Geography & History

The Boreal Avian Modelling Project (BAM) was initiated in 2004 to address knowledge gaps associated with the management and conservation of boreal birds in North America.

BAM is built on the foundation of boreal bird data. The BAM database was created by collating and harmonizing avian data from the Breeding Bird Survey, Breeding Bird Atlases, and individual research, monitoring, and inventory efforts conducted across the Canadian and US boreal and hemi-boreal region.

BAM is working to develop rigorous analytical model-based approaches to support the conservation of the boreal forest region and the bird populations and communities that depend upon it. We have developed specialized statistical approaches to harmonize these datasets by correcting for survey methodology and species detectability to estimate density.

BAM models have a myriad of applications: they allow us to draw relationships between birds and their environment (e.g. vegetation, climate, disturbance) from regional to national scales, to predict their response to changes through time and across geographic areas, to explain population trends, to determine which habitats are important and why, to design monitoring efficiently and effectively, to assess how management decisions made now may affect birds in the future...just to name a few.

Scope & Scale

The geographic scope of our research projects and associated data products varies with the research question and the available data.

At project initiation in 2004, the geographic scope of BAM encompassed the boreal forest across Canada.

In 2011, BAM expanded our study region by almost 30% to include boreal and hemiboreal portions of Alaska, southern Canada, and the Upper Midwest and Northeast United States.

Our work to project avian responses to climate change included avian data from southern Canada.

More recent projects tend to include larger study areas.

 

Continental and Regional Perspectives

The BAM datasets encompass the full geographic scope of the North American boreal and hemiboreal forests. Models based on the continental avian dataset and corresponding biophysical data are useful for understanding the broad scale patterns and processes affecting avian populations. For example, we can map species' distributions, define species' ranges, estimate population sizes, make predictions about impacts of land use activities at the national or continental scale, and compare habitat use by a particular species in different locations across its range in the boreal forest region.

BAM also undertakes analyses at regional scales. A regional-scale focus allows us to incorporate more detailed biophysical data into modelling efforts and identify regionally specific relationships between birds and their environment. This in turn is valuable for predicting the response of bird populations to industrial development, land-use planning and regional-scale conservation efforts.

Temporal Coverage

The BAM database is an collection of independent datasets. Aside from the Breeding Bird Survey and a few long-term monitoring projects such as the Calling Lake Fragmentation Experiment, most datasets were collected in one or a few years. That said, we are actively exploring ways to extract trend and meaningful patterns from this temporally sparse compilation. The earliest records included in the avian point-count database come from surveys conducted in the 1990s. We welcome data from historical or current bird surveys to help fill temporal and spatial gaps and to create longer time series.

BAM Accomplishments

Fundamentals

  • Avian data: Assembled point count data for boreal and hemi-boreal North America from >100 projects, >1.5 million bird survey records, and >250,000 locations, including Breeding Bird Survey and provincial Atlases (Barker et al. 2015).
  • Biophysical data: Assembled, evaluated, and documented major sources of remote-sensed biophysical data for vegetation, productivity, climate, and disturbance data.
  • Common Attribute Schema for Forest Resource Inventory data (CASFRI): Synthesized 20 corporate, provincial, territorial and federal forest resource inventories into a single harmonized database. This allows consistent use of FRI data in regional or national studies (Cumming et al. 2010, 2014a, Cosco 2011).
  • Statistical methodology: Developed statistical methods to improve density estimates derived from point count data, harmonize data collected using different point count protocols, and integrate data collected using Acoustic Recording Units (Matsuoka et al. 2012, Sólymos et al. 2013, 2018b, Van Wilgenburg et al. 2017).
  • Intellectual capacity: BAM has actively cultivated a high degree of rigour, credibility and innovation. The resulting scientific expertise represents expert advice and technical support for monitoring, SAR, sector impacts (forestry, energy, agriculture) and environmental assessment issues in boreal forest.

Key BAM Accomplishments

Boreal landbird monitoring design

  • Recommended optimal point count protocols for density estimation (Matsuoka et al. 2011a, 2014).
  • Informed or developed several model-based sampling designs, including for oil sands monitoring, Moose Cree First Nation Homelands, and forest companies in BC and Alberta.
  • Conducted analyses to assess (1) where additional BBS routes could fill in spatial coverage data gaps (Matsuoka et al. 2011b), and (2) locations of geographic and environmental data gaps (Territories).
  • Quantified the impacts of roadside bias on population size estimates and developed methods to adjust for these biases (Matsuoka et al. 2011b).
  • Evaluated relationships among species’ detectability, phylogeny, and traits to inform conclusions from community studies and allow derivation of placeholder detectability estimates for species with currently insufficient data (Solymos et al. 2013c).

Population status and trends

  • Derived expected species’ density maps for >80 species based on climate and land-use factors (Stralberg et al. 2015).
  • Assessed feasibility of meeting BCR population objectives in a landscape managed for forestry, concluding they were not feasible for most species assessed (Mahon et al. 2014).
  • Estimated population trends for 96 species within BCR 6 in Alberta (Sólymos et al. 2018a).

Habitat requirements

  • Determined habitat associations for >90 songbirds at BCR or national scales (Cumming et al. 2014b).
  • Confirmed that six priority species exhibit regional differences in habitat selection, and that differential habitat selection needs to be incorporated into density estimates.

Threats assessment

  • Forecasted species’ densities with climate projections to estimate impacts of climate change on >80 songbird species (Stralberg et al. 2015). Associated spatial data products are guiding conservation prioritization (Lisgo et al. 2017, Stralberg 2018). Several regional projects are now exploring impacts of climate change on birds and caribou.
  • National-extent, multi-scale impacts assessment of anthropogenic disturbances (both successional and alienating) found negative effects for several forest bird species.
  • BAM population estimates were used to estimate incidental take for >80 forest birds at national and regional extents by terrestrial oil and gas exploration and extraction, forestry operations, roadside maintenance operations, mining operations, and wind energy sector (Longcore and Smith 2013).
  • Evaluated and improved incidental take risk assessment tools for the forest industry in BC and AB.
  • Regional and national population consequences of forestry on songbirds and waterfowl are being investigated using landscape simulation projects led by 6 post-doc and graduate student projects.
  • The effects of multiple land-use sectors, including the energy sector, on AB bird populations were estimated in cumulative effects frameworks (Sólymos et al. 2015, Bayne et al. 2016, Alberta Biodiversity Monitoring Institute and Boreal Avian Modelling Project 2018).

Identification of priority wildlife areas & protected areas design and evaluation

  • Evaluated potential priority areas for migratory birds using BAM models for > 90 species (Stralberg et al. 2018).
  • BAM climate change vulnerability products (Stralberg et al. 2015, 2018, Stralberg 2018) used in conservation planning and prioritization exercises by BEACONs and ECCC, including the Canadian Boreal Forest Agreement (CBFA), Northwest Boreal Landscape Conservation Cooperative (Lisgo et al. 2017).
  • Evaluating potential contribution of Sustainable Forestry Initiative-certified forests to bird populations (collaboration with the SFI, NGOs, industry, academic, and government partners).
  • Prairie Habitat Joint Venture identified priority waterfowl areas using BAM-DUC waterfowl models (Prairie Habitat Joint Venture 2014).

Species at risk conservation, recovery planning, and multi-species management

  • Developed a conceptual approach to critical habitat identification for wide-ranging species; models and approach are now being tested for Canada Warbler.
  • Estimated population trends for Canada Warbler at national scale.
  • Species density models, population estimates, and habitat associations informed recovery strategies, status assessments and conservation of Canada Warbler, Olive-sided Flycatcher and Common Nighthawk (Haché et al. 2014, Westwood 2016, Ball et al. 2016), Westwood et al. 2017.
  • Moose Cree First Nation homelands land planning process requested BAM’s involvement and models for SAR, including Canada Warbler, Olive-sided Flycatcher, Common Nighthawk and Rusty Blackbird (Dénes 2018[just waiting on permission to post this PDF]).
  • Examined the effect of caribou-specific harvest management plans on avian populations.

Full annual life cycle

  • Conducted analyses to assess the relative contributions of breeding and non-breeding ground climate and land-use change on inter-annual variation in breeding ground density; demonstrated importance of breeding grounds.