New regional models developed for the Bering Sea are designed to better anticipate and incorporate climate change into resource management.
The majority of models estimate changes in the center of distribution for several commercially important species. They predict that most species’ summer distributions will shift north by between 50 and 200 kilometers by 2080-2089. Scientists also project large declines in the amount of area occupied by red king crab and snow crab and potentially northern rock sole in the summer months, a substantial increase in the area occupied by arrowtooth flounder, a key predator of walleye pollock, and declines in probability of occurrence for most species in areas with low pH and oxygen concentration.
These changes are altogether more extreme than previous species distribution model projections, which accounted for fewer climate effects.
‘As a subarctic ecosystem at the sea ice margin, the eastern Bering Sea is warming faster than much of the global ocean, resulting in the rapid redistribution of key fishery and subsistence resources,’ said Maurice Goodman, lead author and NOAA Affiliate, University of Alaska, Cooperative Institute for Climate, Oceans and Ecosystems Studies.
‘We need to provide resource managers, fishermen, and coastal communities information so they can make informed decisions about how to adapt to these changing conditions.’
Scientists built species distribution models for eight common and/or commercially important species of groundfish and crabs in the eastern Bering Sea (adults and juveniles). These include walleye pollock, Pacific halibut, Pacific cod, arrowtooth flounder, northern rock sole, yellowfin sole, snow crab, and red king crab.
To date, most studies projecting marine species distributions rely principally on temperature and static habitat characteristics such as depth. This can potentially lead to significant underestimation of species vulnerability to climate change.
However, for this study, ecologists combined 40 years of scientific surveys with a high-resolution oceanographic model. This model was adapted to the eastern Bering Sea by scientists at NOAA’s Alaska Fisheries Science Center as part of the Alaska Climate Integrated Modeling project. They examined the effects of bottom temperature. But they also incorporated information on oxygen, pH, and a regional climate index (the extent of the eastern Bering Sea “cold pool”). They considered all of these factors to produce a range of different climate projections through the end of the century. Model projections also anticipated warming under both low and high greenhouse gas emission scenarios.
‘A big challenge for this modelling effort was to determine how likely certain outcomes are if some aspects of the system are not exactly known,’ said Jonathan Ream, co-author and fisheries biologist at the Alaska Fisheries Science Center.
‘We compared projections among different types of models to quantify the sources of uncertainty when including these novel factors (pH, oxygen, and the cold pool) in species range projections.’
The Cold Pool
The dynamics of the Bering Sea ecosystem are tightly coupled with the annual extent of sea ice and the cold pool that forms beneath it. This colder mass of water affects primary production in the Bering Sea. It also affects the spatial distribution of important groundfish and their prey, including krill and forage fish.
The cold pool can also act as a barrier to the movements of groundfish along the shelf. For example, models demonstrated that the cold pool can block arrowtooth flounder from reaching suitable environmental conditions on the inner Eastern Bering Sea shelf. With the projected loss of the cold pool under some future carbon dioxide emission scenarios, a much larger portion of the shelf could become accessible to flounder. Accordingly, scientists found that models which accounted for variation in species distributions associated with the cold pool projected much larger future range shifts than those that did not.
The oceans absorb about 30% of global carbon dioxide emissions, and warmer water holds less oxygen. Climate change is also leading to the acidification of deoxygenation of much of the global ocean. All animals need oxygen to survive, and many species are expected to shift towards deeper, cooler waters to keep up with climate change. Lower dissolved oxygen content at depth may constrain their ability to do so. Reduced pH in water has the potential to impair organisms by changing their metabolism and physiological function. For crabs and other calcifying organisms, it can decrease calcification and shell formation rates.
Yet, few studies projecting future changes in species distributions integrate the effects of oxygen and pH. In many cases, these variables are not available to modelers, but recent advances in oceanographic modeling have made it possible to include their effects.
The authors found that the estimated effects of oxygen and pH were largely consistent among species. Where environmental oxygen and pH levels were lower, groundfish and crabs were less likely to be observed in scientific surveys. However, they also found the effects of oxygen and pH were difficult to disentangle using survey data, so they modeled their effects using separate models. In projecting future climate-driven changes in species distributions, they gave more say to models that did a better job reproducing past trends.
Where Scientists Hope to Go Next with this Research
These results build on—and in many cases agree with—previous distribution modeling efforts in the Bering Sea. However, they demonstrate that models that account for factors beyond temperature can result in more pronounced range shift projections.
‘What’s really exciting about this research is we are now able to construct long-term species range forecasts, which incorporate a wider array of climate impacts,’ said Kirstin Holsman, co-author and research fishery biologist, Alaska Fisheries Science Center.
In future work, species distribution models may be used to improve the representation of species interactions in multispecies stock assessment models. Scientists also hope to be able to produce short-term forecasts and long-term projections that incorporate a better understanding of predator-prey overlap.
