Science Advisory Report 2015/063

Estimates of abundance and total allowable removals for Hudson Bay-Davis Strait and South and East Hudson Bay Atlantic walrus (Odobenus rosmarus rosmarus) stocks

Summary

• Walrus are a challenging species to enumerate owing to their aggregated distribution and correlated haulout behavior resulting in highly variable proportion of animals hauled out at one time.
• Several methods have been used to estimate walrus abundance. All methods that estimate the number of animals on land or ice, require a correction factor to adjust the counts for animals that are in the water (not visible) when the surveys are conducted.
• Based on simulation studies, the Simple Count (SC) method, using mean counts provides a reliable and unbiased estimate of the number of animals hauled out.
• Aerial surveys were flown during September 2014 to count walruses in the northern Hudson Bay-Hudson Strait portion of the Hudson Bay-Davis Strait stock, and the South and East Hudson Bay stock.
• The uncorrected counts for the surveyed portion of the Hudson Bay-Davis Strait stock and the South and East Hudson Bay stock were 2,144 and 58 animals, respectively.
• These counts were adjusted to an estimate of total numbers using an estimated proportion of animals hauled out of 0.3. This estimate was obtained by taking the average haulout proportion from seven published studies.
• The corrected survey estimates of walrus abundance were 7,100 (95%CI: 2,500-20,400) (rounded to the nearest 100) for the Hudson Bay-Davis Strait stock and 200 (95% CI:70-570) (rounded to the nearest 10) from the South and East Hudson Bay stock.
• A 2007 survey of the east Baffin component of the Hudson Bay-Davis Strait stock resulted in an estimate of 2,500 (95% CI 1,800-3,500)(rounded to the nearest 100) for that region.
• Over 85% of the animals counted were detected on haulout sites on a complex of islands bounded by Southampton and Coats islands in northern Hudson Bay and by an island complex formed by Nottingham/Salisbury islands in western Hudson Strait. These counts are similar to counts from similar surveys of these island areas completed during the 1950s, 1970s, 1980s.
• Counts of walruses from the South and East Hudson Bay stock from the Sleeper Islands area were low, but were also similar to previous reports from this area. The survey did not cover the Hudson Bay arc or James Bay, although walrus have been seen in these areas in the past.
• A surplus production population model, that included reported harvest information, was fitted to the time series of abundance estimates from the northern Hudson Bay-Hudson Strait area of the HBDS stock, using Bayesian methods. One of the strengths of the Bayesian approach is that it provides a framework to account for uncertainty in model inputs.
• The model indicated that the population declined from approximately 10,400 (95% Credible Intervals=6,400-19,800) animals in 1954, reaching a minimum of 3,600 (95% Credible Intervals=2,600-6,100) in 1993, but has increased to 7,000 (95% Credible intervals=4,100-10,700) (rounded to the nearest 100) animals in 2014.
• Total Allowable Removals of walrus in the northern Hudson Bay-Hudson Strait component of the HBDS stock were estimated using the Potential Biological Removal method. In previous walrus assessments, a maximum rate of increase (Rmax) =0.07 has been used. This review concluded that an Rmax of 0.08 is more appropriate. Harvest options using a Recovery Factor of 0.5 and 1.0 are also presented.
• A PBR estimate based on the estimated 2014 population size from the model for the area of the HBDS stock that was surveyed was 114 and 228 for Recovery Factors of 0.5 and 1, respectively.
• A previous review, in 2013, for the east Baffin Island component of the HBDS stock presented a PBR estimate of 38 animals, but this was based on using a Rmax of 0.07. The PBR increases to 44 and 88 for Recovery Factors of 0.5 and 1.0 respectively, when recalculated using an Rmax of 0.08.
• The PBR calculation includes removals from all sources. The average reported harvest over the last 5 years from the HBDS stock has been 85 animals per year. This implies a total removal of 121, assuming a Struck and Loss (S&L) rate of 30%, but does not include removals due to non-reporting or other possible human sources of mortality, e.g., ship strikes.
• A population model was not fit to the South and East Hudson Bay stock because abundance information is more limited. PBR estimates for this stock ranged from 2 and 4 animals with Recovery Factors of 0.5 and 1, respectively. Reported average removals over the last 5 years have been 4 animals per year, implying a total removal of 6 animals, assuming a S&L of 30%, but does not include removals due to non-reporting or other human sources of mortality.
• Samples are needed to confirm if the walrus in southeast Hudson Bay are a separate stock as currently considered or are connected to the HBDS stock.
• In addition to research on stock structure, additional surveys are needed to evaluate abundance of walrus in the SEHB stock.
• A Precautionary Approach framework with clear management objectives is needed. This would allow for more risk-based advice to be provided. In the meantime, the model should be periodically updated with new survey data and model outputs used to provide PBR estimates.

This Science Advisory Report is from the October 20-23, 2015 Annual Meeting of the National Marine Mammal Peer Review Committee (NMMPRC). Additional publications from this meeting will be posted on the Fisheries and Oceans Canada (DFO) Science Advisory Schedule as they become available.

Complete document (PDF, 754 KB)

Inuktitut Atuinnaummijuq (PDF, 862 KB)

Accessibility Notice

This document is available in PDF format. If the document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.