The Value of Information

Science plays an important role in natural resource management by identifying key scientific uncertainties that affect decisions. While better science can reduce those uncertainties, this takes resources away from other components of a management program, which may harm its overall success. Environmental managers can evaluate this trade-off by considering the ‘value of information’ better science brings.

Scientists Gretchen Hanson and Michael Jones address this theme in their recent paper, “The Value of Information in Fishery Management” (Fisheries, July 2008) using an example from efforts to control larval sea lamprey populations in the Great Lakes.

Sea Lamprey Control in the Great Lakes

Sea lampreys are an invasive fish species and fish parasite which have caused severe ecological and economic damage to Great Lakes fisheries. The Great Lakes Sea Lamprey Management Program now spends US $15 million per year to control them. A key part of this program is treating tributary streams with lampricide to kill larval lamprey before they enter the lakes.

This treatment is expensive, but it is neither necessary nor cost effective to poison each stream every year. Another key part of the control program is an annual biological assessment to rank streams for treatment.

Reducing Scientific Uncertainty in Larval Lamprey Assessments

During assessment, biologists estimate how many larvae are in a stream and how close they are to leaving it. To do this, scientists sample streams for information like larval density, the number of larvae per unit area of stream habitat. This and other information, is entered into a computer ‘assessment model’, which calculates stream rankings.

Data collected through sampling carry ‘scientific uncertainty’ in the form of imprecision, or variability, and bias, or lack of accuracy. Though a stream’s larval density can be estimated as a single average value, this value falls within a range called a ‘confidence interval’, the width of which is determined in part by how many samples are taken.

The wider the confidence interval, the less precise is the estimate of larval density and the greater the uncertainty about the actual larval density of a stream. This sort of uncertainty applies to all sampled data and is compounded as they are combined in the assessment model making the stream rankings uncertain too. As a result, not all streams that should be treated will be treated. This ‘ranking uncertainty’ affects overall program objectives.

Is Reducing Scientific Uncertainty Worth It?

Better science in the form of increased sampling during assessment would narrow confidence intervals and ultimately reduce ranking uncertainty. However, assessment costs one third of the annual lamprey control budget, so increased sampling would mean fewer streams could be treated. This raises the question, is improved assessment the best way to meet the overall program objective of controlling lamprey numbers?

Hanson and Jones answered this question using a management experiment to compare the original assessment program to a less costly version. The latter increased scientific uncertainty due to less precise and possibly less accurate estimates, but allowed more streams to be treated. Surprisingly, more larval lampreys were killed using the cheaper assessment, which better met the overall objectives of the Lamprey Management Program.

This example shows that better science is not always worth it when it comes to achieving the overall objectives of management programs. Hanson and Jones emphasize that this is not always true. For example, increased assessment for early detection of new invasive species can reduce the future costs of irreversible economic and ecological damages. In either case, ‘value of information’ analyses are an important tool for evaluating trade-offs in the allocation of management resources.

Sea Lamprey

Print Article

Email Article

View All Articles

How to subscribe to feeds