The costs of conservation

A little empirical truth to help ground the theory

Children understand trade-offs. Ask them if they would like to have more chocolate cake and more candy, and it’s a no-brainer. However, ask them if they would like more chocolate cake or more candy and they could deliberate for some time. Conservation decision-making is the same. Asked if we would like to protect two beautiful habitats rich in biodiversity and ecosystem services and the answer is easy. But if we can protect only one and must choose between the two places, then it all gets much harder. We have to choose because we don’t always have enough resources to do both. The different costs of the two actions will obviously influence our decision. But just how good are we at evaluating those costs?

Over the last 15-20 years, conservation scientists have made great strides at adapting decision-support tools to account for the costs of different conservation options. For example, the spatial prioritization software Marxan allows users to assign different conservation costs to different bits of the landscape (see Decision Point #62, p12,13). However, somewhere in the rush to develop theoretical tools to handle cost data, empirical efforts to estimate conservation costs have lagged behind.

Recently, I undertook a review of the cost estimates that were being used in conservation planning studies (Armsworth 2014). I focused on costs of establishing and managing terrestrial protected areas. Studies ranged from global to local in extent and the spatial resolution of the cost data being used was extremely variable.

In the rush to develop theoretical tools to handle cost data, empirical efforts to estimate conservation costs have lagged behind.” 

The lowest cost that authors often assumed is that conservation would cost nothing in some locations. The upper bound could be in the range of an eye-watering $10,000 to $60,000 per hectare (Australian dollars in 2014). However, the studies in question relied on very different ways of getting these cost estimates and the values they report are influenced by the estimation method used. For example, all of the studies focus on some components of the costs of establishing and managing protected areas only in the hopes that these will reflect spatial patterns in the overall costs that would come into play. Many of the seemingly ‘free’ conservation areas that authors were incorporating would bring with them significant cost burdens when more components of overall costs were included.

The conclusions that I drew from the review were just what you might expect from a dour Scotsman. To account for costs in conservation planning, we are going to need:

1. better estimates of conservation costs,

2. better reporting of cost estimates,

3. better analysis of those estimates and

4. greater criticality about how we incorporate those estimates into conservation planning tools.

Natural values can come with a significant ongoing management cost. The extremely high (up to 50 species per square meter) plant species richness in the longleaf pine savanna ecosystem depends on fire reoccurring within the system every one to three years. The Garcon Point Preserve in north Florida, pictured here, is managed intensively with prescribed fire in order to sustain the fire-adapted species such as these carnivorous pitcher plants. (Photo by Gwen Iacona)

Natural values can come with a significant ongoing management cost. The extremely high (up to 50 species per square meter) plant species richness in the longleaf pine savanna ecosystem depends on fire reoccurring within the system every one to three years. The Garcon Point Preserve in north Florida, pictured here, is managed intensively with prescribed fire in order to sustain the fire-adapted species such as these carnivorous pitcher plants. (Photo by Gwen Iacona)

For example, when relying on proxies for conservation costs, we need those proxies to preserve the variance in cost data and patterns of association between costs of conservation and data on biodiversity benefits. Yet, when I compared the most commonly used proxies (eg, average agricultural land value nearby) to the actual costs of establishing protected areas, these basic standards were not met. Also, many authors estimate one cost component (eg, acquisition costs) and assume that the patterns it contains adequately reflect what they would find if including other cost components (eg, ongoing stewardship costs associated with managing a site). But, when testing the merits of this assumption, again I found it sometimes performed poorly.

Practitioners of course know all about the costs of doing conservation, routinely account for them in conservation decision-making and have done so for decades. However, this often seems to be done off-line of some larger scale conservation planning initiative, after broad priority regions have been identified and when scaling down to decide just which parcels of land within those regions to target. Moreover, when I have explored the methods that different conservation organizations use to project the future cost burden associated with protecting different parcels, I have observed a lot of variability in how different individuals and organizations do it. And, sometimes, practitioners’ intuition does not resonate with what we see when we look at the data.

The principle that accounting for differences in costs of different conservation options would greatly increase the effectiveness of budget-limited conservation plans is well-established, and the conservation planning community now commonly casts their recommendations in terms of the “return-on-investment” offered by investing in different projects (for example, consider David Pannell’s approach to ranking environmental projects, see Decision Point #75, p4,5). However, somewhere along the way the conservation planning mantra that “some cost data are better than no cost data” seems to have slipped into “any old cost data will do”.

Just as conservation planning has learned the importance of being critical about the biodiversity data being used (eg, how representative are indicator taxa of wider biodiversity trends we care about? or how were the data collected and are there inherent biases – eg, sampling bias – we need to worry about?), we now need a similar maturity in how we think about data on conservation costs, where many similar considerations apply about data quality.


Variable management costs

The cost of managing protected areas can vary from little to very large so it’s important our assumptions on what real management costs into the future will be are based on solid evidence. For example, compare the two protected areas below, both managed by the same organization, The Nature Conservancy (TNC).

Below left: Managers of TNC’s Brush Mountain Preserve in Pennsylvania, have fenced sections of the preserve to exclude deer and allow for oak regeneration. Fencing like this can cost a small fortune. In addition, there has been extensive thinning of early successional trees, and a prescribed burn is planned to promote the historic forest type that was dominated by oaks and pines. The ploughed section in the foreground of the photograph is a firebreak for the upcoming burn.

Below right: In contrast, many remote mountain protected areas receive very little management effort. For example, TNC’s conservation objective for the Tally Preserve (below right), is to promote forest intactness within the northern Cumberlands region of Tennessee. The only management that occurs on this preserve was boundary marking and annual visits for ecological inventory.

(Photo by Eric Larson)

(Photo by Eric Larson)

(Photo by Eric Larson)

(Photo by Eric Larson

 


More info: Paul Armsworth p.armsworth@utk.edu 

Reference 

Armsworth PR (2014). Inclusion of costs in conservation planning depends on limited datasets and hopeful assumptions. Annals of the New York Academy of Sciences. doi: 10.1111/nyas.12455

Note: This guest editorial was written by Dr Paul Armsworth, an Associate Professor in Ecology and Evolutionary Biology at the University of Tennessee, Knoxville, USA. Paul has collaborated with many of EDG’s researchers over the years and Gwen Iacona (see page 4) was Paul’s student before joining the EDG last year. We hope to bring you more international perspectives on environmental decision science over time. 

Leave a Reply