of the region’s forest consisted of immature regrowth, still recovering from recent disturbance. These forests did not provide the biodiversity resources of mature forests. Therefore, Parks Victoria and other regional managers sought to manage the forests to maximise the speed of recovery of mature box-ironbark.
Compared to other forest types, box-ironbark forests have less silvicultural value and their dynamics are less understood. This makes predicting the outcome of management relatively uncertain. The 2001 report which recommended the creation of the new national parks also recommended ecological thinning be used to manage the box-ironbark forests and assist the development of mature forests.
Ecological thinning is the active reduction of tree density by removing some trees (see Decision Point #86). Forests are thinned to a density that approximates the density at maturity. The reduction in density aims to improve forest health as the remaining trees should have greater available resources and should achieve faster growth rates.
Given uncertainty in the outcome of managing the box-ironbark forests, a model was built based on expert opinions (Czembor & Vesk, 2009). The model simulated the outcome of three management strategies, a pre-protection strategy with some extractive practices remaining, a post-protection non-intervention scenario, and protection with the addition of ecological thinning. The models were used to predict the outcome of each strategy after 150 years. These models indicated that given the current understanding of these ecosystems the non-intervention strategy had the best expected outcome.
However, the experts acknowledged and were able to quantify considerable uncertainty in predicting box-ironbark forest dynamics. And furthermore, of the three categories of uncertainty, between expert (experts disagreeing with one another), within expert (a given expert being unsure) and system stochasticity (the inherent variability in the ecosystem due to chance), the two expert driven components were the largest (Czembor et al, 2011).
Having a model with expert opinion-based uncertainty presents a compelling case for improving management via learning. But how should managers decide whether improving the model with new information is worthwhile? And, if it is, what should they try to learn about and how much new information should they attempt to acquire? We sought answers to each of these questions by taking the box-ironbark forest simulation model and using it to perform a ‘value-of-information analysis (Morris et al, 2017).
A value-of-information analysis simply asks what is the value of acquiring new information. It compares the expected outcome of management if no new information is sought with what is the expected outcome of management if a decision is made after learning something new. The analysis accounts for the possible results of that learning.
For a manager, the first requirement is to show that the expected outcome after learning is greater than the expected outcome without new information. In this case we found that if we could resolve all the uncertainty in the model we would expect management to result in an increase in the percentage of mature box-ironbark forests from 19% to 42% after 150 years (whereas without resolving uncertainty the increase would only, on average, be from 19% to 35%).
But it is probably being too optimistic to expect that resolving all uncertainty will be cost effective. For it only makes sense to resolve uncertainty if the cost of acquiring the new information is lower than expected benefit. An alternative to resolving uncertainty completely, is to target learning at specific aspect of the system. For example, when seeking more information about box-ironbark management it was found that targeting learning about the effects of the non-intervention scenario had a greater expected outcome for management than targeting learning towards either of the other management options.
The box-ironbark case-study is yet another example that demonstrates the utility of value-of-information analysis. In this case, as in others (see the box on information cost and benefit), a value-of-information analysis represents an important guide for managers seeking to update the understanding of system they are seeking to manage cost-effectively. Our analyses showed it is more cost‐effective to monitor low‐density regrowth forest than other states and more cost-effective to experiment with the no‐management alternative than the other management alternatives. Importantly, the most cost‐effective strategies did not include either the most desired forest states or the least understood management strategy (ecological thinning). This implies that managers cannot just rely on intuition to tell them where the most value of information will lie, as critical uncertainties in a complex system are sometimes not in plain view.
More info: Will Morris firstname.lastname@example.org
Czembor CA & PA Vesk (2009). Incorporating between-expert uncertainty into state-and-transition simulation models for forest restoration. Forest Ecology and Management 259:165– 175. https://www.sciencedirect.com/science/article/pii/S0378112709007269
Czembor CA, WK Morris, BA Wintle & PA Vesk (2011). Quantifying variance components in ecological models based on expert opinion. Journal of Applied Ecology 48:736–745. https://besjournals.onlinelibrary.wiley.com/doi/10.1111/j.1365-2664.2011.01971.x
Morris WK, MC Runge & PA Vesk (2017). The value of information for woodland management: updating a state–transition model. Ecosphere 8:e01998 https://esajournals.onlinelibrary.wiley.com/doi/full/10.1002/ecs2.1998
Information benefit and cost
The decision to gather new information is only rational if the expected benefit when making a decision with new information outweighs the cost of learning. Determining the ‘value of information’ facilitates calculating this benefit and there is a range of decision‐theoretic tools to help make this estimation. While these are often applied in fields like economics and medicine, ecology and natural resource management have been slower in incorporating them in their practice. But this is changing and there have been several examples of its application in past issues of Decision Point:
How much time and money should be invested in gathering biodiversity data?
“We explored the influence of different biodiversity survey investments of money and time using a landscape simulation approach.” Decision Point #27 (p8,9)
The value of information for conservation planning under sea level rise
“When developing a conservation adaptation plan for sea-level rise as we have here, investing in more detailed information is a highly advisable action.” Decision Point #67
Does better information save more koalas?
“The researchers found that gaining new information about survival and fecundity rates and the effect of habitat cover on mortality threats will do little to improve koala management.” Decision Point #87