Discussion

The rule-based, expert-driven model used in an exotic pest risk analysis context (North American Forestry Commission 2004) was adapted for use in assessing risk of oak wilt introduction to and potential for subsequent spread within oak timberland based on spatial, temporal, and site factors. Such a system may be useful for analyzing spread and impact risks in the management of other significant forest diseases. The Wisconsin DNR plans to use the same approach to analyze risk and develop guidelines for reducing spread of Heterobasidion annosum in pine forests of the State. The model was also considered for use in modifying existing guidelines for managing oak wilt in urban and periurban forests of Wisconsin. Existing guidelines were, however, considered sufficiently robust and did not warrant such an effort.

The success of our approach relied on a collaborative planning and decision-making environment. The participatory method sought and obtained the involvement of multiple experts, stakeholders, and end users. The committee responsible for developing the criteria, conducting the risk analysis, formulating guidelines appropriate to risk ratings, and reviewing the prerelease product met for 4 hours on each of 4 days. Solicitation of stakeholder and user response and suggestions to the proposed system occurred over a 7-month time period through presentations and subsequent comment sessions held at numerous meetings, e.g., the Wisconsin Chapter of the Society of American Foresters’ annual meeting and the Wisconsin Woodland Owners’ Association annual meeting.

Several research questions were raised during the exercise of developing criteria, conducting the risk analyses, and developing guidelines appropriate for the assigned risks. The need for observed frequency or estimated probability for successful overland transmission of the oak wilt fungus between mid-July and early October is being addressed in a 3-year study initiated in summer 2006. Questions were also raised about the ultimate quantitative impact of oak wilt introduced during shelter wood preparatory cuts or clear cutting on future oak stocking in stands regenerated on dry and dry-mesic sites. On the basis of results of a West Virginia study (Tyron and others 1983), we hypothesize that the impact would be low in areas where regeneration is mostly of seedling origin. However, where coppice or stump-sprout regeneration predominates, the ultimate impact of oak wilt on stand stocking would likely be higher. A long-term study is needed to address these questions. New knowledge or previously overlooked scientific knowledge pertinent to our risk assessment system will be considered in future revisions of the product.