The State of Circumpolar Walrus Populations

60
2014a, b; Lindqvist et al. 2016). The genetic placement of these walruses in relation to other
Atlantic walruses should be a top research priority. These data should be used to delineate where
in the Russian Arctic the separation between the Atlantic and Pacific walrus subspecies occurs.
Further research on stock structure in the BCS population of Pacific walruses is also needed
(Garlich-Miller et al. 2011; USFWS 2014).
Mortality (anthropogenic and natural)
Hunting by humans is the greatest known cause of mortality in most areas of walrus range within
Canada, Greenland, Russia and the United States. The hunting mortality that the different
Atlantic walrus and Pacific walrus populations can sustain is not known. In all areas where
walruses are legally harvested, removal rates are uncertain due to uncertainties in the reported
harvest. This is due to factors such as reporting bias and lack of systematic data collection effort.
Even where extensive data on harvests are available, the proportion of the population being
harvested is typically uncertain due to biased population estimates. These uncertainties are
further compounded by significant uncertainty in struck and loss rates (animals injured or killed
but not secured) (D.B. Stewart et al. 2014a). There are few estimates for loss rates in subsistence
hunts, and none are recent. Accurate seasonal and habitat specific (e.g., land-based, and over
deep and shallow water) data on struck and lost rates are needed for all hunted populations.
For populations that are not legally harvested (e.g., KS-SBS-NZ population of Atlantic walruses;
LVS population of Pacific walruses), data on illegal removals (poaching) would be of value.
However, the collection of data needed to understand abundance and distribution should take
precedence.
The natural survival rate of walruses is thought to be high, due to their low reproductive rate and
long life span, but this has not been directly estimated (Taylor and Udevitz 2015). Rigorous
survival rate estimates do not exist for any population. Lack of this information increases
uncertainty in important population management tools used to predict demography, estimate
abundance, and calculate total allowable removals.
Rates of walrus predation by polar bears and killer whales are not well known, nor are rates of
natural mortality from sources such as disturbance, which can stampede walruses causing
trampling mortality. Efforts to study and monitor walrus behaviour at haulouts (see below) can
provide important data on disturbances. Mortality rates from pathogens and parasites are
generally unknown. The susceptibility of walruses to viral and bacterial diseases is poorly
understood for both subspecies (e.g., Nielsen et al. 2000, 2001a, 2001b, 2004; Calle et al. 2002;
Phillipa et al 2004). Research could improve understanding of potential risk to populations from