Detecting Furbearer Population Trends in Alberta using Trapper Logbooks, 2018–2024
Interim Report
Author(s)
Michael Jokinen, B.Sc., P.Biol.
Robert Anderson, M.Sc., P.Biol.
Sue Peters, B.Sc., M.Sc., P.Biol.
Summary
Detecting furbearer population trends is a difficult task when relying on harvest records alone. Alberta Conservation Association (ACA) has been working with Alberta Trappers Association (ATA) to develop an approach that uses trapping effort to better understand furbearer trends. Since 2018, a subset of trappers within Alberta has voluntarily provided detailed harvest and effort information recorded in personal logbooks submitted annually. We synthesized these logbook data to estimate furbearer trends over large spatial scales. These logs are cast to gain a measure of catch-per-unit effort (CPUE), and over time a change in CPUE can indicate whether harvest is sustainable. We learned that many trappers were already keeping detailed personal records of their effort before using standardized logbooks. By using a standardized approach, we assess if the harvest and effort data provided by trappers enables us to detect CPUE for a given furbearer species, as well as detect population metrics at meaningful spatial scales.
ACA began the annual collection of trapper logbooks with a focus on marten (Martes americana) beginning in winter of 2017/18. Two years later (2019/20), we added four additional furbearer species to logbook entries including lynx (Lynx canadensis), fisher (Martes pennanti), otter (Lontra canadensis), and wolverine (Gulo gulo). Four years after this study began (2021/22), we also added wolf (Canis lupus) trapping effort and harvest to the logbook data collection.
Harvest records in the form of export sales and registrations have historically been used for speculating about changes in population trends for furbearer species. We began to see evidence emerge from the winter of 2019/20 that harvest records alone may be unreliable for detecting population trends for marten without also factoring in the effort associated with that harvest. The record of provincial exports for marten pelts displayed a year-to-year decline from 2017/18 through to 2019/20 although CPUE showed an increase during that same timeframe. This suggests that other factors beyond the apparent abundance of marten may have influenced the decline of exported marten pelts between 2017/18 and 2019/20, and that annual harvest alone may be an unreliable indicator of population change.
On average, it took 118 trap nights to harvest one marten based on the seven-year mean of CPUE among all reporting traplines (2018 to 2024). This is equivalent to setting 17 traps for a week and catching one marten. On average and among years, trappers in the Boreal Forest Natural Region had lower marten catch rates when compared to those in the Foothills and Rocky Mountains regions. In comparison, it took 200 trap nights to harvest one fisher and 139 trap nights to take one lynx when averaged over five years (2020 to 2024).
We trialled an analytical approach that made use of harvest and effort data and used estimated survival metrics to predict marten abundance across Fur Management Zones (FMZ). For this population reconstruction model (PopRecon), we used seven years of logbook data from 2018 to 2024 and ran models that explored scenarios that assumed an annual harvest represented by high (30%) versus low (10%) portions of the population. We examined marten abundance in five spatial areas (FMZs 1, 2, 3, 4, and 5 & 6) and found that it varied yearly within each FMZ. Regardless of which harvest level we assumed for population reconstruction modelling, trends in abundance were generally stable to possibly increasing, similar to trends based on CPUE. Statistical population reconstruction offers several advantages over traditional index-based monitoring; however, we lack Alberta-specific auxiliary data, including harvest probability. We will therefore treat these results with caution and will refine our approach as new information becomes available. We will also continue to explore population reconstruction models such as PopRecon as we receive additional years of marten harvest and effort data to better understand the reliability of these outputs over time.
Trappers have submitted logbooks from 389 individual RFMAs up to 2024, although the consistency of submissions on an annual basis from individual traplines is well below this number. Logbook submissions have varied from year to year ranging from 126 in the first year, to a high of 207 in year three, and then dipped well below 200 for years four through seven. This decline in logbook submissions is a concern if this trend continues. To gain robust estimates of marten abundance among all FMZs we predict that at least 300 logbooks are needed annually.