Research Document - 2007/059
Risk Assessment of Bythotrephes longimanus establishment in Muskoka Lakes: an Assessment of QBRAT, the Quantitative Biological Risk Assessment Tool
By Johannsson, O.E.
Abstract
The purpose of this risk assessment exercise was to examine the usefulness and limitations of QBRAT v2 and provide feedback to the creators. In so doing, I developed a risk assessment based on the question: ’What is the probability of establishment, with negative impacts, of Bythotrephes longimanus in Muskoka Lakes next year (2006)?’ I specifically addressed the EB2 watershed as topographical and invasion data were available from Norm Yan, Allegra Cairns and Melisa Elliot (York University). This is a very specific use of the QBRAT model, as it turned out, and avoided many of the problems associated with trying to quantify impacts and costs. It did highlight the flexibility and quantitative nature of the model. The v2 model contains four sequential steps in the invasion process: arrival, survival, establishment and spread. However, these boxes can be used to estimate the cumulative probability of any sequence to a maximum of four steps, provided that each is independent of the others. If variables are not independent, the resultant interactions need to be calculated outside of QBRAT and the final independent variable used in the program. The four steps considered in the Bythotrephes risk assessment were the probability of arrival, survival of the individual, establishment related to propagule pressure, and establishment related to mortality. Each of these is the result of a complex set of variables and QBRAT could be redefined to estimate the probability of these processes as well.
Risk assessment often involves the assessment of different management options or conditions. QBRAT v2 can only consider one set of conditions which means that the data must be systematically changed and the program run numerous times in order to evaluate multiple scenarios. It is recommended that a front end and back end be constructed so that multiple scenarios can be examined at a time and that the resultant data be accumulated in a logical manner in an output file which opens on completion of the run. Presently, an output file is created for each run and is not open on completion of the run. The output from the program could be made more user friendly if the numeric values (e.g. probabilities etc) could be presented in a table which could be more easily read and copied for use in other programs.
The great strength of QBRAT is the inclusion of confidence information, which can be input as a quantitative number or ranked estimate, and the flexibility given to the programmer to define the characteristics of these estimates. This means that the final confidence distributions and sensitivity analysis around the cumulative risks are more informative for guiding management actions.
The Bythotrephes risk assessment indicated that 82% of all lakes were likely to experience severe impacts if Bythotrephes arrived. The probability of arrival at a lake was dependent on the shape of the function describing lake accessibility : the rate of invasion was nearly half if the scale was linear compared with logarithmic. The impact of compliance with boating regulations (washing of boats and gear when moving between lakes) was signigicant : roughly a third the number of lakes were invaded at 80% compliance compared with 40% compliance. Sensitivity analyses indicate that the projections of newly invaded lakes were most sensitive to the value of P1 (arrival), and this sensitivity increased as lakes became less accessible. High access lakes were equally sensitive to the seasonal patterns in boat traffic.
Accessibility Notice
This document is available in PDF format. If the document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.
- Date modified: