Conservation and the sampling problem for rare species on managed forest landscapes

Species of conservation concern are usually rare, which makes it difficult to assess whether forest management or other conservation activities are influencing individuals or populations. A screening or survey is often completed before undertaking certain forest management actions to assess whether potential species of concern are present and to avoid negatively influencing those species or disturbing key habitat elements. Especially relevant for species of conservation concern, absence of some species may be determined using survey protocols via a probability of failing to detect. Therefore, to best inform potential risk to rare species during forest operations, it is important to understand limitations of surveys to detect these species. To explore this, we used survey data for lepidoptera and birds from experimental forest management treatments in Oregon, USA. The sample of 12,211 lepidoptera individuals and 322 species over two years included 192 species observed less than 10 times, of which 58 were observed only once. For birds, 23,860 individuals of 96 species were observed over 12 years, of which 30 were observed less than 10 times and 14 only once. To have 95 % certainty of a singleton lepidoptera or bird species being absent would have required 2276 trap-night samples for the lepidoptera or 10,351 sample points for the rare bird. It is important to understand the limitations of surveys for rare species from a regulatory perspective. Further, in many cases, conservation benefits generated via sustainable forest management may be greater than the potential risks incurred from forest management activities for extremely rare species.