The bowhead whale (Balaena mysticetus) is the only baleen whale endemic to the Arctic. Bowheads are long lived (to approximately 200 years [14]) and fully adapted to arctic conditions, e.g., capable of breaking through sea ice (up to 18 cm thick) to breathe. Thus, it may seem counterintuitive that this pagophilic (ice-loving) species appears to be thriving during a period of rapid sea ice loss, at least in the Pacific Arctic region. There, population size has grown, calf counts have increased, and body condition of individual whales has improved over the last quarter-century [15, 16]. These positive outcomes have been attributed to overall expansion of primary production and an augmented food supply for bowheads due to increased zooplankton advection into the Pacific Arctic, accompanied by upwelling of prey during the extended open-water season [5]. While still recovering from over-harvest during the commercial whaling era, bowheads in the Davis Strait–Baffin Bay region appear to be increasing [7], and copious singing (up to 24 h/day) recorded throughout the winter in Fram Strait east of Greenland suggests a rebounding population there [17].

The situation for the beluga, or white whale (Delphinapterus leucas), is less clear. Trends in abundance are known for only 6 of 22 populations [18], and studies of diet and body condition are rare [19, 20, 21]. Beluga populations can be either local or migratory [18], with those that are local often considered to be at greater risk. Of note, habitat selection by 2 populations that undertake long migrations in the Pacific Arctic was associated primarily with bathymetric features rather than ice conditions during the recent period of sea ice loss [22]. A trend in one population toward longer, deeper dives was thought to be an indirect effect of sea ice loss, assuming that ecological changes shifted foraging opportunities to deeper water.

Cetaceans that migrate seasonally to arctic waters are also considered winners. In the Pacific Arctic, the gray whale’s (Eschrichtius robustus) use of continental shelf habitats has been the focus of study since the 1980s [23], during which time population size has steadily increased [24]. With the open-water season extended by 2–4 weeks [6], humpback (Megaptera novaeangliae), fin (Balaenoptera physalus), and minke whales (B. acutorostrata) are now commonly seen north of Bering Strait, unlike 30 years ago [25]. The recent surge in sightings is probably due to a combination of increased survey effort, the growth of whale populations previously hunted commercially, and climate-driven environmental changes. These same 3 baleen whale species have long been common in parts of the Atlantic Arctic [26], especially the Barents Sea [27]. Competition for prey between them and the endemic bowhead may be mediated, at least in the near term, by differences in migration timing, prey preferences, and feeding behavior [5].

Polar bears are the iconic “losers” in reports of climate change impacts [3], with walruses and ringed seals (Pusa hispida) running close behind. While some polar bear populations show signs of stress, including extreme emaciation and reproductive failure, others still appear to be in good body condition [28]. Continued loss of sea ice will lead to range contraction and increasing isolation of some populations [29], and it is clear that polar bears rely on a fat-rich diet of marine mammals that cannot be easily obtained on land [30]. Walruses, ringed seals, and bearded seals (Erignathus barbatus) are also rapidly losing sea ice habitat, but while these pagophilic species may have to swim farther to feed, they have some capacity to adapt by hauling out on land, as walruses in the Atlantic Arctic and Russia often do [31] and as ringed and bearded seals sometimes do in Svalbard [32], parts of the Okhotsk Sea [33], and Alaska [34]. The endemic narwhal (Monodon monoceros) is another loser in the rapidly changing environments of the eastern Canadian Arctic, Greenland, and Svalbard [35]. With their restricted distribution and strong fidelity to pack-ice habitat where they feed in winter, narwhals are considered the most specialized of arctic cetaceans [35]. The extreme physiological adaptations of their skeletal muscles may make them especially sensitive to climate change [36].

Resilience, in the present context, denotes the capacity to adapt to environmental change. Fundamentally, a population’s resilience to habitat alteration depends on a combination of how it responds to perturbation (adapts, moves) and how sensitive it is to perturbation (life history traits, physiological limits) [37]. Generalist foragers with broad distributions are usually considered more resilient than feeding specialists with a restricted range [38]. A simple index of resilience can be devised based upon 4 metrics: (i) population size, (ii) range, (iii) behavior, and (iv) health (Fig 1). Greater resilience is associated with large populations that display behavioral flexibility (including diet) and show resistance to disease and stress, while the reverse generally signifies lesser resilience. We suggest applying a simple 5-point ranking scale, whereby a score of 1 denotes a large, wide-ranging population that displays considerable behavioral plasticity and resistance to disease, while a score of 5 indicates the opposite. By this method, the bowhead whale population in the Pacific Arctic would receive a score of 1, while the smaller Eastern Canada–West Greenland and Svalbard–Barents Sea populations might score as 3 and 4, respectively. Similarly, large, wide-ranging polar bear and beluga populations would have a lower score (i.e., higher resilience) than small, local populations (i.e., lower resilience). This method of ranking resilience is similar to the sensitivity index developed a decade ago [34], but it is based on only 4 metrics (instead of 8), one of which is related to animal health [39]. While we recognize that such simplification brings the risk of overlooking or obscuring factors that could prove decisive in a given instance (e.g., loss of critical habitat or phenotypic uniqueness), the ranking method’s strength lies in the capacity of marine mammals to integrate and reflect complex ecosystem changes through their ecological and physiological responses [4].

Arctic marine mammal populations are often difficult to define and count, with comparatively few reliable estimates of numbers or trends [7]. Furthermore, detecting even precipitous declines in marine mammal population size is unlikely because surveys are too infrequent, and the estimates obtained are too imprecise [40]. Given these challenges and the rapid pace of environmental change in the Arctic, a more holistic approach to assess population status is urgently needed to guide conservation and management actions. We suggest a framework that links the best information available on population size (even if imprecise or only qualitative) to the other 3 metrics that contribute to resilience (Fig 2). A key strength of this framework is the inclusion of both ecological (geographical range and behavior) and physiological (health) metrics, which broaden the foundation of population assessment beyond demography alone. Further, the framework relies upon multidisciplinary science, which increases the likelihood of detecting changes in population status. For example, a shift in migratory timing, a switch in diet, or an outbreak of disease could alert resource managers to a problem that would go unnoticed when relying solely on trends in population size [40]. Importantly, the reliance of indigenous people upon arctic marine mammals makes urgent comprehensive marine mammal health monitoring due to food safety concerns [39]. This “marine mammal connection” also creates opportunities to improve understanding of the nature and trajectories of fast-changing ocean ecosystems through partnerships between conventional science practitioners and the holders of indigenous knowledge [41]. Marine mammals were recently recognized as “essential ocean variables” within a Global Ocean Observing System (GOOS) that is relevant for science and public awareness [13]. An arctic marine mammal tracking framework such as we describe could provide the GOOS with essential evidence on the status of these ecosystem sentinels, and the observatory could in turn provide an online portal for quick delivery of information to resource managers. While the goals set forth by GOOS are challenging [13], they offer a path toward sustainability through improved prediction, more precaution, and wiser policy in this era of global environmental change [42].