|Eight of the 31 warbler species that occur in BC are restricted to the northeast corner of the province. Five of these (the Bay-breasted, Connecticut, Black-throated Green, Canada, and Cape May warblers) have been designated under the Forest and Range Practices Act and regulations as ‘at risk’. These species thus have a direct impact on forest management practices but their status and trends are difficult to assess because they are hard to detect using roadside surveys, are regionally rare, and occur at the western end of their range. Moreover, it is not clear whether current monitoring efforts implemented by industry and government will be effective at detecting changes in warbler abundance over time or evaluating the effects of forest management practices. We address the question: How can we improve our knowledge and conservation of these species through a more refined adaptive monitoring and modeling framework that can provide direct and timely feedback to management? Our strategy uses two complementary approaches: a design-based approach aimed at improving the effectiveness of field surveys and a model-based approach aimed at using habitat models to target sampling in areas that will reduce uncertainty. Both approaches contribute to creating a framework for assessing the status and trends of warblers in the northeast and evaluating the effects of management practices. Canadian Forest Products (Canfor) initiated roadside breeding bird surveys in 2002 (Preston et al. 2006) and, in each subsequent year, has established additional routes so that coverage is now broad across the Peace and Fort Nelson Forest Districts. All survey data met or exceeded RISC standards. Recently, new protocols have been piloted (e.g., off-road forest interior plots; orthophotos to spot map detections) to improve effectiveness of the monitoring. The design-based approach consists of: 1) evaluating the effectiveness and efficiency of Canfor’s current monitoring design at detecting the five ‘at risk’ warbler species and, based on the results of the evaluation, 2) refining the design to improve its ability to detect changes in species occupancy and abundance. We will evaluate the power of the current monitoring design to detect: i) trends in warbler occupancy and abundance, and ii) differences in warbler occurrence among broad habitat or treatment types. We also will assess how well existing bird survey routes represent the habitats within Canfor’s tenure because that has implications for developing, testing, and applying models (model-based approach). Subsequently, we will evaluate the use of forest interior plots and orthophotos to determine if these more costly techniques are justified in terms of increasing the number of warbler detections and the power to detect trends. The model-based approach consists of developing and validating habitat-based models for each of the five ‘at risk’ warbler species. We will use two existing large-scale databases from the boreal forest region of BC and Alberta. Initially, we will develop species-habitat models using 2001-04 data from Alberta Remote Areas Project (Schmiegelow and Cumming 2004). The data consist of a spatially extensive network of point count stations located in some of the most remote areas of the province. Current digital forest inventory data from Alberta (AVI) and BC (VRI) will be used to ensure that the habitat covariates are similar in both study areas. We will adapt/calibrate the models prior to use in BC to ensure that the range of covariate values also is similar. We then will validate the models using independent (out-of-sample) data collected over the past 5 years within Canfor’s biodiversity monitoring program in northeast BC. Based on the results of the evaluation, we will refine the models to enhance their reliability as management tools for use in northeast BC. The models also will be used, in conjunction with the design-based approach, to identify gaps and help refine the monitoring design to target survey sites that will result in the greatest gains in terms of precision and accuracy of field detections and model predictions. Together, the design and model based approaches will enable us to provide practical guidelines for monitoring, modeling and managing warbler species ‘at risk’ in northeast BC. To extend these findings, we will collaborate with industry and government partners to develop a web-based biodiversity monitoring/DSS (Decision-support System) that can be used by forest managers, researchers, and policy makers to access: 1) up-to-date information on the status and trends of warblers, 2) tools for designing and evaluating alternative monitoring protocols, 3) habitat models that can be linked to forest inventory data and simulation tools, and 4) links to other relevant resources for conservation planning and management. |
Schmiegelow, F.K.A. and Cumming, S.C. 2004. The Remote Areas Project: a retrospective study of avian indicators of forest change. SFM Network Project Report. Preston, M.I., Vernier, P.R., and Campbell, R.W. 2006. A 4-year summary of summer bird surveys in TFL 48 and the Fort St. John Timber Supply Area. Report prepared for Canadian Forest Products Ltd.