Demography and Population Dynamics of Bering Sea Krill COMPLICATED LIFE HISTORIES ACROSS THREE OCEANOGRAPHIC DOMAINS

UNDERSTANDING ECOSYSTEM PROCESSES IN THE BERING SEA 2007–2013

Krill are an important food and death relate to the demo- Fig. 1 source for many larger animals graphic structure of krill? such as whales, seals, and fish. As In the Bering Sea, three major for many other organism popula- euphausiid groups occupy different tions, krill populations and their habitats. Thysanoessa raschii were demographic structure in the ocean found in abundance in the middle are due to their growth and death and inner domains. T. inermis rates. Under favorable conditions occurred more abundantly in the the krill grow fast and build up outer domain and T. longipes were their lipid storage for use during more abundant through the outer unfavorable conditions, such as domain and beyond the shelf-break the dark cold winters of the Bering (Fig. 1). Sea. In unfavorable conditions, The demographic structure krill populations have decreased varied among different krill species. growth rates and may even shrink. In general, the dominant age peaks Krill growth and survival are struc- for T. raschii and T. inermis were in tured, in part, by food availability. the 3-9 month range. But in spring Important questions of intrinsic 2009, older individuals tended to Spatial distribution of different krill species: T. interest arise: How are growth and be more abundant for both krill raschii (TR), T. inermis (TI), and T. longipes (TL) death rates affected by the chang- species (Figs. 2, 3). in 2008. Two black lines indicate 50 m and 100 m ing conditions in the Bering Sea? bathymetry, respectively. continued on page 2 How do these changes in growth

The Big Picture The Bering Sea is a very productive ecosystem, with many economically important organisms that rely on krill as a prey item. Variability in krill growth and survival is structured as a climate-driven bottom-up control system of food availability and predation. This project examines growth and vital rates in krill species to better understand their population dynamics and their trophic linkage with predators. Three key questions are: (1) How does krill demographic structure (age and size) vary across three oceanographic domains in the Bering Sea: the inner, middle, and outer shelf? (2) How do growth and vital rates vary in the three domains? (3) How do the variations in growth and vital rates contribute to different demographic structures? During the Bering Sea Project field years 2007 – 2010, several key parameters of krill populations were measured (age, lipid content, and growth) from the same individuals, which provided unprecedented detail for modeling vital rates. We developed an individual-based model to simulate demographic structure, and we concluded that depending on the location (e.g., inner shelf versus outer shelf), krill growth and survival respond differently to large-scale oceanographic changes.

INTEGRATION AND MODELING OF SPATIAL-TEMPORAL VARIATION IN VITAL RATES FOR EUPHAUSIIDS IN EASTERN BERING SEA: IMPLICATIONS FOR DEMOGRAPHICS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. (MOCNESS) to collect krill sam- on demographic structure, growth, ples. Samples were preserved and and survival will facilitate our sorted to species level in the lab. understanding of krill’s response to Live krill samples were also collected environmental changes, which in at sea and then frozen for later turn will improve prediction of the age determination. A biochemical health of the predator populations. approach was used to determine krill Hongsheng Bi, University of Maryland Center for Environmental ages in the lab. To examine the rela- Science tionship between growth, survival, Alexei Pinchuk, University of Alaska and demographic structure, indi- Rodger Harvey, Old Dominion University The growth of T. inemis and vidual based models were fit to the T. raschii tended to be faster in observed demographic and size data The Bering Sea Project is a partnership between 2008 than in 2009, whereas the in spring and summer to determine the North Pacific Research Board’s Bering Sea growth of T. rashii was similar in krill growth rate estimates for 2008 Integrated Ecosystem Research Program and the 2008 and 2009. The difference and 2009. National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject in growth could be explained by age structure and survival rates: Why We Did It more young individuals, faster Krill are important prey for growth rate for the population; many predators, such as pollock, higher survival rate for old indi- whales, seabirds. Their abundance viduals, slower growth rate for the will have major impacts on the population. food web. Due to their complicated life history and multiple molting How We Did It cycles, it is difficult to determine At sea, we deployed a Multiple their demographic structure and Opening and Closing Net with estimate their growth using con- Deployment of a Bongo net to collect krill. an Environmental Sensing System ventional methods. Information

Fig. 2 Fig. 3

Spring and summer age structure for T. raschii in 2008 and 2009. In Spring and summer age structure for T. inermis in 2008 and 2009. Note spring 2008, 6-9 month old krill were common. In spring 2009, older krill that older krill were more abundant in spring 2009. were more abundant, but disappeared in summer. No data were available for T. raschii in summer 2008.

What is the Crystal Ball Saying about the Bering Sea? NOTHING, BUT CLIMATE MODELS ARE CALLING FOR VARYING AMOUNTS OF WARMING

It is a safe bet that the future How We Did It will include a warmer Bering Sea. Our approach featured high- But it is uncertain exactly how resolution ocean model simulations climate change will be manifested, using the Regional Ocean Modeling and in particular, how fast it will System (ROMS). This model warm in summer versus winter, includes interactions among physical and in the north versus the south. water properties, nutrient concentra- Nevertheless, these details in the tions, and the growth and consump- climate forcing are key in terms of tion of groups of plankton crucial their impacts on plankton com- to fish, sea birds and marine mam- munity structure and distributions mals. The regional simulations were and, ultimately, the entire marine embedded in large-scale atmospheric ecosystem. We addressed the for- and oceanic conditions from global midable problem of how climate climate model predictions. ROMS is change is liable to impact lower- much more realistic than the global trophic levels, i.e., the base of the models in representing smaller-scale Sandra Parker-Stetter Euphausiids, also known as “krill.” food web, using groundbreaking effects of bottom topography on the methods and massive computing currents and temperature (Figure 1). resources. continued on page 2 The Big Picture Fig. 1 While global climate models provide consistent global-scale predic- °C °C tions over the next few decades, they differ significantly in their predictions on regional scales. By using an ensemble of such models to drive a coupled physical-ecosystem model for the Bering Sea region, we were able to achieve consistent estimates of how the euphausiid population, an important food source for commercial Surface temperatures for the present climate during February from the Canadian Centre for Climate Model- fish species, would change on those ing and Analysis (CCCMA) global climate model (left panel) and from ROMS (right panel) using the CCCMA same time scales. for the large-scale atmospheric and oceanic forcing.

DOWNSCALING GLOBAL PROJECTIONS TO THE ECOSYSTEMS OF THE BERING SEA WITH NESTED BIOPHYSICAL MODELS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 2 Fig. 3

Milligrams Milligrams °C carbon per m3 carbon per m3

Surface temperature changes for August from the Near surface concentrations of euphausiids in August from ROMS projections using the present climate forc- present climate to the 2030s from ROMS using the ing (upper left panel), and from ROMS using the climate forcing of the 2030s from the CCCMA climate model CCCMA (top panel) and Model for Interdisciplinary (upper right), ECHOG climate model (lower left) and MIROC climate model (lower right). Research on Climate (MIROC) (bottom panel) global models for the climate forcing.

The climate model forecasts that euphausiid populations are likely the development of improved have been carried out are mostly to decline on the eastern Bering forecast models. similar in terms of their projec- Sea shelf. On the other hand, there tions of global means, but they is conflicting evidence from the Nick Bond, University of Washington/JISAO Georgina Gibson, University of Alaska Fairbanks predict different future climates model with respect to the sense of Al Hermann, University of Washington/JISAO from a regional perspective (Figure the expected changes in euphausiid 2). There is little justification for populations over the deep basin of selecting one of these models over the Bering Sea. The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea others to specify the large-scale Integrated Ecosystem Research Program and the future climate forcing of the Bering Why We Did It National Science Foundation’s Bering Ecosystem Sea. It is therefore prudent to take a Our project represented an Study. www.nprb.org/beringseaproject multiple-model approach, and focus ambitious effort, and we have on the range of probable outcomes. learned a lot along the way about An illustration of this range is the crucial interactions and choke- provided by a set of ROMS projec- points in the physical forcing of tions of euphausiid distributions the Bering Sea ecosystem. While in August (Figure 3). Euphausiids we may not be able to assert represent key prey for a number of exactly how climate change will species, including young walleye play out in the region, our research pollock. There is consensus from provides insights for effective mon- the ROMS model projections that itoring of this system and towards

The Spring Bloom Matters THE IMPORTANCE OF THE SPRING BLOOM TO THE OVERALL PRODUCTIVITY OF THE BERING SEA The success of the highly produc- the spring bloom to the Bering Sea What We Found tive Bering Sea fishery depends on ecosystem and to predict how these We found the spring ice-associ- massive blooms of tiny, single-celled blooms might be altered for better ated bloom (Figure 1) to be of vital plants that occur each spring when or worse in a warmer Bering Sea. importance to the productivity of increasing light and abundant nutri- the Bering Sea. It begins the plank- ents enable these plants to flourish What We Did ton growing-season and supplies a both in the ice (ice algae) and in the We collected samples over a large large and dependable food source water column (phytoplankton) as region of the shelf, using ice cores, to which the life cycles of many of the sea ice retreats. These blooms water samplers and nets to identify the important zooplankton, and support a zooplankton commu- and quantify the biomass of the continued on page 2 nity that has just awoken from a planktonic ecosystem components. Fig. 1 period of rest during the long, dark, Shipboard experiments measured cold winter. This community is important biological rates, such as made up of the small, unicellular zooplankton feeding, growth and microzooplankton and the larger, reproductive rates. Datasets then multicellular, mostly crustacean were integrated and synthesized to mesozooplankton dominated by determine regional patterns and year- copepods and krill. The zooplank- to-year variability in the biomass, pro- ton community, in response to the ductivity and consumption rates of spring bloom, dramatically increases different planktonic components. A its numbers and biomass and pro- new planktonic ecosystem model was vides an abundant, highly nutritious developed to better understand the food source for seabirds, mammals food web dynamics and to predict the and fish. This project seeks to bet- response of the planktonic ecosystem ter understand the importance of to future climate changes.

The Bering Sea planktonic food web in spring (not to scale). (Top) A mixed diatom assemblage The Big Picture from the Bering Sea spring bloom. Diatoms are Using a modeling approach combined with extensive data collection and at-sea experiments, the dominant component of both ice algal and we tackled a conundrum: why do years with warm ocean temperatures result in low overall success phytoplankton communities during spring. Photo of copepods and other large zooplankton, despite better food supplies and faster growth? Copepod credit: E. Sherr. (Middle) Dinoflagellates like this overall success is extremely sensitive to winter prey concentrations, even though those concentrations Protoperidinium sp (left) and ciliates similar to the Leegardiella sp. (right) are important members of are orders of magnitude lower than those during the spring bloom. Increased mortality during warm the microzooplankton communities. Photo credit: years and/or lower growth rates at warm temperatures that exceed the optimal thermal tolerance for E. Sherr. (Bottom) The copepod Calanus glacialis/ the animals also are important. Our model provides almost as many new questions as answers, and marshallae (left) and the euphausiid (krill) Thysa- can be used to guide future research directions. noessa raschii (right) are dominant components of the mesozooplankton. Photo credit: C. Gelfman.

LINKING SEA-ICE RETREAT TO PLANKTON COMMUNITY STRUCTURE AND FUNCTION IN THE BERING SEA: DATA SYNTHESIS, BIOPHYSICAL MODELING, AND MULTI-DECADAL PROJECTION A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. benthic, species are timed. production. A warmer ocean could Fig. 2 The planktonic food ebw in the result in tighter coupling between Bering Sea is far more complex the planktonic producers and than simply large zooplankton consumers with detrimental conse- feeding on large phytoplankton. quences for the benthic (sea floor) Copepods and krill all readily feed community. on ice algae, phytoplankton and Our planktonic ecosystem microzooplankton. Frequently model indicates that large zoo- microzooplankton, not phytoplankton have greater success in plankton, are their preferred food. cold years in spite of, not because The life cycles of many impor- of, spring-summer conditions tant zooplankton species are (Figure 2). In warm years, total timed to take advantage of the primary and microzooplank- spring bloom. Peak reproduc- ton production are higher and tion of the large copepod Calanus warmer temperatures mean that glacialis/marshallae coincides with growth and development of the the bloom. Adult females that large zooplankton are faster. Yet have survived the food-limited the large zooplankton have lower winter are mature and ready to overall success in warm years. A take advantage of the rich bloom new model of copepod life history food environment. These animals tradeoffs is narrowing down the respond rapidly to the increased potential reasons for this. One idea food supply by producing up to 50 is that warm winter temperatures eggs per female per day. The eggs decrease copepods’ overwinter- and early developmental stages of ing success by making them burn copepods are an important food through their energy reserves too source for larval fish. fast, but the model suggests that The spring bloom provides an this direct temperature effect is almost inexhaustible food supply outweighed by the positive effect allowing copepods to increase their of higher temperatures on spring- biomass by up to 10-fold between summer growth and development. early spring and summer. Even so, Another idea is that cold years the zooplankton community leaves might favor the production of ice much of the spring bloom produc- algae prior to the spring bloom. tion un-grazed. This excess produc- Robert Campbell, University of Rhode Island tivity falls to the sea floor where it Carin Ashjian, Woods Hole Oceanographic Institution supports a rich benthic community Neil Banas, University of Washington including commercially important Evelyn Lessard, University of Washington crustaceans such as king crab. Alexei Pinchuk, University of Alaska Fairbanks What if a future warmer ocean Barry Sherr, Oregon State University Evelyn Sherr, Oregon State University Model results. Time evolution of an intense ice-edge upsets this balance? We believe Jinlun Zhang, University of Washington bloom, from spring-summer 2009 observations that one reason that the spring (red) and the model (gray). Gray lines show bloom is so productive is that it modeled community evolution along Lagrangian gets a jump on the zooplankton The Bering Sea Project is a partnership between transport pathways that intersect the observed late the North Pacific Research Board’s Bering Sea grazers, which are not very abun- April bloom; the spread among them shows the Integrated Ecosystem Research Project and the effect of small-scale patchiness and variability in dant after the long winter and National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject ice-retreat timing. Observations are courtesy of the cannot consume all of the primary Mordy, Lomas, Sambrotto, and Sherr groups.

Does Water Temperature Influence Pollock Spawning? INVESTIGATING “BOOM” AND “BUST” YEARS

Walleye pollock is a vital compo- We suspected that variability in nent of the food web in the Bering water temperature contributes to Sea, providing food for myriad walleye pollock spawning success fish, bird, and marine mammal and changes of spawning distribu- species, as well as humans. But tion, suggesting that climate change pollock management is challenged could influence when and where by notoriously variable spawn- pollock spawn. We were most ing success and the subsequent interested in determining whether survival of young pollock. In fact, individual pollock conserve a the particular sequence of “boom” memory of their previous or paren- Illustrations by Beverly Vinter and “bust” years largely determines tal spawning locations or whether the success of the fishery and the they exhibit flexibility in choosing ecology of the Bering Sea for many their spawning sites. If we could The Big Picture years. Spawning conditions influ- understand how water tempera- Environmental variability is ence a series of events that set year ture influences pollock spawning increasingly recognized as a regulator class strength. continued on page 2 of marine fish spawning success and subsequent growth and survival of Fig. 1 eggs and larvae. Using long-term egg collections and spawning adult catches, we examined the relationship between walleye pollock spawning distribution and success in relation to variability of spawning season (from March to May) and water temperature. Using a novel statistical analysis we predict that pollock spawning activity progresses from the Aleutian Basin to the shelf region of the Bering Sea from March to May. We also found that pollock spawning

Illustration by Beverly Vinter increases modestly throughout the study area as mean annual water temperature increases, but this increase is spatially homogeneous. So the overall spatial distribution does not change in relation to water temperature. We found four areas of pollock spawning in the eastern Bering Sea based on long-term egg collections.

POLLOCK AND COD DISTRIBUTION A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. dynamics (abundance and distribu- catches of eggs or spawning adults ture throughout the study region tion), then we could better predict increased or decreased under dif- (Fig. 2). However, because the the influence of climate change ferent conditions of water temper- increase of spawning activity is on the ecological dynamics of the ature. Here we only show data and spatially homogeneous, the distri- Bering Sea. results from egg catches, which bution of pollock spawning does are similar to those obtained from not change considerably in relation How We Did It spawning adult pollock catches.. to changes of water temperature. We used a novel modeling In the eastern Bering Sea, most approach to relate the catch of pollock spawning activity occurs Why We Did It pollock eggs or spawning adults during spring (March to May). Understanding the influence to progression of the spawning There are four main spawning of a changing climate on pollock season and to water temperature, aggregations (Fig. 1) going from spawning is an important and after accounting for other potential the Aleutian Basin to the Pribilof timely topic of research because influences on pollock spawning. Islands in the shelf region of the pollock is a key component of the Data for this study consisted of southeast Bering Sea. Pollock Bering Sea food web and is heavily 19 years of pollock egg and larval spawning progresses from the harvested. We were most interested collections, as well as 22 years of Basin in March to the shelf in in determining whether individual adult pollock spawning season May (Fig. 2). On average, pollock pollock conserve a memory of their catch data. Our models allowed spawning is positively influenced previous or parental spawning loca- us to understand when and where by an increase of water tempera- tions or whether they exhibit flexibility in choosing their spawning Fig. 2 sites. We found that environmental variability (e.g., temperature), while affecting the overall success, did not much alter the spatial assemblage of spawning locations, so we concluded that individuals do conserve a memory of their spawning sites and have limited flexibility to respond to interannual variations of environmental conditions.

Lorenzo Ciannelli, College of Earth, Ocean, and Atmospheric Sciences, Oregon State University Nathan Bacheler, National Oceanic and Atmospheric Administration (NOAA) Southeast Fisheries Science Center Kevin Bailey, NOAA Alaska Fisheries Science Center Pollock spawning progressed from the Aleutian Basin in March to near the Pribilof Islands in May (left panel). Increased water temperature results in greater spawning activity but does not influence its location (right panel). The black contour lines and grey shading denote the average predicted egg density The Bering Sea Project is a partnership between (from multiple years); darkest color corresponding to lower density. Overlaid on the image are red or the North Pacific Research Board’s Bering Sea blue bubbles, the size of which is proportional to an expected increase (red) or decrease (blue) of the egg Integrated Ecosystem Research Program and the density as time progressed from March to May (left panel), or as water temperature increases by one National Science Foundation’s Bering Ecosystem degree (right panel). Grey lines are bathymetric contours. Study. www.nprb.org/beringseaproject

Circulation on the Bering Sea Shelf Revealed by Temperature and Salinity Measurements AUGMENTING NOAA FISHERIES ANNUAL BOTTOM TRAWL SURVEY The Big Picture Summer 2008-2010 measure- probably due to sea-ice melt, sug- Unlike other sampling during the ments of the ocean currents inferred gests a clockwise circulation around Bering Sea Project, the bottom-trawl from mass density differences on the island. That less-dense lens was survey CTD measurements are unique in the eastern Bering Sea continental absent in 2009, and saltier, denser two regards. The new measurements are shelf show predominantly north- water intruded across the shelf to made at no additional cost in ship time westward flow (Figure 1). The the 100-m contour. Measurements at established trawl sites where the ships current is strongest seaward of the in 2010 went farther north and are already scheduled to sample. Also, 100-m depth contour that crosses reveal dense water and implied flow they cover the Bering Sea shelf on a uni- the shelf from the Pribilof Islands entering Bering Strait (~65°N). form grid that is reoccupied annually, in (~57°N) toward St. Matthew contrast to process-oriented cruises that Island (~60°N) giving a cross-shelf How We Did It sail at varying times of the year and seek component to the flow. There are Ocean currents are driven by to study specific processes along scat- differences between the years. In wind, tide and horizontal differ- tered transects. This gives a shelf-wide 2008 and 2010, low-density water ences in the water’s density (the scan of the temperature, salinity and density, providing new understanding of surrounding St. Matthew Island, continued on page 2 the geostrophic currents. Fig. 1

Density–1000 (kg/m3) Density–1000 (kg/m3) Density–1000 (kg/m3) Maps of geostrophic velocity vectors drawn on a colored background of seawater mass density at the sea surface for the summer bottom trawl surveys of 2008- 2010, with purple denoting lower density and red higher density. White arrows show the geostrophic transport (in Sv = 106 m3/s) across sections S1-S4. PMEL EcoFOCI mooring sites (M2, M4, M5 and M8) along the 70m isobath are plotted in red, and depths are contoured at 30, 50, 100, 200, 1000 and 2000 m. Credit: Cokelet, E.D., 2014. 3-D water properties and geostrophic circulation on the eastern Bering Sea shelf. submitted to Deep Sea Research Part II: Topical Studies in Oceanography.

FISH FORAGE DISTRIBUTION AND OCEAN CONDITIONS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. weight of a volume of water). by the effects of the Earth’s rota- Each summer NOAA’s Alaska Temperature and salinity together tion, resulting in an ocean current. Fisheries Science Center conducts a determine the density. Ocean water This is similar to the winds that bottom trawl survey, sampling fish does not have the same density circulate around a high-pressure at over 350 sites spaced 37 km apart everywhere. Warmer, less-salty system in the atmosphere. Such to determine commercial fish stocks water is less dense than colder, currents are called ‘geostrophic’ on the eastern Bering Sea conti- saltier water. One might expect currents, from the Greek for Earth nental shelf. We attach ruggedized less-dense liquid to flow out over (‘geo’) and turning (‘strophe’). CTD (conductivity-temperature- denser liquid until it reaches a uni- In the northern hemisphere, the depth) instruments to the headropes form thickness and stops flowing, current flows with the low-density of the bottom trawl nets to measure as seen in an exotic drink made water on its right (when looking temperature and salinity profiles with layered, colored ingredients. downstream), and the sea surface through the water column (Figure However, in the big ocean some- slopes upward to the right, as well. 2). From these measurements, we thing else happens. The pressure The flow is faster where the hori- compute the water density at each force generated by horizontal dif- zontal density difference is larger site and then apply the known effect ferences in density can be balanced and extends over a greater depth. of the Earth’s rotation to infer the current. Fig. 2 Why We Did It Ocean currents transport nutrients, plankton, fish eggs and larvae – important elements of the Bering Sea ecosystem. Adding CTD measurements to the existing bottom survey provides a relatively low-cost method with broad coverage to infer ocean currents. Those observations help us to understand the ecosystem, to measure its variability and to calibrate predictive computer models that estimate future conditions under different climate scenarios.

Edward D. Cokelet, NOAA Pacific Marine Environmental Laboratory

A ruggedized CTD being attached to the bottom trawl net on the NOAA contract survey vessel F/V Arcturus.

Late Winter in the Northern Bering Sea LOW PRODUCTION, BUT WIDESPREAD FORAGING OF STORED FOOD It is well-known that the Bering northern waters their winter home. Fig. 1 Sea is productive, but it is also How does this ecosystem func- ! ! ! ! ! !! ! !!! ! ! !! !! ! ! ! ! ! ! ! ! expected that the timing of pro- tion before the onset of the spring ! !! St. La ! ! ! ! wrence I ! ! ! ductivity is tied to the availability bloom? Where are these top preda- Bering Sea ! ! s. ! !! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !! ! ! ! ! of light. The availability of light tors foraging? What other birds are ! !! ! ! ! !! ! ! ! ! ! !! !!!! ! ! ! ! ! ! ! ! !!! ! ! ! ! !! ! !!! ! ! ! ! !! ! ! !!!!! ! ! ! ! ! ! !! ! ! ! ! !! !!!!! ! ! ! ! ! ! ! ! !! ! ! !! !!!!! !!! ! ! ! !! ! ! ! ! ! !!!!!!!!!! ! ! ! ! !! !!!! ! ! ! ! !!!!!! !!!!!!!!! ! ! ! ! ! ! ! !!! ! ! ! !!! !!!!!!!!!!!!!!!!!! !!! ! ! ! ! ! ! ! ! ! !!! ! ! !! ! !! !!!! !!!!!!!!!!!!!!! ! ! ! ! ! ! !!!! ! !!! !!!! !!! !! !!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!! ! ! ! ! ! ! !! ! ! ! ! !!! ! ! !!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!! ! !! !! ! !!!!!! !!!!!!!!!!!!!!!!! !! ! ! !!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!! ! !! !! ! to drive photosynthesis is related ! ! ! !! !!!!!!!!!!!!!!! !!!!!!!!!!! !! ! ! ! !!!!!!!!!!!!!! !!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!! using these waters? Where there is ! !!! ! ! !! ! ! ! !!!!!! ! !! !! ! !!!! !! ! ! !!!! ! !!!!!!! !!!!!!!!!!!!!!!!!! ! ! !!!!!!!!!!!!!!!! !! !! !!!!!!!!!!!!! !!!!!!! ! !!!! ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! ! !! !!!!!!!!!!!!!!!!!!!!!!!! !!!!! !!!! !! ! ! ! ! ! !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! ! !! !!!!!!!!!!!!!!!!!!! !!!!! !!!!!! ! ! ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!! ! ! !!!! !! !!!!!!!!!!!!! !!!!!!! !! ! !! ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! ! !!!!!!!!!!!!!!!!!!!! !!!!!!! !! ! !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! ! !! !!!!!!!!!!!!!!!!!!!!!!! !! !!! ! ! !!!!! !!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!! !!!! ! ! ! !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! ! !! !!!!!!!!!!!!!!!!!!!!!!!! ! ! !! ! ! ! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! !!!!!!!!!!!!!! !! !!!!! !!!!! ! !!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! ! !!!!! !!! !! ! !!!!!!! !!!!!! ! ! !!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!! ! !! ! !!!!!!!!!!! !! !!! !! !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! !! ! !! !!!!!! !! ! !!!!!!! !!!!!! ! !!!!!! !! !!!!!!!!!!!!!!!!!!!! ! ! ! ! ! !! ! !! !!!!!!! !!!! !!!! ! !!!! !!!!!!!!!!!!!!!!!! ! ! ! ! !!!!!!!!!! !!! ! to both the return of sunlight as ! ! !! ! !!!!!!! !!!!!!!!!! !!! !! !!! ! ! ! ! ! open water within the ice (polyn- !! ! ! ! ! ! ! ! ! ! ! ! ! !!! !!!!! !!! ! !!! !!!!!! !! ! ! !!!!!!! ! ! ! ! ! ! !!! ! ! ! ! ! !! !!!! ! ! !!!! !! ! ! ! !!! ! !!! ! ! ! ! !!!!! ! !! ! ! !!!! ! ! ! !! !! ! ! ! ! !!!!! ! !!!! !!!!! ! !! ! ! ! ! ! !! !!! ! ! !! !!! ! ! !! !!!!! ! ! ! ! !! !!! !! !!!!!!! ! ! ! !!! !! ! ! !!! ! ! ! ! !!!! ! !! ! ! !! !!!! the spring equinox approaches, yas), is there enough light (and !! !!! ! !!!! !! !! ! ! ! ! !!! !!!!! !!!!! ! ! ! ! ! ! ! !!! ! ! ! ! ! ! ! !! !!!!! ! ! ! !!!!!! !!! ! ! ! ! !!!!!! ! ! ! ! ! ! !!!! ! !!! ! ! ! ! !! ! ! ! as well as the retreat of sea ice as nutrients) to stimulate production? !

!! ! winter’s hold begins to ease. These ! are the expectations, but before the How We Did It Bering Sea Project, conditions and For three consecutive years ! 0150 00 animals using the northern Bering (2008-2010), we had the extraor- km Sea (north of St. Matthew Island) dinary opportunity to sample the Spectacled eider satellite telemetry locations in late winter were actually poorly ice-covered northern Bering Sea in (n = 3,229) received from the primary winter- known. We did know that the March and make observations of ing area in the Bering Sea south of St. Lawrence shallow shelf supports some of the the water column, marine sedi- Island, Alaska in September-May in 2008-2009, 2009-2010, and 2010-2011. See Fig. 2 for study most extensive marine invertebrate ments and the animals living in this area inset map. communities in soft sediments in ecosystem. One of the objectives the world ocean, and that certain of the research was to determine Fig. 2 specialized benthic-feeding preda- where walruses and spectacled tors, including walruses, spectacled eiders were feeding on the sea floor, eiders, and bearded seals, call these continued on page 2 100 Nuculanidae (N. radiata)

80 Nuculidae (E. tenuis)

Tellinidae (M. calcarea) 60 100 Nuculanidae (N. radiata) The Big Picture Maldanidae 40 Nuculidae (E. tenuis) The Northern Bering Sea, roughly from St. Matthew Island north, is a distinct ecosystem80 that + + Sipunculida

functions differently than the open seas to the south. This difference is particularly striking in 20 Tellinidae (M. calcarea) Linear regression 60 (Nuculanidae, N. radiata) of total station biomass (gC m ) station of total the winter, when walruses, ice seals, and spectacled eiders congregate in large numbers to take 0 Maldanidae y = -2.5359x + 5118.8 40 r =0.64509, p<0.01 Percent of dominant benthic infaunal taxa benthic of dominant Percent 1985 1990 1995 2000 2005 2010 advantage of abundant food supplies on the seafloor, and also, in some cases, from under-ice prey + + Sipunculida Year such as arctic cod and euphausiids. Productivity in the water column is low due to light 20limita- Linear regression Decline of clam populations (Nuculanidae, N. radiata in the) prime spectacled of total station biomass (gC m ) station of total tions, but west-to-east decreases in chlorophyll and nutrients are already present, as they0 are later y = -2.5359x + 5118.8 eider winter foragingr =0.64509, area p<0.01 south of St. Lawrence Island in the seasonal cycle when massive sea ice edge blooms occur. The biomass of the seafloor infaunal taxa benthic of dominant Percent 1985 biologi- 1990 1995 2000 2005 2010 Yearover the past several decades. Dashed line is possible cal communities have been in decline over the past several decades, so additional changes in the (not statistically significant) increase since 2003 in ecosystem or the movement of industrial fishing northward may have negative consequences for Ennucula tenuis, a bivalve species not favored by the the entire food web of the region. eiders. Figure is modified from Grebmeier, 2012, Annual Review of Marine Science 4:63-78.

BENTHIC ECOSYSTEM RESPONSE TO CHANGING ICE COVER IN THE BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. and to match that feeding with the distribution of food resources, as well as the ever shifting sea ice that might impact the ability of these air-breathing predators to return to the sea surface.

Why We Did It The northern Bering Sea shelf is fundamentally different from more southerly Bering Sea shelves where commercial fisheries dominate. Cold bottom water temperatures Both photos by Matt Sexson Matt Both by photos influenced by ice formation are in Spectacled eiders use openings in the sea ice of the northern Bering Sea to reach clam populations on part responsible for the ecosystem the sea floor 40-60 m below. Other seabirds that use the open leads in winter and spring include black structure, which includes walruses, guillemot and Kittlitz’s murrelet, which feed on small fish, euphausiids and amphipods that are aggregated at the ice edge. Inset: A close-up view; spectacled eiders use the ice for rest between feeding gray whales, and other bottom bouts and use less energy to remain on the ice than to rest in the open water. feeding predators that depend upon abundant biological communities on the seafloor. However, satellite Fig. 3 observations indicate that the duration of seasonal sea ice, particularly a HLY0801 b HLY0901 c Polar Sea 1001 - 2010

St. Lawrence St. Lawrence St. Lawrence Island Island Island ! north of St. Lawrence Island, is HLY0901 ! !

! ! ! !! ! St. Lawrence ! ! ! ! ! ! ! ! Island ! ! ! ! ! decreasing. Continuation of these ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !! ! ! ! ! ! ! ! ! ! !! ! ! ! ! ! ! ! ! ! ! ! !! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! patterns could bring fish north, ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! as well as commercial interests. ! ! ! ! ! ! ! ! ! Industrial trawling could negatively

St. Matthew St. Matthew St. Matthew impact these rich benthic com- Island Island Island Bering Sea Nunivak Bering Sea Nunivak Bering Sea Nunivak Island Island Island St. Matthew munities that Yupik and Iñupiat 60∞N 60∞N ∞N Island Bering Sea Nunivak Island

175∞W 170∞W 175∞W 170∞W 175∞W 170∞W communities on both St. Lawrence Depth (m) Predicted Integrated 0315 06090 Chlorophyll-a (mg/m2) 0 - 30 Nautical Miles Island and the mainland depend 31 - 50

5 0 5 51 - 80 < 10 > 25 10 - 1 15 - 2 20 - 2 81 - 200 upon indirectly through subsistence Predicted integrated Chlorophyll-a in the study area in March of 2008 (panel a), 2009 (b), and 2010 harvests of top predators such as (c). Black dots are sampling stations. Water column chlorophyll (plotted in mg/m2 over the whole water walruses. column) is characteristically one-to-two orders of magnitude lower than is observed during the peak of the spring bloom in May, but west-to-east decreases are also evident due to differences in water mass Lee Cooper, Chesapeake Biological Lab, University of Maryland Center for Environmental Science productivity and nutrient content. Jackie Grebmeier, Chesapeake Biological Lab, University of Maryland Center for Environmental Science Fig. 4 Matt Sexson, USGS Alaska Science Center Chad Jay, USGS Alaska Science Center Anthony Fischbach, USGS Alaska Science Center Jim Lovvorn, Southern Illinois University Kathy Kuletz, US Fish and Wildlife Service Cal Mordy, NOAA-University of Washington JISAO

The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject Benthic biomass per square meter (left) and dominance of clams (right, brown color in circles) from 2009. Note the alignment of clam populations with spectacled eider distributions determined using satellite telemetry (see Fig. 1).

New Insights into Bering Shelf Circulation Structure WINTER WIND DIRECTION ELICITS STRONG OCEAN CURRENT RESPONSE

Southeasterly winds (winds summer months (May-September), blowing from the southeast to the winds are lighter, the shelf stratifica- northwest) tend to promote shelf tion is stronger, and the flow field is flow toward Bering Strait that more strongly controlled by other originates south and east of St. processes. Lawrence Island, and waters in the Gulf of Anadyr are more likely to How We Did It flow west past Cape Navarin. In We deployed eight oceanographic contrast, northwesterly winds tend moorings equipped with profiling to promote flow toward Bering current meters on the Bering Sea The Bering shelf. Vectors, emanating from the eight mooring deployment locations, show mean verti- Strait that originates west of St. shelf from July 2008 to July 2010. cally averaged currents during southeasterly (red) Lawrence Island, while waters east Analysis of the current meter data and northwesterly (blue) winds from July 2008 - and south of St. Lawrence Island in conjunction with the local wind July 2010. Isobaths are drawn at 200, 100, 70, 50 reverse and flow southward. These field showed close connections and 20 m depth levels. results are applicable during winter between the winds and the currents. months (October-April). During continued on page 2 The Big Picture The Bering Sea ecosystem is fundamentally dependent upon the physical mechanisms and characteristics that determine the shelf habitat: ice extent and timing regulates light penetration into the water column and provides a seasonal platform for marine mammals; temperatures control metabolic processes and set limits on geographic distributions; currents carry zooplankton and fish larvae onto the shelf from the slope; frontal systems aggregate prey items; stratification in spring and wind mixing in fall promotes phytoplankton blooms. All of these processes depend in Vertically averaged hindcast current vectors from the 3D model for a month with strong southeasterly winds (December 2000, left) and strong northwesterly winds (December 1999, right). The shelfbreak part on shelf currents. (200 m isobath) is denoted with a black line.

STRATIFICATION ON THE BERING SHELF AND ITS CONSEQUENCES FOR NUTRIENTS AND THE ECOSYSTEM: THE EFFECTS OF ICE AND COASTAL WATER ADVECTION A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. We were able to reproduce the basic nature of the observed current response using a very simple, “ideal- ized” numerical model. More complex (fully 3-D) numerical models run in hind-cast mode demonstrate how the greater shelf circulation responds in areas far removed from the mooring array.

Why We Did It A fuller understanding of the Bering Sea ecosystem requires knowledge of the continental shelf flow field and what controls it, because currents are responsible for conveying nutrients, plankton, eggs, and larvae from one place to another. Are nutrients carried onto the shelf continuously or in pulses? Do fish eggs or crab larvae get carried to the same places every year? Better knowledge of the circulation field and its variations will help us answer these questions.

Seth Danielson, Institute of Marine Science, University of Alaska Fairbanks Kate Hedstrom, Arctic Region Supercomputing Center, University of Alaska Fairbanks Knut Aagaard, Polar Science Center, Applied Physics Laboratory, University of Washington Tom Weingartner, Institute of Marine Science, University of Alaska Fairbanks Enrique Curchitser, Institute of Marine and Coastal Sciences, Rutgers University

The Bering Sea Project is a partnership between UAF and UW mooring technicians David Leech, Kevin Taylor and Jim Johnson, with the assistance of the North Pacific Research Board’s Bering Sea Coast Guard personnel, prepare to deploy a bottom-anchored sub-surface taut wire mooring in July Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem 2008. The mooring is held in place by a railroad wheel anchor; a large orange float provides buoyancy Study. www.nprb.org/beringseaproject and houses the data loggers. This mooring made hourly measurements of temperature, conductivity, pressure, and currents until it was recovered in July 2009.

Young Fish in a Warm Bering Sea THE FATE OF WALLEYE POLLOCK LARVAE A single female walleye pollock program. NOAA has been con- Fig. 1 can produce millions of eggs in her ducting plankton surveys in the lifetime. If even three of her mil- eastern Bering Sea since the mid lions of potential offspring survive to 1980’s. Fish eggs and larvae (ich- adulthood, the female has not only thyoplankton) are collected with replaced herself and her mate, but small mesh nets that strain ocean she has added one more to the over- water and accumulate the early life all population. In this case, popula- stages of fish (Figure 1). Samples tion growth is positive. However, were preserved and identified, and numerous factors act to cull the data were archived in a database number of young that survive, and of larval fish collections. We used evidence suggests that walleye pol- database-derived data on walleye lock populations are either stable or pollock eggs, larvae, and early juve- declining in the North Pacific. At niles collected on plankton surveys the same time, there is mounting conducted between 1988 and 2010 evidence for gradual warming in the to calculate the mean geographic Bering Sea, a major spawning area center-of-distribution for eggs, for walleye pollock. We asked the larvae, and juveniles over the con- question, “Do warming conditions tinental shelf during warm periods affect the survival of young walleye and cold periods. We also deter- pollock? Do they affect their distri- mined mean size of pollock larvae bution? Their growth?” and mean mortality rates during warm and cold periods. Finally, How We Did It we examined shifts in the timing We examined larval walleye of peak egg abundance to address pollock (Gadus chalcogrammus) distribution and abundance continued on page 2 Walleye pollock eggs and larvae are collected using under colder-than-average and small mesh plankton nets. warmer-than-average conditions in the Bering Sea. To examine The Big Picture long-term trends, we relied on a Our study demonstrates that shifts in ocean temperatures affect young walleye pollock larvae series of historical samples col- both directly and indirectly. Direct impacts include growth effects, metabolism, and development. lected by the National Oceanic and Indirect effects include temporal shifts in the timing of spawning and climate-mediated influences Atmospheric Administration/Alaska on ocean currents that deliver larvae to nursery areas. Our conclusion: future changes in ocean tem- Fisheries Science Center (NOAA/ peratures will alter rates of growth, development, and survival of pollock larvae, and can contribute AFSC) Fisheries Oceanography to eastward shifts in the distribution of eggs and larvae. and Coordinated Investigations

ICHTHYOPLANKTON SURVEYS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. the hypothesis that the timing of Why We Did It that available in years when tempera- pollock spawning may be delayed Walleye pollock larvae hatch out tures are cold. Prolonged feeding by under cold conditions. relatively underdeveloped, lacking fast-growing larvae on low-quality the fins that promote swimming prey jeopardizes overall survival and What we found: abilities, so they tend to be trans- recruitment. In fact, recent evidence • There is evidence of a shift in ported at the mercy of predominant suggests that fewer pollock larvae ulti- the timing of spawning of adult ocean currents. Climate-mediated mately survive to become 1-year olds walleye pollock by as much as shifts in ocean flow deliver larvae to when conditions were warm during 30 days between warm and cold different habitats during warm and the larval period compared to when years, with timing of peak egg cold periods, potentially affecting the they were cold. abundance occurring in March type and densities of zooplankton in warm years and April in cold prey that developing larvae need for Janet Duffy-Anderson, NOAA/AFSC years. growth and survival. This is impor- Franz Mueter, University of Alaska Fairbanks (UAF) • All stages of larval walleye pol- tant for young fish since survival to Tracey Smart, University of Washington (UW) Elizabeth Siddon, UAF lock were distributed over the the juvenile phase of life is a critical John Horne, UW middle shelf in warm years and step in successful recruitment to the over the outer shelf in cold years fishery. (Figure 2). In addition, fast growing larvae The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea • Mean growth rates of larval that occur at warm temperatures Integrated Ecosystem Research Program and the walleye pollock were reduced in have higher metabolic requirements, National Science Foundation’s Bering Ecosystem cold years relative to warm years necessitating access to ample, high Study. www.nprb.org/beringseaproject (Figure 3a). quality prey resources to sustain good • Mean mortality rates of larval growth. However, other work has Fig. 3 walleye pollock were elevated in determined that the zooplankton prey cold years relative to warm years available to walleye pollock larvae in A (Figure 3b). warm years is of poorer quality than Fig. 2

B LATITUDE

LONGITUDE

Distributions of all early life history stages (eggs, yolksac larvae, preflexion larvae, Larval growth rates (A) are reduced in cold years rela- postflexion larvae, juveniles) of walleye pollock are shifted eastward over the tive to warm, and larval death rates (B) are increased middle continental shelf in warm years and westward over the outer continental in cold years relative to warm. shelf in cold years.

Hidden Food in the Coldest of Times THE NUTRIENT ROLE OF SEA ICE Copepods, tiny lipid-rich obtaining sufficient food under Fig. 1 crustaceans in the Bering Sea, the ice to initiate feeding and are a favored meal of larval and reproduction. A second question juvenile Pollock. One copepod is “what is this food source”? One that dominates the zooplankton possibility was the layer of ice on the Bering Sea shelf, and shelf algae—a diverse community of areas around the Arctic Ocean, microscopic plants and animals is Calanus glacialis (Figure 1). that grow under the ice during We know that the abundance of spring (Figure 2)—rather than this species fluctuates between from the more usual phytoplank- years. Surprisingly, colder years, ton community in the water when ice cover is more extensive column beneath. and persists longer during spring, appear to favor growth of the How We Did It copepod population. Why is this? We collected zooplankton We set out to answer this ques- samples to see what the C. glacialis Fig. 2 tion during a cruise in late winter population was doing, whether of 2009 through early spring of the adult females were laying eggs 2010, when ice covered most of or not, and how much food was the Bering Sea shelf. in their guts. We determined the Since reproduction and identity of individual prey spe- growth of this copepod is con- cies in their guts from their DNA trolled by the availability of food, and quantified the amount of this we thought that they must be continued on page 2

The Big Picture The high feeding rates of the copepod Calanus glacialis that we observed during a cruise in late Blocks of sea ice turned winter and early spring of 2009/2010 could not have been sustained by the low levels of phyto- upside down to reveal plankton in the water column. This, and the presence of ice algae found in their guts, indicates that a thick layer of ice the copepods were obtaining their nutrition from ice algae. The higher feeding rates appeared to be algae. associated with warmer air temperatures, which are, in turn, associated with the release of ice algae into the water. Before this ice algae is diluted by dispersion into the water column below, it is likely that it provides a dense layer of food for the zooplankton. In years when when ice cover is more extensive and persists longer, there is an extended period of higher food availability for C. glacialis, compared with the brief ice-edge or water column phytoplankton bloom. This results in a longer period of population growth, resulting in greater abundance later during the spring.

A NOVEL MOLECULAR APPROACH TO MEASURING IN SITU FEEDING RATES OF COPEPODS IN THE SOUTH EASTERN BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. prey DNA to calculate consump- • C. glacialis eggs began to appear Fig. 4 tion (Figure 3). The amounts of in the water column during the phytoplankton chlorophyll in their cruise. At the same time we found guts provided an independent large amounts of phytoplankton measure of consumption (Figure 4). chlorophyll in the guts of adult We also measured the amount of female C. glacialis, indicating phytoplankton present in the water elevated feeding rates. column beneath the ice, and used DNA analysis to characterize the • The prey DNA in the guts of C. potential prey species present both glacialis was mostly from ice algal in the water column and in the ice species, indicating that ice algae algal community growing on the were an important source of food. underside of the ice. • Quantification of this prey DNA Estimated daily consumption by Calanus glacialis What we found: followed the same pattern of varia- adult females of the four diatom prey species, tion over time as the chlorophyll a based on the 18S DNA copy numbers in their • We found phytoplankton in guts, plotted against estimated phytoplankton extremely low concentrations pigments, indicating that DNA consumption, based on gut pigments. within the water column under can be used to provide a measure the ice, while a dense layer of ice of feeding rate on individual prey algae was at the base of the ice species. Why We Did It at most locations. Population growth of the dominant Arctic copepod C. glacialis Fig. 3 appears to be dependent upon seasonal ice cover and its associated ice algae during spring. Changes in the extent of this seasonal ice cover associated with climate change will adversely affect higher trophic levels that feed upon this key species. The Bering Sea Shelf is a region of rapid climate change. Knowledge of how key species respond to this change will help predict overall ecosystem response, and how important fisheries, as well as endangered species, may be affected.

Edward Durbin, University of Rhode Island Graduate School of Oceanography

Mean water column chlorophyll a, Calanus glacialis adult female gut pigments, abundance, egg abundance in the water, and estimated egg production rates between March 17 (Day 70) and March 31 (Day 90) 2010.

Subsistence Harvests Show Continuity and Change HARVESTS REFLECT ECOSYSTEMS AND SOCIETY

Our work documented relatively range of personal, economic, and high and diverse subsistence harvests, environmental factors when compar- The Big Picture consistent with earlier research and ing subsistence uses in the study year Alaska’s Bering Sea coasts and islands confirming the continuing economic, with previous years. These factors are home to Aleut, Yup’ik, Iñupiaq, and social, and cultural importance of included increasing costs of fuel and St. Lawrence Island Yupik peoples. Their subsistence uses of wild resources. purchased food, commercial fisheries subsistence practices are essential to The research also found differences harvests and bycatch, more persistent their cultures, heritage, and well-being, in subsistence use patterns compared storms and less predictable winds, and are recognized by customary rights to previous years’ studies, such as and reduced sea ice. Such condi- and by various laws and policies. An integrated study of the Bering Sea ecosystem tions affect resource abundance and harvest levels, harvest composition, is incomplete without understanding the and diversity of resources used. The locations as well as access to fish and people whose ways of life are part of the nature of these differences varied wildlife populations, and may shape stunning ecological and cultural richness among communities, with some long-term trends. So far, as in the and diversity of the region. In addition, increases and some decreases, sug- past, families and communities have the findings of such a study are of great gesting local influences in addition to adapted to changing economic, social, importance to those whose lives and potential region-wide changes. Survey and environmental conditions, but livelihoods are most directly affected by changes in that ecosystem. respondents identified a complex continued on page 2

Fig. 1 Fig. 2

Subsistence harvests of fish, wildlife, and wild plants, pounds usable weight Akutan, Emmonak, and Togiak: total subsistence harvests in pounds per per person, in the study communities in 2008. These substantial harvests person in 2008 compared to previous study years. provided between 170% (Togiak) and 500% (Savoonga) of daily protein requirements for community residents.

SUBSISTENCE HARVEST, USERS AND LTK ECOSYSTEM PERSPECTIVE A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Alaska Department & Game of Fish Antone Shelikoff Antone Alaska Department & Game of Fish Local research assistant Daisy Lamont interviews an Sockeye salmon drying on racks, Akutan. Jars of oil rendered from sea lion (left) and harbor Emmonak family about their subsistence harvests. seal, Akutan. the future is less clear if such changes trends in the community and the patterns is crucial to documenting intensify or accelerate. Local commu- ecosystem, and other topics that how people use and interact with their nity residents should be essential part- provided insight to help explain environment, what those interactions ners in future efforts to understand harvest patterns (example responses mean for individuals and communi- these complex processes that affect the are shown in blue boxes below). In ties, and how social and ecological natural resources of the Bering Sea. a fifth community, St. Paul, annual change may be affecting people. programs to document two key James A. Fall, Alaska Department of Fish and Game, Division of How We Did It subsistence resources, fur seals and Subsistence, Anchorage To document and quantify sub- sea lions, were continued, revealing Nicole M. Braem, Alaska Department of Fish and Game, Division of sistence harvests of fish and wildlife trends at an annual level. Subsistence, Fairbanks Caroline L. Brown, Alaska Department of Fish and Game, Division resources, comprehensive household of Subsistence, Fairbanks harvest surveys were conducted in Why We Did It Lisa B. Hutchinson-Scarbrough, Alaska Department of Fish and four Alaska Native communities on Hunting, fishing, and gathering Game, Division of Subsistence, Anchorage David S. Koster, Alaska Department of Fish and Game, Division of the Bering Sea: Akutan, Emmonak, have provided food for Alaska Native Subsistence, Anchorage Savoonga, and Togiak. The surveys communities since time immemorial. Theodore M. Krieg, Alaska Department of Fish and Game, Division of Subsistence, Dillingham used a detailed questionnaire to ask They continue to play a major role in participants about their subsistence nutrition, culture, identity, and social activities in the previous year (see connectivity within and among com- The Bering Sea Project is a partnership between Figs. 1 and 2). In addition, inter- munities. Subsistence is thus a major the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the viewers asked open-ended questions part of human use of the Bering Sea National Science Foundation’s Bering Ecosystem about factors affecting subsistence, ecosystem. Understanding subsistence Study. www.nprb.org/beringseaproject

Traditional values continue to shape subsistence hunting and A very active hunter and fisherman from Akutan described fishing activities in the study communities. As a resident of changing weather patterns that have impeded subsistence Togiak explained: activities.

“My grandparents used to tell me that the things I “Storms are more frequent and less predictable. got from this land and water don’t belong to me. It For example, the usual pattern, up to a few years was given to me to use and to respect it all the time. ago, would be a big storm with lots of wind that The first rule from my grandpa is take only what you would stay a few days, then would clear for a few can use. Even if there is abundance of whatever take days before the next would come. Lately, the last only what you can use, what you can handle. Never couple of years, the storms seem to come back-to- waste, and respect the animals, so like with the fish, back and are mixed with each other. This makes they can come back year after year after year.” it harder for us to get out and hunt because the storms make the sea too dangerous.”

Whales and Porpoise in the Bering Sea CHANGES WITH TEMPERATURE

The Bering Sea has gone through in 2002, 2008 and 2010 on the east- estimate abundance and to evalu- significant environmental changes ern Bering Sea shelf. When we saw a ate whether numbers changed over over the past decades, some of group of cetaceans, we recorded the time. We computed the abundance which may be driven by global location, the species and the number of these species by estimating how warming. Understanding distribu- of individuals. We plotted sight- many groups were seen within a tion and abundance is important for ings to examine the distribution of certain distance from the ship and evaluating how these changes will each species. Five cetacean species extrapolating to the whole survey influence the habitat and behavior (humpback, fin and minke whales, area. Finally, we estimated changes in of cetaceans (whales, dolphins and Dall’s and harbor porpoise) were abundance between 2002 and 2010 porpoise). Also, because year-to- seen in sufficient numbers to also using traditional regression methods. year variation in ocean temperatures continued on page 2 and productivity has been shown to Fig. 1 affect their prey of zooplankton and 2002 (Warm) 2008 (Cold) 2010 (Cold) Trend schooling fish, determining how 231 436 675 12.0% cetaceans interact with their envi- Humpback Whales (39 to 1370) (177 to 1073) (150 to 3040) (-9.8% to 34.0%) ronment will help predict how they 419 1368 1061 14.0% will respond to these changes. Fin Whales (219 to 802) (695 to 2692) (493 to 2283) (1.0 to 26.5%) Our study shows that the abun- 389 517 2020 15.6% dance and distribution of cetaceans Minke Whales (147 to 1030) (146 to 1831) (520 to 7855) (-6.2% to 38.6%) changed with the temperature of 35,303 14,543 11,143 -14.4% the environment. In colder years, Dall’s Porpoise (12,989 to 95,946) (7598 to27,837) (5788 to 21,451) (-29.0% to 1.0%) there were more whales and fewer 1971 4056 833 -0.7% Harbor Porpoise (798 to 4870) (1844 to 8920) (230 to 3018) (-33.6% to 24.9%) porpoise in our study area (Figure 1). The distributions of humpback Estimated abundance by species and year, with 95% confidence intervals in parentheses. (Megaptera novaeangliae) and fin whales (Balaenoptera physalus) were similar regardless of temperature, The Big Picture but the distributions of minke We examined the Bering Sea Project hypothesis that climate and ocean conditions will whales (B. acutorostrata), Dall’s affect cetacean prey and, consequently, will influence cetacean foraging habits. Will whales (Phocoenoides dalli) and harbor remain in their typical feeding areas in years when their prey is not as abundant? Because porpoise (Phocoena phocoena) other studies have found that prey populations change with ocean temperature, we use seemed to shift toward deeper temperature as a proxy for change in cetacean prey and foraging habitats. We examined the waters in colder years (Figure 2). distribution and quantified the abundance of whales and porpoise in warm and cold years and by oceanographic domain. Our conclusion: the abundance of whales increased in cold years, How We Did It which corresponds to periods when higher abundance of their potential prey seems to occur As part of NOAA’s walleye pol- as revealed by other Bering Sea Project studies on zooplankton and fish. The abundance of lock (Gadus chalcogrammus) assess- porpoise decreased in cold years, but the reasons are not yet fully understood. ment cruises, we conducted surveys

WHALE BROAD-SCALE DISTRIBUTION A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Study. www.nprb.org/beringseaproject National Science Foundation’s Bering Ecosystem Integrated Ecosystem Research Program andthe the North Pacific Research Board’s Bering Sea The Bering Sea betweenProject isapartnership preting future changes. forinter population size isnecessary on current cetaceanoccurrence and Sea. For thisreason, information warmer conditionsintheBering change causesprolonged periodsof like whalesare expectedifclimate dance oflargemarinepredators on thedistributionandabun- their prey changes.Profound effects composition andenergeticvalue of intimeand space ifthe will vary ceans andtheirpreferred habitats suspect thattheabundanceofceta- the distributionoftheirprey. We Sea, isassumedtobedriven by high latitudes,includingtheBering Why We Did It remained constant. abundance ofporpoisedeclinedor over thestudyperiod,but dance ofbaleenwhalesincreased We foundevidencethattheabun- both outerandmiddledomains. and harborporpoisewere foundin more commonintheoutershelf Bering Sea. Dall’s porpoisewere scattered throughout theeastern tinental shelf. Minke whaleswere occurred primarilyintheoutercon- the AlaskaPeninsula andfin whales concentrated incoastalwatersalong What WeFound Phillip J. Clapham, NOAA AFSCNMML Ecosystems Division Moore, E. Sue NOAA, Science of Office & Technology, Marine Janice M. Waite, NOAA AFSCNMML Collective Alexandre N.Zerbini, NOAA AFSCNMMLandCascadia Research National MarineMammalLaboratory (NMML) A.Friday,Nancy NOAA Fisheries Alaska Center Science (AFSC), The distributionofcetaceansin Humpback whaleswere more A component oftheBEST-BSIERP Project, Sea Bering funded by the National Science Foundation from Pacific participants. support Research andtheNorth Board with in-kind - Fig. 2 2008, green for 2010. ofthefiveSightings cetacean inthestudyarea, species by year—red dots for 2002,blue for WHALE BROAD-SCALE DISTRIBUTION Humpback Whale Harbor Porpoise Dall’s Porpoise Minke Whale Fin Whale

National Park Service Kate Stafford Gerry Sanger Dave Ellifrit Dave Ellifrit UNDERSTANDING ECOSYSTEM PROCESSES IN THE BERING SEA 2007–2013

Where You Are Is More Important Than Where You Started WINDS PROMOTE THE TRANSPORT OF OCEANIC ZOOPLANKTON ONTO THE BERING SEA SHELF, BUT IN-SITU PROCESSES MAY CONTROL THEIR SHELF BIOMASS On-shelf transport of oceanic Fig. 1 zooplankton onto the eastern Bering Sea shelf is elevated around submarine canyons traversing the shelf break, and around Cape Navarin in the northern Bering Sea (Figure 1). The extent of on-shelf transport depends primarily on wind direction. Southeasterly winds that blow along the Bering Sea shelf break from January-April result in increased on-shelf transport along the length of the shelf break (Figure 2), but reduced transport at the northern end of the shelf break around Cape Navarin; northwesterly winds have the opposite effect Conceptual diagram of oceanic zooplankton transport pathways onto the eastern Bering Sea shelf. (Figure 3). Southeasterly winds are Shaded regions indicate likely source areas for oceanic zooplankton, which remain relatively constant generally associated with warmer air despite inter-annual variability in wind direction. Shading intensity indicates likelihood that a region supplies zooplankton to the southern (blue) and the northern (green) Bering Sea shelf. Arrow size temperatures, while northwesterly indicates relative transport volume. winds are associated with colder air temperatures. Net tow observations of zooplankton abundance on The Big Picture the Bering Sea shelf indicate that Zooplankton are a major link in the food chain of the Bering Sea shelf. Therefore, the species Neocalanus spp. abundance and bio- composition, and the abundance of zooplankton over the shelf can impact the pelagic community at mass actually increase in cold years. a variety of trophic levels, from fish to birds and marine mammals. The distribution of zooplankton The cold years examined did not species on the shelf is influenced by processes moving water masses, along with their constitu- stand out as having periods of strong ent zooplankton communities, onto and off of the shelf. Our findings suggest that transport of the zooplankton onto the shelf appears to be enhanced in the vicinity of canyons and that wind direction SE wind, which promotes on-shelf is the primary driver in determining the on-shelf transport of large oceanic zooplankton that inhabit flow during January-April, a criti- the upper wind-mixed layer for much of their life cycle. However, the success of oceanic zooplankton cal time period for transportation once on the shelf will depend on conditions encountered, i.e. the amount of food they find and the of seasonally migrating larvae. This predation they experience. It appears that these in-situ conditions must be at least as important as suggests that changes in oceanic transport processes in determining the biomass of oceanic zooplankton over the shelf. continued on page 2

DOWNSCALING GLOBAL CLIMATE PROJECTIONS TO THE ECOSYSTEMS OF THE BERING SEA WITH NESTED BIOPHYSICAL MODELS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. zooplankton biomass on the shelf foraging habitat for higher trophic along the Gulf of Alaska and Bering may be more dependent on in-situ levels. Sea shelf breaks. Float trajectories processes promoting growth and resulting from alternate climate survival than mechanisms promot- How We Did It forcing scenarios were compared ing transport. Despite inter-annual Using a three-dimensional to determine which environmental differences in the magnitude of on- oceanographic model, coupled to a variables and conditions were most shelf transport, the relative impor- model of ‘virtual’ floats, designed to influential in controlling cross-shelf tance of source areas to supplying have ontogenetic vertical migration transport. zooplankton to the Bering Sea shelf behavior similar to the large-bodied did not vary greatly from year to oceanic zooplankton Neocalanus, Why We Did It year (Figure 3). A relatively con- we explored the mechanisms, tim- The Eastern Bering Sea shelf is sistent supply of oceanic copepods ing and location of the transport of divided into distinct hydrographic to the outer shelf of the southern Neocalanus onto the eastern Bering domains by structural fronts. Bering Sea produces a favorable Sea shelf from overwintering sources Despite frontal obstructions to cross-shelf transport, each year Fig. 2 large oceanic copepods—primarily Neocalanus spp.—are known to dominate the biomass of the outer- shelf zooplankton communities, and in some years are advected into the middle-shelf domain; the mechanisms promoting on-shelf transport of oceanic zooplankton were poorly understood. The oceanic zooplankton are an important prey source for higher trophic levels such as birds, whales and commercially important fish. Inter-annual variability in environmental conditions promoting shoreward transport of oceanic zooplankton onto the outer a) Predominant wind direction in 2001 and 2002. Blue indicates NW wind while red indicates SE wind. Bering Sea shelf have the potential b) Percentage of all floats released that were on the Bering Sea Southern Outer Shelf from January to affect energy transfer and food through June in 2001 (black) and 2002 (red). c) Domain over which the wind direction index shown in web relationships throughout the (a) was computed. Bering Sea shelf. Fig. 3 G.A. Gibson, International Arctic Research Center, University of Alaska, Fairbanks K.O. Coyle, Institute of Marine Science, University of Alaska, Fairbanks K. Hedstrom, Arctic Region Supercomputing Center E. N. Curchitser, Institute of Marine and Coastal Sciences, Rutgers University

The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the Number of virtual zooplankton floats first crossing the 200m isobath at 50 km binned locations along National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject the length of the Bering Sea shelf break, from the northern end of the shelf (Cape Navarin) to the southern end (Bering canyon) for 2001 (black) and 2002 (red). In both years, a total of 21,240 floats were released at 700m depth along the 1000m isobath.

Bering Sea Ice LIFE IN THE FREEZER While life in association with sea concentrations of algae within helped support the food webs in the ice has been studied for hundreds of these layers that exceeded water water column and at the seafloor. years in many Arctic and Antarctic column phytoplankton concentra- regions, very little information had tions by a factor of 100 to more How We Did It been gathered from the Bering Sea. than 1,000 from mid-March to the During expeditions in spring 2008, This is surprising, as the Bering end of June. The total amount of 2009, and 2010 we sampled dozens Sea has the largest seasonal sea ice plant biomass within the bottom of different ice floes (Figure 2), at extent of any Arctic region. Our 10 cm of the ice is about the same various locations, for the sea ice group focused on the biological as for the phytoplankton integrated algal abundance and growth rates. activity within the sea ice, trying to over the upper 20 m of the water Sea ice samples were taken with understand the amount and fate of column. During the ice covered a specialized ice corer (Figure 3). primary production contributed by period, pelagic crustaceans, mainly Back on the ship, ice cores were tiny algae living within the sea ice. the euphausiid Thysanoessa raschii melted and the melted samples were We discovered that each spring (sometimes known as Arctic krill) is analyzed for concentration of algal vast amounts of sea ice algae accu- likely ingesting sea ice algal mate- pigments, mainly chlorophyll a. mulate in highly concentrated rial. With the onset of ice melt, We compared these data to the algal thin bottom layers of the sea ice the ice algal biomass was rapidly development below the sea ice by floes (Figure 1), and we observed released into the water column and continued on page 2

Fig. 1 The Big Picture The Bering Sea supports an incredibly rich marine ecosystem with a wealth of marine resources exploited by commercial and subsistence harvests. Change in ocean conditions, including sea ice characteristics, impact the functioning of marine systems. To enable predictions on future scenarios, we need to understand the interplay between the current system components like fish, birds, plankton, sea floor, and sea ice plant and animal life. Our project provides one of the building blocks to understand how the Bering Sea Vertical distribution of ice algal pigments in an ice floe collected in the Bering Sea. The ice was 50 cm thick, ecosystem functions and how it might and the highest algal pigment concentrations occurred in the bottom 2 cm of the ice floe. Water column change in the future. phytoplankton concentrations at the same locations were two orders of magnitude lower.

SEA ICE ALGAE, A MAJOR FOOD SOURCE FOR HERBIVOROUS PLANKTON AND BENTHOS IN THE EASTERN BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 2 R. Gradinger Sea ice covering the Bering Sea as seen from an icebreaker in March 2009 .

Fig. 3 collecting water from below the big questions were (1) how impor- sea ice, using a small water sampler tant are these sea ice related ecosys- deployed through holes in the ice. tems for the Bering Sea food web, We also measured when and how (2) how much ice algal biomass is those algae melted out of the ice formed during spring prior to the by collecting sinking material with ice melt, and (3) how is the produc- sediment traps under the ice. We tion linked to the water column and used chemical markers to follow the seafloor communities. fate of the ice-derived matter in the Rolf Gradinger, University of Alaska Fairbanks (UAF) food web. Bodil Bluhm, UAF Katrin Iken, UAF Why We Did It Sea ice is an integral part of Arctic marine ecosystems, serving as The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea breeding and migration ground for Integrated Ecosystem Research Program and the marine mammals, resting area for National Science Foundation’s Bering Ecosystem birds and seals, and as a realm for Study. www.nprb.org/beringseaproject hundreds of different species, from microscopic, unicellular plants to larger animals. Given all the recent

C. Morel changes in temperature and ice Taking ice samples with an ice corer . conditions in the Bering Sea, the

SEA ICE ALGAE, A MAJOR FOOD SOURCE FOR HERBIVOROUS PLANKTON AND BENTHOS IN THE EASTERN BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. UNDERSTANDING ECOSYSTEM PROCESSES IN THE BERING SEA 2007–2013

Origin and Fate of Nitrogen on the Eastern Bering Sea Shelf FERTILIZER FOR AN ECOSYSTEM

The Bering Sea shelf is an excep- Bering Sea shelf is consequently a on the ice-covered shelf. By exam- tionally productive ecosystem, owing nitrogen-limited system, whose fer- ining the stable isotope ratio mea- in large part to high concentrations tility is dependent on the amount of surements of the different types of of nutrients delivered to the shelf nitrogen fertilizer in shelf waters. samples in relation to each other, we from the open Bering Sea. Nutrients were able to start building a picture entrained seasonally onto the shelf How We Did It of the movement and transforma- from the slope are thought of as “fer- We investigated the origin of tions of nitrogen through the shelf tilizer” that leads to prolific spring nitrogen and its cycling on the shelf water, sediment, and organisms. phytoplankton blooms in marginal from measurements of nutrient Our observations revealed that, ice zones. The concentration of concentrations (specifically nitrate, as expected, the annual resupply of bio-available nutrient nitrogen (N) nitrite, ammonia, and phosphate) nutrient N, in the form of nitrate, delivered to the shelf is particularly and from corresponding measure- from the open Bering Sea contrib- important, because its concentra- ments of the natural abundance utes an important fraction of the 15 14 tion in slope waters relative to other stable isotope ratios ( N/ N) of continued on page 2 nutrients is below the physiological discrete nitrogen pools in early requirement of phytoplankton. The spring shelf waters of 2007 and 2008 Fig. 1 The Big Picture Given sufficient light and adequate nutrients, phytoplankton will thrive. The seasonal retreat of sea ice on the Bering shelf allows for sunlight to penetrate the sea surface, and phytoplankton begin to bloom in nutrient-rich waters. The size of the blooms is ultimately determined by the concentrations of nutrient nitrogen in the water, which occurs in lower concentration relative to other plant nutrients. We sought to determine the amount of nutrient N replenished seasonally to the shelf from the slope, and monitor the fate of this N once on the shelf. Our conclusions: Nitrogen sourced from decomposition in sediment is a dominant source of nutrient N for the spring bloom, The proportion of nutrient nitrate originating from the decomposition of shelf material rather than newly increasingly so inshore. imported in waters from the slope, relative to the total nitrate.

NITROGEN SUPPLY FOR NEW PRODUCTION AND ITS RELATIONSHIP TO CLIMATIC CONDITIONS ON THE EASTERN BERING SEA SHELF A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. “fertilizer” available for the spring unavailable N2 gas by denitrifying the ice-covered portions of the shelf bloom upon ice retreat, particularly bacteria in sediment. So-called “deni- than those from areas of open water so on the outer shelf and on the trifiers” actually “breathe” nitrate (Fig. 2). Algae growing under sea seaward portion of the middle shelf. when oxygen runs out in order to ice obtained nitrogen in the form Shoreward on the middle and inner decompose organic material. This of ammonium rather than nitrate, shelf, however, nearly all of the N is paradoxical, in a sense, because which gave them a distinctively in the water column originates from the more fertile the shelf as a result higher δ15N than open-water algae. mineralization in situ (Figure 1). of the amount of nitrogen fertilizer Zooplankton feeding on ice- Through this process, organic mate- in the water, the more this bioavail- associated algae similarly adopted rial accrued in sediments from the able nitrogen gets used up during a higher δ15N than those feeding previous season’s growth is decom- the decomposition of dead algal on open-water algae. Because this posed in sediments during the dark material by denitrifying bacteria in distinct spatial pattern in the δ15N winter, and released as inorganic sediments. This removes the nitrogen of algae is transferred to their preda- nutrients back to the water column. fertilizer from sediments and from tors, it may prove useful for tracing In this way, the shallow continental water above the sediments, convert- the diet and movements of animals shelf recycles and retains nutrients ing this usable nitrate into unusable on and off the ice-covered shelf. through the winter, and re-mobi- N2 gas! This conundrum motivates Julie Granger, Department of Marine Sciences, University of lizes these to the water column, thus our research, to try to understand Connecticut Maria G. Prokopenko, Geology Department, Pomona College fertilizing the spring bloom. how much fertilizer N is “breathed” Laura V. Morales, Department of Plant Sciences, UC Davis to unusable N2 gas during decom- Daniel M. Sigman, Department of Geosciences, Princeton Why We Did It position in sediments rather than University The concentration of N fertil- re-released to the water column to izer relative to phosphorus, another fertilize subsequent algal blooms. The Bering Sea Project is a partnership between nutrient, decreases dramatically We also found that phytoplank- the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Project and the inshore and northward, because ton and their predators had a much National Science Foundation’s Bering Ecosystem bioavailable N is converted to higher δ15N, the 15N/14N ratio, in Study. www.nprb.org/beringseaproject

Fig. 2

15N/14N ratio (δ15N) of POM (Particulate Organic Material = algal material) in the water column of the Bering Sea shelf vs. nitrate concentration, in relation to relative sea ice cover. Phytoplanktonic algae growing under sea ice have a distinctively greater δ15N than open-water algae. Numbers correspond to those of individual shelf stations. Lines delineate the expected 15N/14N of POM derived solely from the partial assimilation of nitrate at the inner, middle and outer shelf.

Biophysical Moorings Fig. 1 TAKING MOTHER OCEAN’S PULSE FROM AFAR The Bering Sea shelf is a big decades-long records of important place. It is bigger than the state of ecosystem variables. California, with weather that chal- During the Bering Sea Project lenges even the saltiest seafarers. we discovered that water is less Ships provide the best platform sharply stratified in the north than for scientists to make most ocean the south because tides are weaker. measurements, but both funding This creates a stable layer above the and seasonal ice cover limit ship- bottom mixed layer and below the based research in the Bering Sea. surface mixed layer, which receives Scientists with the joint research sufficient light to support a sub- EcoFOCI maintains an array of four moorings on the program Ecosystems & Fisheries- surface phytoplankton bloom in southeastern Bering Sea Shelf (M2, M4, M5, M8). Oceanography Coordinated the north during summer. Summer M2 began the 19th year of observation in 2013. Investigations (EcoFOCI), have primary production can affect the used moored oceanographic productivity of the entire food instruments (“moorings”), like web. We discovered how the spring a stethoscope anchored to the phytoplankton bloom is affected by The Big Picture seafloor, to track the health of ice retreat, and that blooms occur Long-term biophysical moorings provide year-round measurements of the Bering Sea year-round since deeper below the surface in the the state of the Bering Sea, filling the 1995. These moorings provide gaps in knowledge between ship-based continued on page 2 observations. These measurements provide a foundation for understanding the mechanisms that drive this productive region. The Bering Sea Project provided the opportunity to look at targeted ecosystem questions about the physical, chemical, and biological changes in climate and ocean conditions in the context of this long-term data set. During the Bering Sea Project, we used data from these moorings to help answer questions about the differences between the northern and southern Bering Sea, and if animals will be able to shift their ranges northward with climate warming; the difference between warm and cold years on the Bering Sea shelf and how the animals that live here are affected; and how the timing of the spring bloom Edward Cokelet Edward will affect everything from the smallest EcoFOCI scientist Scott McKeever removes sensitive equipment on the surface buoy of the M2 Mooring before plankton to the largest whales. bringing it onboard the ship.

BIOPHYSICAL MOORINGS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. north. We learned that the magni- world’s most commercially valuable Fig. 2 tudes of the spring and fall blooms fisheries. This cold, shallow sea is are related, and that the interval also extremely variable, so predict- between blooms can vary by up to ing changes in ocean conditions has two months. This length of time great value from economic, ecologi- between spring and fall blooms may cal, and public safety perspectives. affect the amount of production We continue to measure the vital (i.e., food) that reaches higher tro- signs of the Bering Sea to better phic levels, including fish, seabirds, understand how this ecosystem and whales. responds to change. The Bering Sea oscillates between periods of relative How We Did It warm and cold conditions (Figure EcoFOCI maintains an array of 3). Observing how the ecosystem four long-term biophysical moor- responds to warm periods (i.e., ings in the Bering Sea (Figure 1). 2000-2005) vs. cold periods (i.e., Each mooring hosts instruments 2006-2010) can help us predict that make hourly measurements how this ecosystem responds to of temperature, salinity, nitrate, climate warming. This information chlorophyll (fluorescence), currents, will help us adapt to a changing and sea ice, year-round (Figure 2). climate and ensure that the living The instruments are programmed marine resources of the Bering Sea to take measurements at least every continue to be managed in a sus- hour and then store the data. The tainable way. M2 mooring also hosts acoustic Lisa Sheffield Guy, University of Washington Joint Institute for instruments that record zooplank- the Study of the Atmosphere and Ocean ton size and abundance and marine Phyllis J. Stabeno, National Oceanic and Atmospheric mammal vocalizations. Moorings Administration (NOAA) Pacific Marine Environmental Laboratory Jeffrey Napp, NOAA Alaska Fisheries Science Center are recovered and re-deployed using ships in spring and fall, weather and ice permitting. The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Project and the Why We Did It National Science Foundation’s Bering Ecosystem The Bering Sea supports abun- Study. www.nprb.org/beringseaproject dant and diverse wildlife, coastal communities, and some of the Diagram of a biophysical mooring including surface buoy (ice-free seasons), illustrating how instruments Fig. 3 are arranged along the length of the mooring.

Temperature averaged across all depths measured by the M2 mooring during 1995-2011.

North-South Differences in the Eastern Bering Sea Shelf CHANGES IN LATITUDE; CHANGES IN ATTITUDE Fig. 1 In the ocean, hard physi- We observed that the presence or cal borders do not exist. So what absence of seasonal ice affected the environmental cues tell organisms strength and location of the bound- when they are in a suitable habi- ary, along with water characteristics tat? One feature, which splits the such as currents and temperature on Bering Sea middle shelf in two and either side. The southern shelf was forms a transitional line between sharply stratified by temperature the northern and southern Bering during summer into warm upper Sea, is defined by seasonal sea ice and cooler bottom layers, while the and water temperature. We discov- northern shelf had a more gradual ered that this line, found between change in temperature and salinity. 59° and 61° latitude, is essentially The southern shelf was also much the divide between the cold north, warmer in years without seasonal Annual ice extent and mooring locations in where salt and temperature both ice cover. We explored the tempera- the eastern Bering Sea. Mooring locations are play a role in vertical stratification, ture preferences of Bering Sea fish indicated with the red dot (•) and the north-south and the warmer, more sharply strati- and snow crab to help predict who transect by the broken line (---). Scientists took measurements in the spring and late summer at fied southern shelf. This boundary the “winners” and “losers” might the 50+ stations along the transect line. Also persists through the summer, but be in a warmer climate without shown is the maximum ice extent in three different may become more diffuse due to seasonal sea ice. We discovered that years; 1976 and 2008 were cold years with lots of the horizontal transport of water some fish, such as pollock, which sea ice, 2001 was a warm year with minimal ice onto the shelf. continued on page 2 penetration into the southeastern Bering Sea.

The Big Picture As the climate changes, water temperatures on the northern shelf will likely remain cold during spring and summer due to seasonal sea-ice cover and darkness, making a simple northward shift in the distribution of Bering Sea species unlikely. However, if the global climate continues to warm, the southern shelf will have less ice, though large interannual variation in ice cover is expected. Biological responses to climate warming could include greater north–south differences in zooplankton communities, the transport of some large zooplankton from the outer to the middle shelf, and the disappearance of two important zooplankton prey (large copepods and krill) for planktivorous fish, seabirds, and whales. The response of commercially and ecologically important fish species is predicted to vary. Some species of fish, such as juvenile sockeye salmon, may expand their summer range into the northern Bering Sea; some (e.g., pink salmon) may increase in abundance, while still other species (e.g., walleye pollock and arrowtooth flounder) are unlikely to become common in the north. Warming of the southern shelf will likely make it more hospitable for subarctic species, but Arctic species, such as snow crab, will be restricted to colder northern waters. Baleen whales will likely be able to extend their range to follow their prey (krill and small fishes) into new areas.

BIOPHYSICAL MOORINGS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. avoid the coldest waters, could not maintains an array of four long- Why We Did It shift their ranges northward, while term biophysical moorings in the The Bering Sea supports abun- others, such as pink salmon, may Bering Sea (Figure 1). Each moor- dant and diverse wildlife and some adapt more easily. ing hosts instruments that make of the world’s most commercially hourly measurements of tempera- valuable fisheries. Predicting how How We Did It ture, salinity, nitrate, chlorophyll these animals will respond to chang- We used sea ice data from moor- (fluorescence), currents, and marine ing conditions will help coastal ings and satellites, and data from mammal vocalizations, year-round. communities, subsistence users, ships occupying a north-south The M2 mooring also hosts acoustic and commercial fishers prepare for transect between St. Lawrence instruments that record zooplank- a changing Bering Sea. We wanted Island and Bristol Bay, Alaska. ton size and abundance. Water to understand if some species would EcoFOCI (Ecosystems & Fisheries- column measurements of tempera- simply move north as seasonal sea Oceanography Coordinated ture, salinity, oxygen, nutrients and ice declines, and which species Investigations), a joint research pro- zooplankton were also collected might be most vulnerable in a rap- gram between the Alaska Fisheries from ship-based surveys (Figure 2). idly changing ecosystem. Science Center and the Pacific Fish and crab data were obtained Lisa Sheffield Guy, University of Washington Joint Institute for Marine Environmental Laboratory, from the Alaska Fisheries Science the Study of the Atmosphere and Ocean Center trawl surveys (Groundfish Phyllis J. Stabeno, National Oceanic and Atmospheric Fig. 2 Assessment Program, Resource Administration (NOAA) Pacific Marine Environmental Laboratory Jeffrey Napp, NOAA Alaska Fisheries Science Center Assessment and Conservation Engineering Division). Whale data were from both visual surveys and The Bering Sea Project is a partnership between moored and shipboard acoustics. the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Project and the National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject

Results from the north-south transect line sampled in September 2008. Shown from top to bottom: temperature, salinity, chlorophyll-a, nitrate, and ammonium. The four vertical lines

through each panel indicate the positions of the Lindsey Wright Napp Jeffrey four moorings. Note the strong break in temperature and salinity near mooring M5 at roughly 60 Casting the CTD (conductivity, temperature, and EcoFOCI Scientists Nancy Kachel and Carol Ladd N. This is the feature that separates the northern depth instrument) in a sea of ice and jellyfish deploy a bongo net aboard the R/V Thomas G. and southern portions of the eastern Bering Sea. aboard the NOAA Ship Miller Freeman. Thompson.

Bering Sea Krill and the Impact of Climate Change SOME LIKE THEIR ALGAE ON ICE

We often hear about decreases in Fig. 1 ice coverage and thickness in polar regions, and we can see how climate change is affecting large organisms, like polar bears, but it also affects organisms toward the bottom of the food chain. Euphausiids, more commonly known as krill, are large shrimp-like zooplankton (Figure 1) that are an important food source for larger organisms, including fish, seals, seabirds, and baleen whales. In the eastern Bering Sea, ice typically covers much of the wide shelf in late winter through spring. In early spring, intense blooms of ice algae form in and on the bottom of the ice. There are also algae that thrive in the water (phytoplankton), but Krill species Thysanoessa inermis and T. longipes, separated from a bongo net tow collection taken on the outer shelf of the southeastern Bering Sea. they are present in low amounts when there is ice cover and begin to bloom only later, after the ice plankton from the same location. retreats. We wanted to determine During the incubations, light levels The Big Picture the importance of ice algae in the were adjusted to simulate the light Based on long-term observations, we krill diet and the possible effects of exposure to which the krill were know that krill are more successful in cold an ice-free springtime on krill. acclimated. years, when there is lots of ice, than in To examine the krill’s eating warm years. It is possible that the earlier- How We Did It habits, we measured the abundance blooming, possibly more nutritious, ice We worked primarily at night, and type of plankton—includ- algae kick-start and bolster krill growth when krill migrate to the surface ing ice algae, phytoplankton, and and reproduction. If ice cover and extent water. Krill were captured with the very small microzooplankton diminish in the Bering Sea in the future, a bongo net (Figure 2) that was (protozoa)—before and after krill some species of krill may be less produc- towed behind our research vessel, were allowed to feed. We were able tive, while other species that are not as and then incubated in water con- dependent on ice algae as a food source taining the natural community of continued on page 2 may be less affected.

THE TROPHIC ROLE OF EUPHAUSIIDS IN THE EASTERN BERING SEA: ECOSYSTEM RESPONSES TO CHANGING SEA-ICE CONDITIONS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. to determine what they ate, how as a food source, and the chemical Why We Did It much, and whether or not they signatures of the ice algae are seen Understanding the impact of preferred one type of plankton over as increased lipids within the krill. climate change on krill in the another. We also analyzed the lipid The more oceanic counterpart of T. Bering Sea is important because (fat) profiles in both natural water raschii, namely T. inermis, rely more they are a dietary staple for many samples and in the krill themselves on planktonic algae and micro- organisms in that region and a rich before and after feeding. Different zooplankton. Changes in sea ice source of calories as lipids. Our types of plankton have different dis- extent and duration in the Bering feeding experiments showed that tributions of lipid structures. These Sea, with consequent changes in ice the most abundant species of krill, can be used to track specific groups algae availability and concentration, Thysanoessa raschii, devoured ice of plankton, such as ice algae, may affect the reproductive success algae, when available, at very high through the food chain. Finding of the regional T. raschii population rates, up to five times the rate at these compounds in krill tells us to a greater extent than T. inermis. which they could consume phyto- what was eaten and stored in their Plankton samples were collected plankton. After the ice melted back bodies. The total lipid amount is and taken to the lab, where the and planktonic algae bloomed in an important source of calories for different quantities and types of the water column, T. raschii fed on those that eat them. plankton were counted. Subsamples planktonic algae and microzoo- Through a series of shipboard of the water collected for incubation plankton. Not all species of krill are incubations, we were able to were analyzed for lipid biomark- alike, though. Thysanoessa inermis analyze changes in plankton con- ers. At the end of each experiment, typically lives farther offshore than centration before and after krill the krill were frozen and brought T. raschii. In this region of the grazing. We have seen that some to the lab, along with samples of Bering Sea, ice is not as extensive krill—Thysanoessa raschii, in par- the plankton collected on filters for in late spring as it is inshore, and ticular—rely heavily on ice algae biochemical analysis. therefore ice algae are less available. Not surprisingly, we found T. Fig. 2 inermis fed primarily on planktonic algae and microzooplankton.

H. Rodger Harvey, Old Dominion University (ODU) Ocean, Earth, and Atmospheric Sciences Evelyn J. Lessard, University of Washington (UW) School of Oceanography Rachel Pleuthner, ODU Ocean, Earth, and Atmospheric Sciences Megan Schatz, UW School of Oceanography Tracy Shaw, Hatfield Marine Science Center, Oregon State University

The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Project and the National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject

Bongo net deployment at dusk during a summer research cruise.

Climate and Bering Sea Fisheries: Beyond a Northward March SURPRISING IMPACTS ON BERING SEA POLLOCK AND PACIFIC COD FISHERIES

While some other global scale How We Did It research has suggested that a warm- A significant component of our The Big Picture ing climate will propel marine research has been focused on iden- The BEST-BSIERP Bering Sea Project species northward, our work has tifying the mechanisms by which recognized from its outset that humans demonstrated that for the biggest climate impacts fisheries. We use are an important component of the fisheries in the Bering Sea, this has data on fishing locations, fish and ecosystem, and that we cannot under- not occurred as expected. For pol- fuel prices, and how these interact stand the system without understanding lock between 1999 and 2009, the with biological survey information how they use and adapt to the changing fishery shifted northward in the sum- and environmental data. After col- environment. By examining the response mer, but this occurred in cold years lecting data and talking to fisher- of largest Bering Sea fisheries to the more than warm years. Similarly, for men, we used a variety of statistical changing environment, we have illus- Pacific cod, a larger cold pool (where methods to see how management, trated that people will not respond in a bottom water temperatures are below changing prices, and changing bio- simple manner to the changing environ- 2°C) in cold years has led to fish logical and environmental measures ment. A better understanding of how the fishery behaves in warm, low-abundance being more concentrated in northern have impacted the fisheries (Figures years will help inform how the fishery areas and consequently to more fish- 2 and 3). continued on page 2 will react in the future. Managers can use ing in those areas (Figure 1). this information to better anticipate how Fig. 1 fisheries will interact with other parts of the ecosystem, which can contribute to better-managed fisheries.

The Eastern Bering Sea and the fishing areas of the catcher–processor fleet. Points represent the catch-weighted mean center of the distribution of fishing hauls by season. Note the large distinction in the movement of the fishery over time that occurs in the summer fishery B season as well as the lack of movement in the winter fishery A season. [From Haynie, A. and L. Pfeiffer. 2013. “Climatic and economic drivers of the Bering Sea pollock (Theragra chalcogramma) fishery: Implications for the future.” Canadian Journal of Aquatic and Fisheries Science. 70(6): 841-853, 10.1139/cjfas-2012-0265.]

SPATIALLY EXPLICIT INTEGRATED MODEL OF POLLOCK AND COD A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. From this research, we have has variation in climate conditions Fig. 3 seen that abundance and environ- affected the spatial extent of Bering mental conditions both directly Sea fisheries? How do we expect impact where the fisheries occur. predicted changes in future climate Other BSIERP work has indicated to impact fisheries and fishing that we are likely to see more low- communities? Informing decision- abundance years with a warming makers on how climate and fisheries climate (Mueter et al., ICES 2011), are interacting is essential to the but in recent times, warm years have effective management of marine also been high-abundance years. As resources in the future. The deci- shown in Figure 3 and discussed in sions that managers make now the Haynie and Pfeiffer ICES 2012 will impact the welfare of fishers, article referenced in Figure 2, we communities, the nation, and the have not yet experienced a likely ecosystem over the next century. future state of warm, low-abundance Summary of the effects of the size of the cold pool Alan Haynie, National Oceanic and Atmospheric Administration conditions. (NOAA) Fisheries, Alaska Fisheries Science Center (AFSC) and total pollock abundance on the intensity of Lisa Pfeiffer, NOAA Fisheries, AFSC early A-season (winter season) effort, B-season Why We Did It (summer season) catch per unit effort (CPUE), B-season effort, and B-season travel costs. Years Fishers are the apex predators The Bering Sea Project is a partnership between in the sample characterized by varying abundance of the Bering Sea ecosystem, and the North Pacific Research Board’s Bering Sea and cold pool levels are listed on the horizontal and their spatial behavior can tell us a Integrated Ecosystem Research Program and the vertical axes. [Also from Haynie and Pfeiffer CJFAS great deal about the way in which National Science Foundation’s Bering Ecosystem 2013, referenced above.] fish populations are shifting under Study. www.nprb.org/beringseaproject changing climate conditions. After controlling for other factors, how

Fig. 2

A catch of pollock on the NOAA ship Miller Free- man, a fisheries oceanography vessel that works A conceptual model of how the environment affects the distribution of fishing effort, including the total predominantly in the Bering Sea and the North allowable catch (TAC) and the cost per unit effort (CPUE). Arrows represent the direction of causality, Pacific Ocean. and dotted lines represent mechanisms that may occur on a non-contemporaneous time scale. [From Haynie, A. and L. Pfeiffer. 2012. “Why economics matters for understanding the effects of climate change on fisheries.” ICES Journal of Marine Science, 69 (7): 1160-1167, doi:10.1093/icesjms/fss021.]

Seasonal Bioenergetics in the Bering Sea THE FATTER THE BETTER Being fat is good when you have big and fat by eating fatty prey, and of heat produced directly reflects to survive a long winter with very we found that fatty prey were most its calorie content. We applied this little food to eat. This seems obvious, abundant when conditions in the method to samples of fish and their but it is one of the most interesting Bering Sea were cooler, with longer prey collected from the Bering Sea and important discoveries we have lasting sea ice, rather than when they between 2003 and 2010. During this made from our studies of juvenile are warmer (Figure 1). period, the Bering Sea underwent a pollock in the Bering Sea. Winter shift from “warm years” character- is a time when food is scarce for How We Did It ized by an early sea ice retreat to juvenile fish that must use their fat Fat fish have more calories per “cool years” characterized by late sea stores to survive. We found that unit weight than lean fish. We used ice retreat. When we compared the fish that are fat at the beginning of a method called bomb calorimetry calorie content of the fish and their winter survive better than those that to measure the number of calories prey to these different climatic con- are lean. Apparently, the more fat in a fish. Essentially, we dry the fish, ditions, we saw fish and their prey they have the better, because small put it in a machine that sets it on fire were leaner in warmer years than in fat fish do not survive as well as big and then measures how much heat cooler years. We were also able to fat fish. We also realized that fish get is produced (Figure 2). The amount continued on page 2 Fig. 1

The Big Picture We found a similar story for Pacific cod as we did for pollock. This suggests that a warming Bering Sea is likely to produce less protein for us to consume, or that the protein we harvest from the Bering Sea may have to come from new sources. Predicting how climate change will influence the Bering Sea ecosystem was a primary goal of the Bering Sea Project. The observations we made were consistent with an overall picture that the organisms we depend on from the Bering Sea have evolved life history strategies that rely on the presence of ice in spring. As the Bering Sea warms and ice retreats earlier and earlier, juvenile This is the amount of energy in a gram of juvenile pollock tissue (measured in kilojoules per gram) in each of the years we have surveyed the Bering Sea. The red symbols show the energy content in warm years and the blue symbols forms of the species we depend on will find show the cool years. It is clear that the energy content of pollock has changed between warm and cool years. it more and more difficult to survive.

SEASONAL BIOENERGETICS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. compare the total number of calo- So understanding how climate the availability and quality of their ries in the fish to the number of fish affects fisheries can be thought of food, the rate at which they use that survived and were eventually as a question of food security for energy for daily activities, and how caught in fisheries. We discovered our country. We believe that the much food their predators need. that the years that produced big and impacts of climate on fisheries and Ron Heintz, National Oceanic and Atmospheric Administration fat juveniles were the same years that fish populations are most discern- (NOAA) - Alaska Fisheries Science Center (AFSC), Auke Bay produced more fish for the fishery. ible among juvenile fish because Laboratories they must use energy to grow and Elizabeth Siddon, NOAA – AFSC, Auke Bay Laboratories Why We Did It avoid predation or store it to avoid The pollock fishery in the Bering starvation, especially over the win- The Bering Sea Project is a partnership between Sea is one of our largest fisher- ter. Climate has a profound influ- the North Pacific Research Board’s Bering Sea ies, and it represents an important ence on how fish deal with these Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem source of protein for the country. conflicting demands by influencing Study. www.nprb.org/beringseaproject

Fig. 2

This shows our calorimeter and all the various components. Fish (1) are ground up (2) and dried into a powder (3). A sample of the powder is pressed into a pellet (4) and loaded into the pellet holder (5), which has a fuse installed. The pellet and fuse are loaded into the bomb casing (6) and the casing is filled with oxygen and then placed in the water bath (7). Electrodes heat the fuse, which ignites the powder and generates heat that warms the water bath. A thermometer in the water bath records the change in temperature, and a computer converts the temperature change into calories.

Defining Ecological Regions in the Bering Sea SYNTHESIS OF PHYSICAL AND BIOLOGICAL DATA TO INFORM SPATIAL MANAGEMENT

Issues of scale are important in influenced by the scale of analy- understanding competitive and sis. To better understand species predatory interactions and the interactions, community responses influence of external drivers on to physical drivers, and ecologi- ecosystem dynamics. Our research cal dynamics at regional scales, we evaluated trends in species abun- illustrate a statistical approach to dance for evidence of compensation delineate distinct ecological regions (inverse trends), and synchrony within large marine ecosystems. (common trends). The former suggests competition, while the latter How We Did It Baker Matt suggests an external driver. We also We integrated time-series data on evaluated whether resource parti- species abundance and the physi- tioning occurs, which would suggest cal environment to better under- The Big Picture mutual avoidance to reduce com- stand factors influencing ecosystem Effective management of marine petitive interactions. stability and change in the Bering resources requires tools to identify structure in ecosystems and better Interpretations of these rela- Sea. Using random forest statistical understand processes at regional tionships and trends are strongly continued on page 2 scales. Our methods delineate ecologi- Fig. 1 cal regions on the basis of threshold shifts in the composition of biological communities along a suite of physical gradients. Maps that distinguish marine areas with distinct biogeog- raphy provide an important tool to better understand ecological processes and to facilitate conservation planning and spatial management. Previous efforts have characterized the eastern Bering Sea according to hydrographic patterns or expert-derived delinea- Threshold shifts in the abundance of multiple species along the gradient of select environmental predictions of regional differences. Our tors. Gray histograms display the relative importance of environmental values as a breakpoint for shift results build on this, applying a statis- in distribution for individual species (split values from random forests). Blue lines illustrate a com- tical approach with practical utility for munity turnover rate, estimated as the ratio of relative importance and relative density of breakpoints fishery management. aggregated over individual species (split importance to splits observed). This reflects a rate of change in the aggregate biological community. Peaks above the dashed line indicate locations of greater change in community composition.

FORAGE DISTRIBUTION AND OCEAN CONDITIONS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. methods, we quantified the rela- Why We Did It environmental thresholds for indi- tive importance and marginal effect To effectively manage marine vidual species, we are able to better of physical variables (temperature, resources, we need to understand estimate competitive and predatory depth, substrate and stratification) the relative impact of environmen- interactions in multispecies models. to species abundance, and identi- tal drivers on biological interactions Such analyses may also inform spa- fied critical thresholds. We then and processes at multiple scales. tial management, the importance of integrated results for individual We sought to develop a standard stock structure and the relevance of species to characterize how the dis- approach to identify regional sub- localized environmental drivers. tributions of multiple species shift structure in large marine ecosystems Anne Hollowed, NOAA Alaska Fisheries Science Center along the gradients of a suite of and to inform integrated studies Matthew Baker, Joint Institute for the Study of the Atmosphere environmental variables (Figure 1). of physical dynamics and biologi- and Ocean, University of Washington; NOAA Alaska Fisheries We also identified threshold shifts cal interactions at smaller scales. Science Center in the composition of the aggregate Species distribution (i.e., biogeog- biological community, and used raphy) and interaction are impor- The Bering Sea Project is a partnership between these to delineate distinct regional tant drivers of ecosystem function the North Pacific Research Board’s Bering Sea boundaries (Figure 2). and structure. By identifying Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject Fig. 2

Principal component plots of stations in the bottom trawl survey (left) where coordinate position reflects the inferred biological community associated with environmental predictor gradients (arrows) and color refers to distinct ecoregions. Individual survey stations were grouped via clustering methods to delineate distinct ecological regions in the eastern Bering Sea (right). The top row of plots display weighted species abundance per station for select species, demonstrating how individual species respond to multiple environmental variables.

Climate, Population Dynamics and Predator-Prey Overlap ARROWTOOTH FLOUNDER VS. JUVENILE POLLOCK Climate- and human-induced their habitat. In addition, flounder changes in marine ecosystems have and pollock distributions are influ- detectable impacts on the spatial enced by water temperatures and the distributions of fishes. However, less location of the cold pool of subsur- is known about how shifts in distri- face water that forms across the con- butions might alter predator-prey tinental shelf with the formation and overlap and the dynamics of prey melting of winter sea-ice (Figure 1). populations. Our study revealed that Our findings contribute to the grow- population size and ocean tem- ing evidence that continued increases peratures have a synergistic effect in flounder abundance combined Olav Ormseth Olav on the strength of overlap between with warming ocean temperatures Juvenile pollock. arrowtooth flounder (predator) and could translate into higher predation juvenile pollock (prey) in the eastern mortality on juvenile pollock in the Fig. 1 Bering Sea. Predicted changes in eastern Bering Sea. overlap strength occurred largely as a consequence of flounder movement. How We Did It This result was expected because the The Bering Sea is an ideal sys- abundance of flounder has increased tem to examine the ecological eight-fold over the past three consequences of changes in species decades, prompting expansion of continued on page 2

The Big Picture The potential for changes in species distributions and interactions is pronounced in the Bering Sea ecosystem and subarctic systems in general. Not only is climate-induced habitat variability especially strong in these regions, but some of the largest commercial fisheries in the northern hemisphere are found in these waters. Using existing assessment survey data, we examined species abundance, distribution, and interactions to gain insights into predator-prey overlap. Our methodology provides the ability to characterize the dynamics of species interactions and quantify the impact of predators on prey under different scenarios. This methodology is particularly valuable for understanding ecological processes in the Bering Sea, as it improves our ability to anticipate shifts in predator-prey relationships involving key species such as arrowtooth flounder and pollock. Study Region and Cold Pool. Summer survey bottom temperatures (ºC) in the eastern Bering Sea during a cold year (A; 2007) and warm year (B; 2003). The 50 m, 100 m and 200 m depth contours are shown.

CLIMATE VARIABILITY AND DEMOGRAPHY DICTATE STRENGTH OF PREDATOR-PREY OVERLAP A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. distribution. Like other heavily har- Why We Did It juvenile stages combined with other vested systems, the Bering Sea is the Better knowledge of the mecha- top-down and bottom-up pressures focus of intense assessment surveys nisms that influence the strength on the survival of pollock early life aimed at estimating species abun- of species overlap can improve our stages could have important ecologi- dance, distribution and predator- ability to anticipate shifts in preda- cal and economic consequences. prey interactions. The survey data tor-prey relationships and forecast Mary Hunsicker, College of Earth, Ocean, and Atmospheric are therefore valuable for improving ecosystem-level effects of chang- Sciences (CEOAS), Oregon State University (OSU) our understanding of species spatial ing environmental conditions. For Lorenzo Ciannelli, CEOAS, OSU dynamics and ecological interac- Kevin Bailey, Alaska Fisheries Science Center (AFSC), National harvested species, understanding the Marine Fisheries Service (NMFS) tions in subarctic ecosystems. Using magnitude and variability of natural Stephani Zador, AFSC, NMFS this data, we first characterized pol- mortality can be important for set- Leif Christian Stige, Centre for Ecological and Evolutionary lock and flounder distribution and ting realistic harvest goals. Further, Synthesis, University of Oslo then predicted their overlap in rela- the potential impact of the grow- tion to further increase of flounder ing flounder population on pollock The Bering Sea Project is a partnership between biomass and ocean temperature. We population dynamics has become a the North Pacific Research Board’s Bering Sea found that the predicted changes in real concern. Alaska pollock provide Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem overlap at higher temperatures were sustenance for many species of com- Study. www.nprb.org/beringseaproject greater in years of high flounder mercial and conservation value, and biomass (Figure 2a) compared to support the world’s second largest years of low flounder biomass single-species commercial fishery. (Figure 2b). Increased predation by flounder on

Fig. 2

Predicted changes in species overlap with increasing temperatures. Using a a standardized unit increase in spatially-explicit bottom temperatures, red circles indicate locations where the probability of species overlap is predicted to increase, and black circles indicate decrease . The effect of temperature on the magnitude of species overlap was amplified by high flounder biomass. For example, in years when the flounder stock size and temperatures were high (A: 2005), there were large increases in overlap in the northwest shelf and throughout most of the middle and southeast shelf regions. However, when flounder biomass was low (B: 1987), the change in overlap with an increase in temperatures was mostly to the 100 m isobath in the north shelf region. The 50 m, 100 m and 200 m depth contours are shown.

Subsistence Food Comes from a Vast Area! THE BIG PICTURE OF PRODUCTION Subsistence hunters and fishers How We Did It are drawing on a vast area of the We began with subsistence The Big Picture ocean. Much attention has been harvest records, which told us the An issue of great interest to researchers in the Bering Sea Project is how the given to subsistence use areas, where species that have the largest har- ecosystem affects people. Analysis of people hunt and fish. We also looked vest by weight. Then we looked at subsistence harvests tells us a great deal at the areas of the ocean that help the biological data to identify the about direct human interactions with the produce those fish and animals. We species for which good distribution ecosystem. Looking at “calorie-sheds,” areas called this the “calorie-shed,” the and range data are available. That of the ocean that produce fish and animals area that contributes to the food that gave us good information about sought for food, gives us another way of ends up on people’s plates. Using three important subsistence species understanding how the ecosystem matters subsistence harvest records to iden- in each community. We also had to people. In short, coastal residents have a tify important species, we then used to be specific about the location of great deal at stake when it comes to ecosys- biological data to establish how far interest. For example, when show- tem well-being. This interest extends across those species range from the commu- ing the range of the salmon that the entire region, not just in the areas where nity or area where they are harvested. are harvested in Togiak, we did not people travel, hunt, and fish. Calorie-sheds give us another way of considering how We did this for Togiak (Figure 1) and want to include the full range of changes in the ecosystem may affect the Savoonga (Figure 2), using three spe- all salmon, but only the range of people who are part of that ecosystem. cies for each village. It turns out that the areas are huge! continued on page 2 Marla Greanya

SUBSISTENCE HARVEST AND LTK ECOSYSTEM PERSPECTIVE A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 1 salmon that return to Togiak, or at least to the Bristol Bay region. Note, too, that these “calorie-sheds” only include the species harvested, not the animals and plants farther down the food web. So, in fact, these are minimum areas.

Why We Did It The calorie-shed idea came from wondering about the full geographic extent of people’s interactions with the Bering Sea ecosystem. Subsistence use areas show where people go, but there is more to their use of the ecosystem than that. By showing how much of the ecosystem they draw on, we can also show why an individual community might be concerned about what is happening far away. If those distant activi- Calorie-shed for Togiak, based on the distribution of pink salmon, sockeye salmon, and harbor seal. ties affect the fish, birds, or marine mammals that the community relies on, then they would clearly be inter- Fig. 2 ested in what was taking place. In the future, it may also be possible to look at changes in the calorie-shed in light of environmental change, and better understand the implications of change for subsistence communities.

Henry P. Huntington, Eagle River, Alaska Ivonne Ortiz, School of Aquatic and Fishery Science, University of Washington George Noongwook, Savoonga Whaling Captains Association Maryann Fidel, Aleut International Association & University of Alaska Anchorage (UAA) Dorothy Childers, Alaska Marine Conservation Council (AMCC) Muriel Morse, AMCC Julia Beaty, AMCC Lilian Alessa, Resilience and Adaptive Management Group (RAM Group), UAA Andrew Kliskey, RAM Group, UAA

The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem Calorie-shed for Savoonga, based on the distribution of bowhead whales, Pacific walrus, and pink salmon. Study. www.nprb.org/beringseaproject

Local and Traditional Knowledge of the Bering Sea Ecosystem DETAILS MATTER!

The Bering Sea is a complex and of “hot spots,” or areas with very changing ecosystem. In the south- high productivity. Around St. east, many species are in decline. Lawrence Island, hunters noted In the north, it remains a produc- several such locations, all of which tive ecosystem with abundant are still productive, attracting an fish, seabirds, and marine mam- abundance of fish, seabirds, and mals. The most rapid changes are marine mammals (see Figure 1). occurring at the edge of the sea ice Overall, the results from local and maximum, in the southern Bering traditional knowledge (LTK) are Sea. Ice-associated species, such as consistent with other findings from Caleb Pungowiyi (gray shirt) and Chester Noong- bearded seals, are becoming scarce. the Bering Sea Project. wook (red shirt) discuss LTK over a map of St. Lawrence Island. Ice conditions are also changing in the northern Bering Sea, so we How We Did It were surprised to find that hunters We interviewed experienced reported a thriving ecosystem. Of hunters and fishers in five Bering The Big Picture particular interest were descriptions continued on page 2 In search of local and traditional knowledge (LTK), we interviewed hunt- Fig. 1 ers and fishers from several Bering Sea communities. What they shared shed light on several aspects of the Bering Sea Project’s research. First, broad differences between the southern and northern Bering Sea had been noted in several other analyses of the ecosystem, and these differences were confirmed with LTK, supporting this interpreta- tion. Second, observations about increased summer storms were contrary to the decrease that was predicted in the Bering Sea Project hypotheses, raising interesting questions for further study. Third, changes in abundance and distribution of species did not follow a simple pattern across the Bering Sea, but showed great local variation, indicating that the ecosystem is complex. Specific locations associated with particular ecological features/actions in the vicinity of St. Lawrence Island as reported by Savoonga LTK participants.

SUBSISTENCE HARVEST, USERS AND LTK ECOSYSTEM PERSPECTIVE A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Henry Huntington Henry Huntington Caleb Pungowiyi demonstrates how to net auklets Caleb Puagowiyi holds a least auklet near Savoonga. from an old blind near Savoonga.

Sea communities: Akutan, St. Paul, Why We Did It oceanographers, climatologists, Togiak, Emmonak, and Savoonga. People who live on the shores biologists, and others are finding. We discussed many aspects of the of the Bering Sea, especially those Henry P. Huntington Bering Sea ecosystem, especially who spend a lot of time hunting Nicole M. Braem, Alaska Department of Fish and Game (ADFG), those related to the hypotheses and fishing, have a deep under- Division of Subsistence driving the entire project. Most of standing of the environment. In Caroline L. Brown, ADFG, Division of Subsistence the interviews were open-ended Native villages, this knowledge Eugene Hunn, University of Washington Theodore M. Krieg, ADFG, Division of Subsistence discussions, closer to a conversa- may have been accumulated Pamela Lestenkof, Aleut Community of St. Paul Island Tribal tion than to a poll or a question- over many generations, allowing Government and-answer session. After the people to hunt and fish success- George Noongwook, Savoonga Whaling Captains Association Jennifer Sepez, National Oceanic and Atmospheric Administration interviews, we wrote down what fully and safely. By documenting (NOAA), (retired) we had heard, and reviewed our what Bering Sea residents know Michael F. Sigler, NOAA report with the hunters and oth- about their ecosystem, we can Francis K. Wiese, North Pacific Research Board, Anchorage Philip Zavadil, Aleut Community of St. Paul Island Tribal ers in the communities. Then we learn important local details about Government made any necessary corrections ecological processes and changes. and other adjustments before shar- And we can also check what we ing the results within our group have learned from other types of The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea and with the Bering Sea Project studies by comparing what local Integrated Ecosystem Research Program and the researchers. residents are seeing with what National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject

Changing Wind and Ice Conditions in the Bering Sea WHEN DO WALRUS HUNTERS CHOOSE TO STAY HOME? Fig. 1

The biggest factor in walrus How We Did It hunting success is whether hunt- Our analysis considered wind ers go hunting. This is not sur- speed, wind direction, and sea prising—but what makes walrus ice concentration in relation to hunters sometimes choose to walrus hunting from Gambell and stay home? Changing wind and Savoonga, on St. Lawrence Island ice conditions can affect hunting (Figure 1). We used those vari- success, but hunters are used to ables to see how they affected the dealing with variable conditions. number of hunting trips that were We wondered if conditions attrib- made and the number of walrus utable to climate change may that were harvested. affect walrus hunting by affecting First we compiled daily data on Map of St. Lawrence Island and the eastern decisions about whether or not to walrus harvest, number of hunting Bering Sea, showing the communities of Gambell and Savoonga. go hunting. continued on page 2

The Big Picture Bering Sea Project researchers are very interested in how the ecosystem is changing and what those changes mean, especially for people who depend on the Bering Sea for food and for their livelihood. Changes in sea ice are a prominent part of ecosystem change in the region. By examining the impact of changing sea ice, along with winds, we were able to show that walrus hunters on St. Lawrence Island may indeed be affected by those changes, but also that other factors may be more important, such as the skill and experience of the hunters, who are accustomed to dealing with variability and are quick to adjust and adapt as needed. Marcus Janout Marcus

SUBSISTENCE HARVEST, USERS AND LTK ECOSYSTEM PERSPECTIVE A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. trips, wind speed, wind direction, explained by wind and ice condi- that both together affect how well sea ice concentration, and visibility tions. While other factors com- they are able to hunt walrus. We for both Gambell and Savoonga. bine to explain much more of the wanted to test that idea, and also Then we analyzed these data using variability, wind and ice conditions to see if we could understand the a “generalized additive model.” do matter. Our analysis also helped relationships between those physi- This method allows us to model explain how they matter, in other cal factors and walrus hunting. By several parameters together to pre- words, how a change in wind or ice doing so, we may be able to under- dict an outcome, so we used wind would affect hunting. For example, stand better how changes in climate speed, wind direction, and sea ice higher winds make boating more can affect walrus hunting. concentration at various distances dangerous and difficult, so hunt- from the villages to see what influ- ers tend to stay on shore when it Henry P. Huntington George Noongwook, Savoonga Whaling Captains Association ence they had, individually and is too windy. Similarly, too much Nicholas A. Bond, Joint Institute for the Study of the Atmosphere together, on hunting outcomes. ice makes boat travel difficult, but and Ocean, University of Washington (UW) We also considered visibility, with too little ice can mean there are few Bradley Benter, Marine Mammals Management, U.S. Fish and Wildlife Service (USFWS) the expectation that foggy condi- walrus since the walrus like to haul Jonathan A. Snyder, Marine Mammals Management, USFWS tions were not good for hunting, out on ice; or too little ice can allow Jinlun Zhang, Applied Physics Laboratory, UW but found that the addition of vis- waves to build much higher, again ibility to the model did not appear making it dangerous for hunters. The Bering Sea Project is a partnership between to be much of a factor. the North Pacific Research Board’s Bering Sea One-quarter to one-third of the Why We Did It Integrated Ecosystem Research Program and the variability in the number of hunt- Hunters in Savoonga told us that National Science Foundation’s Bering Ecosystem ing trips that were made could be wind conditions affect sea ice, and Study. www.nprb.org/beringseaproject Henry Huntington Thomas Van Pelt Van Thomas George Noongwook leads a discussion of traditional knowledge in Savoonga. George Noongwook driving his skiff along the north shore of St. Lawrence Island, west of Savoonga.

North to the Arctic ALBATROSSES INCREASE IN THE BERING SEA

Fig. 1 All three species of North Pacific albatrosses are now SHORT-TAILED ALBATROSS BLACK-FOOTED ALBATROSS found in greater abundance and LAYSAN ALBATROSS found farther north than in the 1970s–1990s. The increase in sightings of short-tailed albatross (Phoebastria albatrus), an endangered species, is good news for conservation—and knowing where they go to forage can help us manage their interactions with fisheries in the future. Sophie Webb Sophie Webb Zeller Tamara

How We Did It Counting birds at sea from a variety of research vessels has resulted in over 140,000 km of survey transects in Alaskan waters, extending from the 1970s through the 2000s. Work carried out as part of the Bering Sea Project allowed us to extend at-sea sur- Current distribution of the three albatross species during recent years, primarily during the Bering Sea Project veys in 2008-2010 (Figure 1), and surveys (2008-2010). Red dots are scaled to indicate number of albatross observed; faint gray lines indicate supported examination of decadal survey effort. changes in seabird distribution (Figure 2). We looked at albatrosses The Big Picture because they are large, conspicu- Albatrosses prefer to forage on squid, which may have increased in the Bering Sea. There ous birds, easy to count at sea, and is a close overlap in the distribution of albatrosses (Figure 1) and squid (Figure 3) along the they are near the northern ‘fringe’ Aleutian Islands, the Bering Shelf and near shelf canyons. Additionally, a northward shift in of their ranges, which makes it fisheries could draw some vessel-following birds, like albatrosses, farther north. Such broad- relatively easy to notice changes. scale changes in distribution of an apex predator are indications of ecosystem-level change. Mapping densities of each By having a better understanding of the changes in albatross distribution, and the mechanisms albatross species over four decades driving those changes, we can work with commercial fishers to reduce detrimental interactions revealed increases and northern between albatrosses and fisheries. expansion in all three species continued on page 2

SEABIRD BROAD-SCALE DISTRIBUTION A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. (Figure 2). The short-tailed alba- Why We Did It Fig. 2 tross may be reclaiming its former One of the Bering Sea Project’s range, since it historically occurred predictions is that climate change will LAYSAN ALBATROSS in the Bering Sea and is recover- alter prey distributions, which will ing from near-extinction. But ultimately alter distributions of apex short-tailed albatrosses may even predators. The increase in albatrosses be checking out the Arctic—the in the Bering Sea could be one such BIRDS PER KM first albatross ever recorded in the example. It could also lead to more southern Chukchi Sea (in August, interactions with fisheries in the future. 2012) was a short-tailed albatross. The long-term dataset of the North The more common Laysan alba- Pacific Pelagic Seabird Database, com- 2 tross (Phoebastria immutabilis) has bined with the Bering Sea Project, has also been increasing, especially in allowed us to look at relative abun- the Aleutian Islands, and is now dance of seabirds over decades. commonly encountered along the entire shelf break. The black-footed Kathy J. Kuletz, U.S. Fish and Wildlife Service Martin Renner, Tern Again Consulting SHORT-TAILED ALBATROSS albatross (Phoebastria nigripes), Elizabeth A. Labunski, U.S. Fish and Wildlife Service historically found in the Gulf of G.L. Hunt Jr., University of Washington Alaska, has increased in the Aleu- Olav Ormseth, NOAA Alaska Fisheries Science Center tian Islands, and since the 2000s has been foraging in the southeast- The Bering Sea Project is a partnership between BIRDS PER KM ern Bering Sea during late summer the North Pacific Research Board’s Bering Sea and fall, when waters are warmest. Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject 2

Fig. 3

BLACK-FOOTED ALBATROSS BIRDS PER KM 2

Assessment of the squid stock complex in the Bering Sea and Aleutian Islands. Decadal changes in density for the three albatross Mean Catch Per Unit Effort (CPUE) of all squid species taken in NOAA trawl species found in Alaska. The 1970s, 1980s, 1990s, surveys, 2000-2012, in 20 x 20 km cells. and 2000s are shown in each of the four small panels, with data binned into 50 x 50 km cells. White cells indicate where surveys were conducted but no birds were recorded on transect.

Climate Change Could Stress Kittiwakes and Other Seabirds MODELING TO UNDERSTAND LIMITS IN ADAPTIVE BEHAVIOR

For black-legged kittiwakes Pacific kittiwakes on a region-spe- control nests compared with those (Rissa tridactyla), mortality increases cific basis, and the longevity of these that grew in food-supplemented with increasing levels of stress hor- birds may not always be sufficient nests. Modeling results further mones, and strong relationships exist to buffer their populations from low showed that in some colonies it between indices of environmental reproductive performance. appears birds sacrificed more life- variation and stress hormones. time reproductive success than a These relationships indicate that What We Found prudent parent would, and that less anticipated climate warming might Using our collaborators’ experi- food early in life led to first breeding bring at least short-term demo- mental manipulation of food avail- at a younger age, as well as greater graphic benefits for kittiwakes in the ability during early development, we reproductive effort, compared to Bering shelf region, while having discovered first breeding at younger birds reared with more food. negative impacts on birds breeding ages for kittiwakes that experienced Although we found a positive in the Gulf of Alaska and western suboptimal natal conditions, as correlation between warmer ocean Aleutians. Thus, climate variability well as greater productivity of early waters and higher productivity for is likely to affect survival of North recruiting kittiwakes growing in continued on page 2

The Big Picture Many different populations in the Bering Sea are increasingly likely to experience climate-induced changes in their physical and biological environments. Since adult kittiwakes are central place foragers with high energy requirements, an increased variability of forage patch dynamics, as predicted for polar regions, is likely to influence both the quantity and quality of food available. This would consequently alter the population dynamics of kittiwake colonies, mitigated by stress hormones rising in response to food shortages,

Chris Barger Chris with consequent effects on survival and Black-legged kittiwakes nesting on St. Paul Island. Several chicks are visible in nests, for example at lower left reproduction. and middle left of the photograph..

BEHAVIORAL AND LIFE HISTORY MODELING OF SEABIRDS IN THE NORTH PACIFIC A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 1 if climate change will push animals beyond these limits. If one thinks that responses to environmental variation are fixed and inflex- ible (left panel of Figure 1), then the limits are hard boundaries, but when behavioral flexibility allows animals to adjust to changing environments (right panel of Figure 1)—which is true, even in ‘simple’ animals—then the limits are more complicated to understand. Characterizing them cannot be done observationally and is Behavior is the first response to changing environment. Left hand panel: When organisms are assumed to very difficult to do experimentally. have fixed, stereotyped responses to food availability, the effects of the environment on population demog- That is, we cannot wait until the raphy (growth, survival, and reproduction) are linear (but may still be complicated). Right hand panel: On the other hand, if organisms have flexible responses through foraging and reproductive behavior, the links climate changes and then see how between climate, food availability, and demography become more intricate and less linear. The objective of animals respond if we want to our work was to explicate these linkages using state-dependent life history theory. have a chance of mitigating effects; even if we could do feeding experiments on caged animals, it would the colonies on Bogoslof Island and hormones. Through experimental be difficult to interpret them and the Pribilof Islands, a remaining manipulation of food availability, we scale them up. Modeling pro- puzzle is understanding how the studied aspects of reproductive per- vides a natural way for projecting regime shifts in the Northeastern formance associated with food avail- the behavioral responses and the Pacific of the late 1970s, and the ability. Population modeling helped boundaries beyond which changing associated changes in food, drove us determine whether the mortality climates will have seriously delete- colony declines in the Bering Sea, rates associated with persisting in a rious effects. while other colonies in the Aleutian breeding attempt despite high levels Authors: Archipelago increased in size. of stress hormones caused the birds Marc Mangel, University of California, Santa Cruz to sacrifice more lifetime reproduc- William H. Satterthwaite, University of California, Santa Cruz How We Did It tive output than they gain from Simone Vincenzi, University of California, Santa Cruz Our study combined mathemati- one year’s breeding. Modeling also cal models, statistical analysis and helped relate the effects of environ- Co-authors: Scott A. Hatch, US Geological Survey, Alaska Science Center experimental manipulation. We ment and energy resources on kit- Alexander S. Kitaysky, University of Alaska Fairbanks examined the statistical relationship tiwake growth, fledging age, survival John F. Piatt, US Geological Survey, Alaska Science Center between the stress hormone corti- from hatching to first breeding and costerone and the mortality of birds. productivity. The Bering Sea Project is a partnership between We also used statistical methods to the North Pacific Research Board’s Bering Sea test if inter-annual changes in the Why We Did It Integrated Ecosystem Research Program and the Pacific Decadal Oscillation, winter Although animals have evolved National Science Foundation’s Bering Ecosystem ice cover, or local sea-surface tem- to deal with environmental stress, Study. www.nprb.org/beringseaproject perature predict changes in produc- there are limits to their ability to tivity (fledglings per nest) or stress do so, and it is important to know

The Impact of Changes in Sea Ice Extent in the Eastern Bering Sea MODELING A LARGE-SCALE ECOSYSTEM RESPONSE TO GLOBAL OCEAN WARMING

Phytoplankton form the base of Fig. 1 the food chain in the sunlit ocean and support higher trophic levels, such as fisheries. Previous studies have linked changes in phytoplankton community to indices of natural climate variability (e.g., El Niño), but little is known about ecosystem responses to ocean warming. We used a combination of new field measurements and an ecosystem model to estimate changes in phytoplankton production and removal under actual cold and simulated warm years over the southeastern Bering Sea shelf. Using ecosystem model simulations, we observed that phytoplankton production over the Bering Sea shelf in warm years was only slightly higher than during cold years. Associated with this increased phytoplankton productivity was a Focus regions for our modeling simulations (panel A). Rates (mg C m-2 d-1) of phytoplankton primary production (panel B) and particulate carbon export flux (panel C) for warm (left plots) and cold (right simulated increase in export of phy- plots) years in each study region and for each 8-day average time window. toplankton material to the ocean floor (Figure 1). Simulated phytoplankton carbon export in both The Big Picture warm and cold years showed strong The Bering Sea supports one of the world’s most productive ecosystems and sustains a seasonal patterns; a result validated large fraction of the total U.S. fisheries harvest. The Bering Sea is potentially susceptible to by direct observation in the cold future climate change, but it is not known how, or to what extent, a warmer Bering Sea might years of the Bering Sea Project. alter the fate of phytoplankton production within the ecosystem and hence affect the yield of Phytoplankton carbon export was this important fishery. A key hypothesis of our Bering Sea Project is that climate change shifts low in the marginal ice zone (MIZ), the fate of organic matter from the pelagic to the benthic environment; and, further, that such external forcing on the ecosystem is highly dynamic, non-linear, and unpredictable. continued on page 2

THE IMPACT OF CHANGES IN SEA ICE EXTENT ON PRIMARY PRODUCTION, PHYTOPLANKTON COMMUNITY STRUCTURE, AND EXPORT IN THE EASTERN BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 2

Spring (a) and summer (b) export ratios determined during 2008-2010. Solid and dashed lines represent ice-edge maximum and minimum during spring sampling periods. Solid bars indicate trap-derived and open bars represent Thorium-derived export ratios. Colors: red (2008), blue (2009), and green (2010). unless the area was experiencing an their physiology. Using sequential Why We Did It active phytoplankton bloom, and ocean color images and a math- The combination of field measure- increased through late spring and ematical model constraining the ments and model analysis has led to early summer. This phytoplank- relationships between elements in an improved understanding of the ton carbon export to the benthos the ecosystem model, we estimated regional and temporal (warm vs. cold represented a significant fraction of the partitioning of organic carbon periods) variability in the magnitude primary production, i.e., the export between higher trophic levels and of phytoplankton production and its ratio (Figure 2). the ocean floor (Figure 1). These fate within the Bering Sea ecosystem. ecosystem model simulations were From this study, we are developing a How We Did It validated for cold years by com- more mechanistic understanding of We conducted spatially extensive parison to directly measured par- how carbon and energy flow through measurements of phytoplankton ticulate carbon export derived from the plankton community to commer- production, community structure, sediment traps (Figure 3). Other cially important species in a changing and associated particulate carbon carbon fluxes are currently being Bering Sea. export during spring and summer compared to distributions of data S. Bradley Moran, Graduate School of Oceanography, University of from 2008-2010. We supplemented collected by the Bering Sea Project Rhode Island (URI) this observational dataset with to better understand the partition- Michael W. Lomas, Bigelow Laboratory for Ocean Sciences phytoplankton production model ing of carbon within the Bering Matthew S. Baumann, Graduate School of Oceanography, URI simulations derived from spatial dis- Sea ecosystem so that the potential Roger P. Kelly, Graduate School of Oceanography, URI Chunli L. Liu, Marine College, Shandong University, China tributions of remotely sensed phyto- implications for the fishery may be plankton biomass and knowledge of assessed. The Bering Sea Project is a partnership between Fig. 3 the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject Images of Eastern Bering Sea plankton, common diatoms and dinoflagellates, (left panel) and the sediment trap used to capture them as they sink from the surface ocean (right panel).

Return of the Zooplankton RECENT COLD CONDITIONS A BOON FOR CRUSTACEAN ZOOPLANKTON Copepods (e.g., Neocalanus the food web linked to changes in Fig. 1 cristatus) and krill (e.g., Thysanoessa climate and ocean circulation? inermis) are miniature shrimp- like animals that are critical to the How We Did It diets of commercially valuable fish, Our approach was to analyze marine birds and cetaceans. They bottom-up (food supply) and top- are an essential link between the down (predation by fish) controls base of the marine food web and of LCZ standing stocks, including larger animals. But the population climate, physics, primary produc- of these large crustacean zooplank- tion, micro-zooplankton produc- ton (LCZ) in the Bering Sea varies tion, and predation, and to examine depending on ocean conditions. how LCZ production was parti- Associated with a change from warm conditions (2000- The population of LCZ crashed tioned among top predators under 2005) to cold conditions (2007-2010) was an increase in during a string of years with warmer varying climate scenarios. Because the number of Calanus copepods and krill on the eastern water (2000-2005), and has recov- the eastern shelf has different physi- Bering Sea shelf. Vertical bars represent the standard ered in recent years as water tem- cal domains (regions with different deviation of the data. peratures cooled (Figure 1). ocean properties) these questions What caused such a large swing were examined in defined regions in LCZ population? Was there of the shelf to elucidate how differ- The Big Picture insufficient food during the warm ences in water column structure and The Bering Sea shelf supports one of the world’s most productive fisheries and years (less phytoplankton and tiny mixing processes affect the flow of accounts for a large fraction of U.S. fisheries micro-zooplankton), or was there carbon and energy. landings. This system is highly susceptible more grazing from fish and mam- Using data from the past decade, to climate change, but our understand- mals keeping LCZ populations we examined spatial and tempo- ing of that susceptibility remains poor. low? And how are such changes in ral distributions of predator and In this study, we addressed several key prey fields, and the influence of Bering Sea Project hypotheses, includ- Fig. 2 climate and currents ing the influence of climate and ocean on those distribu- processes on food availability for fish and tions. Hypotheses mammals (bottom-up processes), and and questions were dynamic ecosystem controls from predation (top-down processes). We examined how continued on page 2 the presence or absence of sea ice over the eastern shelf in spring influenced the flow In cold years, krill were more of energy through the pelagic ecosystem abundant and more widely dis- in the eastern Bering Sea, particularly the tributed across the shelf compared distribution, standing stock, and trophic role to warm years as determined by of large crustacean zooplankton (LCZ). acoustic surveys of krill biomass.

IMPACT OF SEA-ICE ON BOTTOM-UP AND TOP-DOWN CONTROLS OF CRUSTACEAN ZOOPLANKTON AND THE MEDIATION OF CARBON AND ENERGY FLOW IN THE EASTERN BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. also addressed through integrated phytoplankton were smaller and eastern Bering Sea and its ability models and by expert panels at two micro-zooplankton were the major to support major fisheries. From interdisciplinary workshops. food source for LCZ. Energy flow this study, we hope to develop new We found that the spatial dis- through the ecosystem appeared mechanistic and conceptual models tribution of krill differed between to be different in warm and cold of carbon and energy flow, and to warm and cold years with greater conditions (Figure 4). In warm provide improved predictions of abundance over the shelf during years, the phytoplankton bloom the magnitude and fate of second- cold periods (Figure 2). This may occurred later, and sea ice and ice ary production in an ever-changing be the result of changes in ocean algal communities were less exten- Bering Sea. circulation as there was more sive. In cold years, algae growing on Calvin Mordy, University of Washington (UW) southward flow during cold years the bottom of the ice, and earlier Seth Danielson, University of Alaska Fairbanks that brought ice and colder water ice edge blooms, gave the LCZ an Lisa Eisner, National Oceanic and Atmospheric Administration over the southern shelf, which in early boost of food, helping sustain (NOAA) turn excluded some predators from egg production and survival of George Hunt, UW Michael Lomas, Bigelow Laboratory for Ocean Sciences the shelf. However, when using a juveniles. This may partially explain Jeff Napp, NOAA multivariate regression analysis of the return of LCZ during recent Patrick Ressler, NOAA predator-prey biomass, it did not cold years. Mike Sigler, NOAA appear that Walleye Pollock (Gadus Phyllis Stabeno, NOAA chalcogrammus), the major fish Why We Did It predator, exerted top-down control Results garnered from these The Bering Sea Project is a partnership between on krill populations (Figure 3). studies will provide a better under- the North Pacific Research Board’s Bering Sea In spring, phytoplankton and standing of regional and temporal Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem ice algae were the main food (seasonal, interannual) variability Study. www.nprb.org/beringseaproject source for LCZ, but in summer, in secondary production in the

Fig. 3 Fig. 4

Three scenarios of ice retreat and its influence on the timing of the spring phytoplankton bloom in the southeastern Bering Sea. If sea ice (blue) is present after mid-March (Scenarios A and B), a phytoplankton bloom (green) is present during sea ice retreat. If ice retreat is early (Scenario C), a spring bloom usually occurs in May.

Partial effects of pollock biomass and bottom temperature in a multivariate model (GAM) of krill biomass density. Taken together, the flat pollock curve and steep temperature curve suggest that krill abundance is greater at colder temperatures, but is not tightly linked to changes in pollock biomass, casting doubt on top-down control by predation.

Hungry Fish Make a Difference LINKING CLIMATE AND KRILL ABUNDANCE Many fish, seabirds, and whales Krill abundance is higher dur- feed on krill, but there is only so ing cold years and lower during much to go around. Every year, krill warm years. The amount of energy (or euphausiids) abundance peaks fish need to grow also changes with in late spring – early summer, and temperature. To grow the same bottoms out at the end of winter. amount, fish require less energy in Migrations and movement are tuned cold temperatures, more in warm to the seasons, but what happens temperatures, thus eating less krill in when there is overall less or more cold years and more in warm years. REEM/AFSC/NOAA krill, as can happen in cold and This creates large areas where krill is Hungry fish – warm temperatures increase fish warm years? Do fish make a notice- grazed down in warm years but not metabolism, meaning they eat more krill in warm able dent on the available krill? How in cold years, impacting krill preda- years, changing the availability of krill to other much and where? tors such as forage fish, seabirds, and predators throughout the Bering Sea shelf and slope. marine mammals (Figure 1). Fig. 1 continued on page 2 The Big Picture Forage fish are the link between zooplankton and many larger fish-eating predators such as large fish, seabirds and marine mammals. For example, walleye pollock is the single most abundant fish on the eastern Bering Sea shelf, with an estimated 6.5 million tons per year consumed by predators. It also supports a fishery of over 1,000,000 tons annually, with revenues upwards of 2 billion dollars. Keeping track of krill and forage fish response to different climate conditions, and the cascading effects on the food web, builds on our understanding of processes such as population growth, feeding grounds, “hot spots,” consequences of fat and skinny krill, as well as fishermen’s behavior. Combined with climate forecasts, it has the potential Euphausiid Biomass mgC/m2 to complement current conservation and management in the Bering Sea with more Average krill biomass in the eastern Bering Sea shelf and slope for 2004 (warm year) and 2008 (cold year) assuming proactive and strategic actions. zooplankton mortality is proportional to biomass (uncoupled) and linking a bioenergetics fish model (coupled).

FORAGE AND EUPHAUSIID ABUNDANCE IN SPACE AND TIME (FEAST) A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. How We Did It everything from oceanography to zooplankton or forage fish, it is We used a 3D model for ocean- plankton dynamics, fish numbers, important to understand how much ography, nutrients and plankton distribution, length, and weight. and where zooplankton (like krill) (NPZ) constructed for previous This requires a lot of calculations, is consumed by forage fish (small work, and we added data for several so we use a supercomputer, which fish like young pollock, capelin and species of fish at different lengths means we divide the whole region herring) year round and in multiple based on historical databases from into small squares and send them years. We wanted to quantify the the National Marine Fisheries out to 384 processors that talk to difference between assuming that Service and National Oceanic each other. One simulated year fish predation is proportional to and Atmospheric Administration. takes about 16 hours to run. krill biomass (uncoupled mode) Rather than assuming zooplankton versus using bioenergetics (coupled gets eaten in proportion to their Why We Did It mode) (Figure 2), and to measure biomass, we assumed it gets eaten In the eastern Bering Sea most of changes in the spatio-temporal according to fish energy needs or what we know about fish occurs in availability of krill (Figure 3). bioenergetics. We gave the differ- summer and early fall, and relates Ivonne Ortiz, University of Washington (UW)/ Alaska Fisheries ent types of zooplankton (such as to their feeding habits, species Science Center (AFSC), National Oceanic and Atmospheric krill) and fish values in calories, and abundance and their distribution. Administration (NOAA) then based the fish consumption We know very little about the rest Kerim Aydin, AFSC, NOAA Al Hermann, UW / Pacific Marine Environmental Laboratory, NOAA and growth on how many calories of the year, including interactions they ate and how they spent them with climate, winds, currents or on swimming, living and growing, zooplankton. We are now working The Bering Sea Project is a partnership between all of which is affected by tempera- on integrating oceanography with the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the ture. We then ran the model for zooplankton and fish dynamics. National Science Foundation’s Bering Ecosystem the entire Bering Sea, estimating Because many predators eat either Study. www.nprb.org/beringseaproject

Fig. 2 Fig. 3

Average krill biomass in the eastern Bering Sea shelf and slope for 2004 Variability in space and time of krill biomass in different regions of the eastern (warm year) and 2008 (cold year) assuming zooplankton mortality is Bering Sea shelf and slope as estimated for 2004 using the fish bioenergetics proportional to biomass (uncoupled) and linking a bioenergetics fish model to estimate predation on krill. model (coupled).

Observation Synthesis and High Resolution Numerical Modeling WHY IT’S NECESSARY IN THE BERING SEA

Over the past several years, the How We Did It Bering Sea Project has accumulated Two approaches were used in an unprecedented amount of ocean our research. The first combines a The Big Picture and sea ice observations. These have technique called Four-Dimensional There has been a significant increase been obtained in different seasons Variational Data Assimilation in operational in situ and satellite using different platforms (e.g CTD, (4DVAR), an existing numeri- observations during the last decade. This moorings, Argo floats, surface drift- cal ocean model, and an optimal creates the potential for more accurate hindcasting and forecasting of circula- ers, moorings). The rich collection interpolation algorithm used in the tion, water, and ice properties in the of Bering Sea Project in situ and sat- Bering Ecosystem STudy ice–ocean region. These capabilities are boosted ellite observations now provide an Modeling and Assimilation System by new and more sophisticated and excellent opportunity for synthesis, (BESTMAS). The technique is efficient data assimilation systems, through modeling and data assimi- based on the least squares fit of the based on high-resolution models that lation (DA), in order to improve model solution to observations. include biogeochemical components. The our understanding of the impacts It demands thousands of model resulting patterns enhance our ability to of changes in the physical forcings runs and significant computational understand and manage the rich eastern of the Bering ecosystem in response resources. However, as an advan- Bering Sea ecosystem. to climate change. Synthesis of tage, 4DVAR allows assimilation available data allows us to improve continued on page 2 estimates of the state of the Bering Fig. 1 Sea and to obtain dynamically balanced fields of all physical parameters. After assimilation, we will be able to quantify the volume, heat, 1 2 1 2 and salt transports over the eastern Bering Sea. High-resolution computer modeling will complement the DA efforts, providing a tool for studies of processes that influence transports, mixing, and hydrographic changes in the Bering Sea on temporal scales from hours (tidal) Snapshots of reconstructed sea surface height (cm; see color bar) and surface circulations in the Bering Sea to years (inter-seasonal). in 2008 (left: 2 January; right: 10 September). Thicker black arrows designate the locations of the moorings in the Bering Strait and the Eastern Bering Sea shelf. Numbers 1,2,3 designate the Anadyr and Spanberg straits and St. Lawrence Island, respectively.

OBSERVATION SYNTHESIS AND HIGH RESOLUTION NUMERICAL MODELING IN THE BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. of any type of observations. In our using the Bering Sea Project moor- Bering Sea shelf. case, the variety of assimilated data ing data along with Argo drifters The cold pool frontal boundary includes both in situ (e.g., tempera- and satellite SST and SSH. This (Figure 2) erodes during summer at ture, salinity, velocity) and satellite model allowed us to analyze vari- varying rates at different locations, (Sea Surface Height (SSH), Sea ability in flows through numerous leading to a highly corrugated fron- Surface Temperature (SST), ice con- passes of the Aleutian Island chain tal boundary. The Pribilof Islands centration, velocity) observations. and provided a dynamic picture of act as a region of enhanced cold Our results indicate intensification the erosion of the “cold pool” (an pool erosion due to tidal mixing. of the Bering Slope flow in winter area in the middle of the Bering Sea The highly variable rate of erosion and enhanced variability of circula- shelf where the dome of very cold of the pool and the characteristics tion over the Eastern Shelf during near-bottom water is capped by the of its boundary introduce ecologi- the summer and fall (Figure 1). surface layer of warmer water in cal patchiness, and potentially alter The second approach uses high- summer; Figure 2). the rate at which larvae and juvenile resolution numerical modeling. The fish can be transported or migrate dominant spatial scale of variability Why We Did It from recruitment regions near the in the Bering Sea (the scale of the Quantifying the currents in the shelf break to settlement zones on width of coastal jets and eddies) Bering Sea is important for depict- the inner shelf. may be as small as 20 km. Because ing accurate patterns of physical Gleb Panteleev, International Arctic Research Center (IARC), of the large area of the Bering Sea and biochemical parameters in the University of Alaska Fairbanks (UAF) and the computational cost of region. For example, strong flow Jacob Stroh, IARC, UAF resolving ocean features on these through the Spanberg Strait during Alex Kurapov, College of Oceanic and Atmospheric Sciences (COAS), Oregon State University (OSU) scales, most previous modeling the late summer and fall (Figure 1) Scott Durski, COAS, OU studies have failed to achieve suf- may significantly affect conditions Jinlun Zhang, Applied Physics Laboratory (APL), University of ficient resolution to represent many in the cold pool that are extremely Washington (UW) important phenomena. We devel- important for local biological oped a 2-km horizontal resolution production. Other, smaller-scale The Bering Sea Project is a partnership between model that describes ocean circula- motions, evident in our reconstruc- the North Pacific Research Board’s Bering Sea tion in the Eastern Bering Sea. This tions, may play a critical role in Integrated Ecosystem Research Project and the National Science Foundation’s Bering Ecosystem model is run for the ice-free period eroding the cold pool and providing Study. www.nprb.org/beringseaproject of July-October 2009. It exhibits nutrients from near the sea floor correct ocean behavior, as verified to the surface mixed layer of the

Fig. 2

This figure illustrates the seasonal retreat of the cold pool, from early July 2009 through early September 2009, in the model simulation. Contour lines indicate the position of the cold pool front, as indicated by locations where the 3.5 ˚C isotherm intersects the sea floor in the high-resolution Bering Sea model. Color indicates the date for which the contour line corresponds. Rapid erosion of the cold pool is observed especially early in the season and around the Pribilof Islands. Erosion along the edges of the cold pool is highly irregular and likely due to a combination of tidal and convec- tive mixing effects.

Distributions of Bering Sea Forage Fish THE MOVING MIDDLE OF THE FOOD WEB Forage fish, which include cape- How We Did It biological, and/or climate factors lin, herring, and the young life stages At sea, we used echosound- on forage fish distributions. Models of walleye pollock and Pacific cod, ers and trawling to map distri- varied by species but, in general, are food for many fish, birds, and butions of forage fish between temperature, bottom depth, and/or mammals in the Bering Sea. These BASIS (Bering Aleutian Salmon zooplankton prey were important predators are ecologically important, International Survey) survey sta- predictors of forage fish presence commercially valuable, and the focus tions in 2008-2010. The analysis and density. Interestingly, annual of traditional harvest. Evidence sug- was expanded to include existing variables, such as storminess in June gests that forage fish distributions acoustic data from 2006-2007. In and sea ice, were sometimes as or (vertically within the water column 2008, age-0 pollock were primar- more predictive than local condi- and horizontally across the Bering ily found in the surface water, less tions at a station. Sea) can change from year to year, than 35 m deep (Figure 1). In both and yet we don’t fully understand 2009 and 2010, highest densities Why We Did It why. Knowing that climate change were found in dense schools in the Environmental conditions (e.g., may impact the available habitat for midwater, more than 35 m deep temperature, salinity), the avail- forage fish, it is necessary to under- (Figure 2). Both age-0 Pacific cod ability of zooplankton prey, and stand the where, how many, and why and capelin had high densities in continued on page 2 of fish distribution to predict how the surface in 2010 as compared to changes may affect forage fish popu- 2009 (Figures 3 and 4), but no or lations and the predators that count low densities in the midwater. We The Big Picture on them as prey. evaluated the influence of physical, Forage fish are the critical middle of aquatic food webs throughout the world. Fig. 1 Changes in forage fish densities or distributions can affect forage fish recruitment, nesting/breeding success of birds, and/ or movements of fish or marine mammal predators that are important for commercial or traditional harvest. Understanding how forage fish distribute themselves is critical when evaluating potential impacts of climate change, and to fulfill the requirements of ecosystem-based approaches to fisheries management. Our baseline information can inform Bering Sea models that predict biological responses to climate change and improve methodologies for future Acoustic echograms showing differences in vertical distribution of age-0 pollock in 2008, 2009, and 2010. Bottom abundance estimate surveys. depth shown is ~80 m in 2008, ~110 m in 2009, and ~150 m in 2010.

ICHTHYOPLANKTON SURVEYS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. vulnerability to predators can all and our ability to obtain informa- Sandra Parker-Stetter, University of Washington (UW) influence survival of forage fish. tion on forage fish distributions John Horne, UW Distributions may result from a from existing surveys will also Ed Farley, National Oceanic and Atmospheric Administration, combination of selection of pre- change. A comprehensive analysis Alaska Fisheries Science Center ferred conditions and the influ- that included physical, biological, ence of water movement in the and climate factors was needed to The Bering Sea Project is a partnership between Bering Sea. If forage fish vertical understand what affects forage fish the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the or horizontal distributions change distributions. National Science Foundation’s Bering Ecosystem with environmental conditions, Study. www.nprb.org/beringseaproject then food availability for predators

Fig. 2 Fig. 3

Distribution of age-0 pollock in the surface (left) and midwater (right) Distribution of age-0 Pacific cod in the surface waters in 2009 (left) in 2009. Larger dots show higher densities. Bottom temperature (ºC) and 2010 (right). Larger dots show higher densities. High densities was an important predictor of midwater pollock density and is shown of age-0 Pacific cod were observed in 2010 in regions that had low on the midwater figure (red is warmest). Although there were few age-0 densities in 2009. pollock in the surface zone in 2009, there were regions of high densities in the midwater zone. Bottom temperature data courtesy of Bob Lauth (NOAA-AFSC).

Fig. 4 Sandra Parker-Stetter Forage fish, flatfish, and jellyfish from a surface trawl catch on the R/V Oscar Distribution of capelin in the surface waters in 2009 (left) and 2010 (right). Dyson in 2010. Larger dots show higher densities. Capelin were found in the same regions in both years, but densities were higher and more continuous along transects in 2010.

The Early Life of Walleye Pollock on the Eastern Bering Sea Shelf DISTRIBUTION SHIFTS IN WARM AND COLD YEARS

Fish eggs and larvae are very vul- Fig. 1 nerable within the first few months after being spawned. They are only a few millimeters long, and rely on ocean currents to take them from the spawning grounds to their nursery areas. Due to starvation and predation, less than one percent survive. To maximize their chances, adult fishes have evolved to spawn their eggs at the times and places that will lead to successful transport and higher survival rates of their larvae. In the eastern Bering Sea (Figure 1), sea ice may affect when and where adult Walleye Pollock (Gadus chalcogrammus) spawn their eggs, and current patterns affect where the eggs and larvae drift. This area recently experienced an exceptionally warm The dominant currents (blue lines) and Walleye Pollock spawning areas (green ovals) of the Eastern period (2001-2005) followed by a Bering Sea. The Alaska coastline is shown in black and the 50, 100, and 200 m isobaths in gray. prolonged cold period (2007-2012). ACC – Alaska Coastal Current; BSC – Bering Slope Current. During cold years, winter sea ice extended farther south and offshore, creating a large, cold pool of bottom The Big Picture water that adult pollock avoid. In the Spawning time, spawning location, and transport by currents affect the location and warm years this cold pool was much survival of fish eggs and larvae. The eastern Bering Sea recently experienced several warmer- smaller and there appeared to be than-average years followed by colder-than-average years. Observations of the spatial stronger flow to the east and onto the distribution of Walleye Pollock eggs and larvae indicated that larval distributions were shifted shelf. Research cruises observed that from the outer continental shelf towards the middle shelf in warm years. We used a computer pollock eggs and larvae were found model to simulate how pollock eggs and larvae are transported by currents, grow over time, further onshelf in warm years than in and move up and down in the water column. Simulations suggest that differences in adult spawning location between warm and cold years play a bigger role than differences in water cold years. We wanted to know how transport alone or differences in the time of spawning. continued on page 2

THE VARIABLE TRANSPORT OF POLLOCK EGGS AND LARVAE OVER THE BERING SHELF: A MARRIAGE OF PHYSICS AND BIOLOGY A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. differences between cold and warm and delays in spawning time are not warm years with less sea ice. We are years resulted in this pattern. sufficient to cause observed changes currently investigating the cause of in distributions. The distribution of these observations. We need a bet- How We Did It simulated eggs and larvae resembled ter understanding of these connec- We developed a model that observations when spawning areas tions because climate change and the simulated the transport, biology, and were expanded in warm years (Figure associated warming of the arctic and behavior of individual pollock eggs 2c). The simulation that produced subarctic not only affects the pollock and larvae. Simulated spawning areas distributions most comparable to population and therefore the entire were based on where adult pollock the observations was when spawning ecosystem of the eastern Bering Sea, in spawning condition have been areas were decreased offshore in cold but also the people who depend on found. Tens of thousands of eggs years (Figure 2d), but the differences pollock for their livelihood. were “released” in the model at seven between warm and cold years were not Colleen M. Petrik, University of Alaska Fairbanks (UAF) spawning times. We compared where as large as those observed. We con- Janet T. Duffy-Anderson, National Oceanic and Atmospheric different size classes of eggs and clude that the dissimilar distributions Administration (NOAA) Alaska Fisheries Science Center larvae were located in warm (1996, of eggs and larvae in warm and cold Franz Mueter, UAF 2002, 2003, 2005) and cold (1997, years most likely resulted from spawn- Katherine Hedstrom, UAF Enrique Curchitser, Dept. of Environmental Sciences and Institute 1999, 2000, 2006, 2008-2012) years ing area shifts in response to changes in of Marine and Coastal Sciences, Rutgers University by calculating the center of the distri- the presence and extent of sea ice. bution of each size class. We consid- The Bering Sea Project is a partnership between ered four different scenarios. In each, Why We Did It the North Pacific Research Board’s Bering Sea the ocean currents were specific to Related studies have shown dif- Integrated Ecosystem Research Program and the the simulated year and were based on ferences in prey availability between National Science Foundation’s Bering Ecosystem observed climate conditions, while warm and cold years, and fewer pol- Study. www.nprb.org/beringseaproject spawning areas and times differed lock surviving to adulthood in recent among scenarios: (a) Spawning time and location were the same for warm Fig. 2 and cold years. (b) Spawning locations were the same, but spawning time was 40 days later in cold years, simulating the possibility that adult fish waited for sea ice retreat before spawning. (c) Spawning time was the same, but the spawning areas were increased in size in warm years, simulating the expansion of spawning adults into areas without sea ice. (d) Spawning time was the same, but the spawning areas were reduced in size in cold years, simulating avoidance of sea ice-covered areas by adult fish. When spawning time and location were held constant (Figure 2a), and when spawning time was 40 days later in cold years (Figure 2b), the centers of distribution of pollock eggs and larvae Modeled centers of gravity of late stage (10-40 mm Standard Length) larvae in cold (blue) and warm (red) did not differ much between warm years for all 4 scenarios: (a) spawning time and location were the same for warm and cold years (the red and cold years, suggesting that climate- dot is on top of the blue dot); (b) spawning locations were the same, but spawning time was 40 days later related differences in ocean circulation in cold years; (c) spawning time was the same, but the spawning areas were expanded in warm years; (d) spawning time was the same, but the spawning areas were contracted in size in cold years.

Zooplankton Populations in the Eastern Bering Sea LINKING CLIMATE, ZOOPLANKTON AND FISHERIES

Zooplankton are large and small System) at night (Fig. 2) to ensure Fig. 1 animals, mostly invertebrates, that representative collections. The sam- drift in the water. Information on ples then were brought to the lab zooplankton abundance is being and preserved critters were identi- used by the Bering Sea Project to fied and counted. These samples assess the health of the Bering Sea will be stored at the University of ecosystem and to aid in understand- Alaska for at least 20 years and ing the potential effects of global cli- made available upon request to mate change on Bering Sea fisheries. future researchers. The data on Since the Bering Sea sustains large zooplankton composition and the commercial fisheries and subsistence surrounding environment, which resources for native communities, were simultaneously collected understanding the potential effects with automated sensors during the of climate change on zooplankton net tows, were uploaded into the NOAA Russ Hopcroft, Pelagic predatory amphipod Themisto libellula and the fish populations that feed Bering Sea Project interdisciplinary flourish in cold Arctic and Bering Sea waters. on them will help policy makers database (beringsea.eol.ucar.edu) plan for and mitigate climate-related for public availability. impacts on the fishing and indig- continued on page 2 enous communities along the Bering Sea coast. The Big Picture How We Did It Ecosystem studies over the last decade revealed substantial declines in populations of We collected 675 zooplankton large zooplankton during the warm period of 2002 – 2006. These declines were accompanied samples from the eastern Bering by declines in the survival of juvenile pollock, a major commercial fish on the Bering Sea Sea, covering all shelf domains shelf. Since large zooplankton are an important food for young pollock, the declines in large and extending from the Alaska zooplankton are a potential reason for observed declines in survival of young pollock. In addi- Peninsula in the south to the St tion, large zooplankton are an important food for salmon, herring, capelin, and other large fish Lawrence Island in the north. species. In the absence of large zooplankton, other large fish were consuming juvenile pollock, Because many zooplankton taxa thus lowering pollock survival and stock size. As fish stocks decline, the supply of fish to the spend daytime in the deep and fishery also declines, resulting in lower incomes and employment in fishing communities. As ascend to the surface during the assessed by the Bering Sea Project, colder temperatures in 2007 – 2010 were accompanied by night, we fished a MOCNESS a recovery of large zooplankton populations. Increases in abundance of large zooplankton dur- (Multiple Opening/Closing Net ing the recent cold period are further evidence that declines in zooplankton during the warm and Environmental Sensing period were temperature-related.

ASSESMENT OF MESOZOOPLANKTON POPULATION AND BIOMASS IN THE EASTERN BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 2 Chris Lindner, WHOI Chris Lindner, Nighttime deployment of the Multiple Opening/Closing Net and Environmental Sensing System (MOCNESS) off the stern of USCG Cutter Healy.

In addition to increases in Why We Did It in a timely manner, minimizing the populations of large zooplankton The species composition and potential impacts of unexpected during the cold period of 2007 abundance of plant and animal changes in fish and wildlife popula- – 2010, large Arctic zooplankton populations in ecosystems are tions on the coastal communities species, such as pelagic amphipod continuously changing in response that depend on these resources for Themisto libellula (Fig. 1), occurred to climate. Since climate warm- subsistence and commercial harvests. in the samples. These arctic spe- ing is predicted to occur rapidly in cies had not been observed in Alexei Pinchuk, School of Fisheries and Ocean Sciences, arctic and subarctic environments, University of Alaska Fairbanks the southern Bering Sea since the these changes in species composi- Ken Coyle, Institute of Marine Science, University of Alaska 1970s. Arctic species can be an tion and abundance are likely to Fairbanks important food source for sea- accelerate and increase in ampli- birds and commercial fish, so their tude. Nevertheless, ecosystems are The Bering Sea Project is a partnership between reappearance on the Bering Sea extremely complex and can change the North Pacific Research Board’s Bering Sea shelf is an indication that climate in unpredictable ways. Therefore, Integrated Ecosystem Research Program and the change can impact the Bering Sea sound resource management in a National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject ecosystem by changing the species changing world requires continuous composition of the constituent assessment of the plant and animal populations in addition to chang- populations to allow resource man- ing population size. agers to modify management policies

ASSESMENT OF MESOZOOPLANKTON POPULATION AND BIOMASS IN THE EASTERN BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. UNDERSTANDING ECOSYSTEM PROCESSES IN THE BERING SEA 2007–2013

Seabird Diets and Reproductive Success in the Pribilofs INSIGHTS FROM A 35-YEAR KITTIWAKE DATASET 2008-2010 were three cold a relatively high year at St. Paul years on the Bering Sea shelf, in 2009, when there was a patch characterized by cold ocean of age-1 pollock to the north- temperatures and high ice extent. west of the island, the propor- The reproductive success of black- tion of pollock in kittiwake diets legged kittiwake (Rissa tridactyla) has decreased since 1975, while was well below average in 2008- that of sand lance has increased.

2009 and slightly above average Long-term, kittiwake diet was cor- Brie Drummond in 2010. A look at our long-term related with some broad-scale cli- Stephanie Walden holds a kittiwake captured for datasets on diet and reproduc- mate variables (Arctic Oscillation diet and survival studies. (To prevent disturbing tive success helped us put these and regional summer sea surface the same bird multiple times, researchers apply a years in perspective. Except for temperature) but not with local dye that wears off in about a week.) continued on page 2 Fig. 1 The Big Picture The seabird colony-based studies of BSIERP relied heavily on intensive diet, foraging trip, and reproductive success data collected during 2008-2010, the three field seasons of the project. However, understanding relationships among climate variables, seabird diet, and reproductive output requires many more years of study. Otherwise, how would we St George black-legged kittiwake reproductive success, 1975-2010. know what is a good year or a normal year? The seabird cliffs in the Pribilofs are part of the Alaska Maritime National Wildlife Refuge, which has an ongoing annual seabird monitoring program at eight sites around the Alaskan coast. This long-term dataset helps us place in context the detailed diet, foraging behavior, body condition, reproductive success and survival data collected as part of the Bering Sea Project. St Paul black-legged kittiwake reproductive success, 1975-2010.

SEABIRD COLONY-BASED STUDIES A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. physical variables. When we sepa- reproductive success. An increase cliffs of both Pribilof Islands, rated reproductive success into its of prey from deeper waters beyond monitoring individually-numbered sequential components, we found the shelf break (mediated by nests to determine success or failure. that success in earlier parts of the travel distance required to access For diet studies, adults are captured nesting cycle and the previous prey) and small invertebrates in bringing food back to the nest sites year were more important predic- diets negatively affected fledging after the chicks have hatched. tors of overall productivity than success, which may indicate low any climate variables. Timing availability of high quality prey Why We Did It was also an important predictor near the colonies. As upper level predators in the of laying success for kittiwakes. marine ecosystem, seabirds reflect These relationships suggest a How We Did It fluctuations in the marine envi- cascade effect, in which adult Most summers since 1975, field ronment that influence their prey condition carrying over from the crews have spent three months shiv- supply. Studies of seabird diets and previous year plays a large role in ering on the fog-shrouded seabird reproductive success thus provide insight into the physical and biological mechanisms that potentially drive population changes in both predators and their prey. The eastern Bering Sea shelf, among the most productive marine ecosystems in the world, has undergone significant restructuring in recent decades that is likely to continue in light of anticipated climatic change.

Heather Renner, U.S. Fish & Wildlife Service (USFWS) Alaska Maritime National Wildlife Refuge (NWR) Brie Drummond, USFWS Alaska Maritime NWR John Warzybok, USFWS Alaska Maritime NWR Vernon Byrd, USFWS Alaska Maritime NWR (retired)

Heather Renner Heather Kittiwake nesting habitat at the Pribilof Islands.

Do Subsistence Harvests Reflect Ocean Ecology? WHAT REGIONAL PATTERNS OF HUNTING AND FISHING REVEAL

When Alaska’s coastal resi- Fig. 1 dents hunt and fish they, in effect, sample their local environment. We wanted to know what the patterns of subsistence harvests could tell us about ecological patterns in the environment, and also whether those patterns revealed any cultural preferences for certain types of foods. We looked at 35 communities along the coasts of the Bering, Chukchi, and Beaufort Seas. We found that the patterns of local harvests appear to follow biological, oceanographic, and geographic patterns, with the precise patterns depending on the type of analy- Villages (dots) shown in the same color share similar subsistence harvest practices. The variance between sis (Figure 1). These results sug- (a) and (b) reflects two different cluster analysis approaches. Both analyses show east-west divisions, which appear to follow patterns of ocean currents and species migrations. The left-hand figure (a) also separates gest that subsistence harvests are North Slope communities, likely reflecting high seal and whale harvests in the area. In the right-hand figure samples of the local environment, (b), the green dot on the North Slope is Nuiqsut, which is located on the Colville River and has similarities with indicating patterns of regional communities on the Alaska Peninsula that also harvest a mix of fish and marine mammals. ecology, physical settings, or other influences on what people harvest. Further studies of harvest levels could reveal patterns over time, The Big Picture reflecting environmental change. Our analysis of the relationship between ecological patterns and subsistence patterns is We also divided the villages into one of many analyses that emerged from all the interactions among researchers throughout the Bering Sea Project. The cluster analysis of subsistence harvest data helps to connect the six regions and compared the regions characteristics of the ecosystem to human interaction with the Bering Sea (and neighboring to see which regions were most seas). It is exciting to see that different approaches to studying the ecosystem produce similar closely related. Not surprisingly, pictures of what is happening. This increases our confidence that we are correctly identifying geography dominated the result patterns, and also reinforces the idea that ecosystem patterns matter to people who live on (Figure 2). Interestingly, the north- the islands and coast of the Bering Sea. ern Bering Sea aligned more closely continued on page 2

SUBSISTENCE HARVEST, USERS AND LTK ECOSYSTEM PERSPECTIVE A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. with the Chukchi and Beaufort with more recent studies typically other factors might also affect har- than with the central and southern identifying harvests by species (e.g., vest patterns. In addition to satisfy- Bering Sea. This differs from an sockeye salmon, or king eider), ing our curiosity, we hoped that the analysis of marine ecology done for rather than by larger group (e.g., results might shed light on whether the same region, but is not surprising salmon, or eiders). For the purpose subsistence harvest characteristics given the migration routes of bow- of establishing a consistent body of could be used as indicators of the head whales and walrus, which are data to work with, we had to sacri- condition of the ecosystem. Further popular subsistence resources in the fice some level of detail and lump studies looking at patterns over time northern Bering Sea as well as along some species together. We then would be useful, but at the moment the Chukchi and Beaufort coasts. conducted the cluster analyses at we do not have enough studies in the village and regional levels to see the same villages at different times How We Did It what patterns emerged. to do that analysis. We started with subsistence Why We Did It Martin Renner, Tern Again Consulting harvest survey results compiled Henry Huntington by the Alaska Department of Fish An earlier analysis of marine and Game. These covered 35 com- ecology data spurred us to wonder munities in the region. Because how subsistence harvest patterns The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea some communities had more than compared with the underlying Integrated Ecosystem Research Program and the one year of data, we had a total of ecological patterns across the same National Science Foundation’s Bering Ecosystem 53 harvest surveys from 1964 to region. While we recognized that Study. www.nprb.org/beringseaproject 2009. The degree of detail about the hunters and fishers have to rely on harvest varied from study to study, what is available, we wondered if

Fig. 2

Communities were grouped into six oceanographic regions, and the results were analyzed to see how similar the subsistence harvests were to one another. The farther left the bar connecting two regions is on the above diagram, the more

closely their harvests resemble one another. The Peninsula, AI (Alaska Peninsula Huntington Henry P. and Aleutian Islands) group is most similar to the Southern Bering Sea group. Subsistence harvests, such as the variety of foods seen here at a These two are the most similar of any pair of groups. These two are also fairly community feast in Wainwright, reflect the ecology of the local similar to the Central Bering Sea group. These three, on top, have relatively little in area. In this picture, maktak (bowhead whale skin and blubber) fills common with the three on the bottom. The Beaufort Sea group and the Chukchi the bowls in the foreground, with a variety of fishes, marine mam- Sea group are closely related, and have some features in common with the North- mal meats, soups, and other delicacies behind. ern Bering Sea group, although this link is not as close.

Documenting Nelson Island Natural and Cultural History CAPTURING A WEALTH OF ANCIENT KNOWLEDGE

Between 2006 and 2010, the Fig. 1 Calista Elders Council (CEC), the primary research organization for Southwest Alaska, worked with elders and community members from five Nelson Island communities on the Bering Sea coast to document the natural and cultural history of their homeland.

How We Did It CEC staff traveled with elders out

on the land to document historic Feinup-Riordan Ann sites and landscape features on and Simeon Agnus points out a land feature near Arayakcaaq at the mouth of the Qalvinraaq River, July around Nelson Island (Figure 1). 2007. Michael John sits to his right and Theresa Abraham to his left. CEC staff also hosted a number of topic-specific gatherings, two- and three-day meetings with elder experts The Big Picture devoted to a single topic, as an effec- Coastal communities throughout Alaska, as elsewhere, are undergoing profound environ- tive means of both documenting mental, socioeconomic, and cultural changes related to their reliance on marine ecosystems and, increasingly, a global economy. Social scientists, as well as community members, traditions and addressing contempo- increasingly seek to understand community vulnerability and sustainability. To do so, it is not rary scientific concerns. Unlike inter- sufficient to say that changes are taking place. We need to understand how community mem- views, during which elders answer bers interpret these changes--not just what is occurring but why people believe it to be so. questions posed by those who often CEC’s collaborative approach, grounded in community initiatives and local elder gatherings, is a do not already hold the knowledge powerful tool that can simultaneously help natural and social scientists understand the unique they seek, gatherings (like academic cultural perspectives that underlie the actions and reactions of coastal residents, and give voice symposia) encourage elders to speak to community understandings of the world in which they live. among their peers at the highest level (Figure 2). Work with Nelson Islanders Riordan and Rearden, 2012) is a variously as weather, world, uni- resulted in two major publications. 450-page ethnography document- verse, and awareness. The book’s Ellavut/Our Yup’ik World and ing the qanruyutet (oral instruc- ten chapters reflect gathering topics, Weather: Continuity and Change tions) that continue to guide Yup’ik including weather, land, lakes and on the Bering Sea Coast ella (Fienup- interactions with —translated continued on page 2

NELSON ISLAND NATURAL AND CULTURAL KNOWLEDGE PROJECT A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 2 Arctic) to link separate mapping Fig. 3 efforts in Bering Sea coastal communities into a comprehensive map web service covering 200 miles of coastline and over 6,000 place names. Like the NCAR site, the new site— http://eloka-arctic.org/communities/ yupik/—has the capacity to display a wide variety of information (audio, video, text, and photographic), and will serve as an invaluable resource Elders and youth discuss place names during for the region, both educational and a CEC gathering in Chefornak community hall, March 2007. capacity building, for years to come. Why We Did It rivers, ocean, snow, ice, survival, and Nelson Islanders express an environmental change (Figure 3). urgent need to document their Our project also produced the unique natural and cultural history. bilingual book Qaluyaarmiuni Many recognize that such docu- Ellavut/Our Yup’ik World and Weather: Nunamtenek Qanemciput/ Our mentation must happen in the near Continuity and Change on the Bering Sea Nelson Island Stories: Meanings future or not at all. Although there Coast by Ann Fienup-Riordan and Alice Rearden. of Place on the Bering Sea Coast, will always be elders, the present winner of a 2012 American Book generation of elder experts are the Award. Elders actively support the last to have received a traditional Fig. 4 documentation and sharing of tra- education in the qasgi (communal ditional knowledge, which all view men’s house) before the advent as possessing continued value in the of organized religion and formal world today (Figure 4). education. Elders were the primary Community members have teachers in the past. Venues to share embraced the idea of using the their knowledge have drastically web to share information gathered declined, and contemporary elders during the Bering Sea Project. In actively seek arenas to share their collaboration with National Center knowledge. Our project provided for Atmospheric Research (NCAR), a unique opportunity for elders, Earth Observing Laboratory, CEC community members, scientists, and has developed a place-based web- local organizations to work together site including the location of over toward this common goal, enriching 400 historic sites and geographic lives locally while at the same time features, as well as oral accounts sharing knowledge globally. relating directly to over 100 sites. Ann Fienup-Riordan, Calista Elders Council (CEC) Community members voted unani- Mark John, CEC mously for open access to their Alice Rearden, CEC site, which can be viewed at http:// Qaluyaarmiuni Nunamtenek Qanemciput/ Our Nelson Island Stories: Meanings of Place mapserver.eol.ucar.edu/best. on the Bering Sea Coast edited by Ann Fienup- Expanding on our Nelson Island The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Riordan, with translations by Alice Rearden. project, CEC is presently working Integrated Ecosystem Research Program and the with ELOKA (Exchange of Local National Science Foundation’s Bering Ecosystem Observations and Knowledge in the Study. www.nprb.org/beringseaproject

Synergies Between Traditional and Western Environmental Knowledge LINKING LOCAL AND GLOBAL

Between 2006 and 2010, the How We Did It Calista Elders Council (CEC), the We interviewed Nelson Islanders The Big Picture primary research organization for to learn their unique, nearshore The emerging question that concerns Southwest Alaska, worked with perspective on the Bering Sea. While both Yup’ik and non-Yup’ik ocean observ- elders and community members oceanographers attempt a compre- ers is: How can we link local observations from five Nelson Island commu- hensive understanding of the ocean, with large-scale environmental issues? nities on the Bering Sea coast to Yup’ik hunters are most concerned Our oceans are now being monitored. document the natural and cultural with surface features of the water and If we add to this a greater understand- history of their homeland. ice cover because of their impact on ing of the seas immediately offshore, The last 60 earsy has seen dra- hunting success and safety of travel. the contrasting scale of global versus matic change in the way Nelson Yet, coastal Yup’ik residents also see local can be bridged. How do we make this potential integration a reality? We Islanders inhabit their land. the ocean as an integral part of ella, can make a start by listening to Yup’ik Perhaps most significant is the a word they translate as weather, community members, engaging those concentration of people into five world, universe, or awareness, whose understanding of the ocean is permanent villages and the aban- depending on context. not only useful but represents a unique donment of hundreds of small In the many warnings elders worldview. They have long accepted per- camps and settlements that were give of a dangerous and unpredict- sonal responsibility for changes in their still vibrant through the 1940s. able ocean, they also identify key homeland. They lead by example. These five villages—ranging in research problems. One example population sizes from 250 to 600— is connecting the response of the are small by modern standards, but nearshore ice regime to ocean swells Fig. 1 huge compared to the tiny settle- and tides. Yup’ik people have many ments of the past. words describing the appearance As people gather closer together and response of ice to currents and on and around Nelson Island, the winds. One opportunity for western island’s resources, although still science/traditional knowledge syn- abundant, are more distant. People ergy could be to start with commu- still harvest from the fishing sites nication that enables sharing such their parents used, but at a cost insights and knowledge, followed many find difficult to afford. Now by focused research partnerships to men often travel miles, either by study the linkage of wind, wave, gasoline-hungry snowmobile or and ice dynamics. Ice mixed with mud on the north shore of Toksook skiff, to set their nets and traps. continued on page 2 Bay, May 2008.

NELSON ISLAND NATURAL AND CULTURAL KNOWLEDGE PROJECT A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 2 shallow, muddy coastal environ- surges, and unusual ice and weather ment (Figure 1). Sea ice scientist conditions, as well as their views Hajo Eicken notes that while coastal on ongoing changes in the Bering erosion is often attributed to a lack Sea ecosystem, will be invaluable in of sea ice, in fact sea ice is the most preparing them for the future. effective mover of sediments in Our work with Yup’ik com- waters with seasonal ice cover. How munity members has been a major the ice interacts with the coast is collaborative effort during which not well understood and cannot be we made a serious attempt to co- captured by satellites. Local observ- produce the knowledge we share. ers recording locations of dirty ice Meetings went beyond consulta- can help with modeling sediment tion and cooperation, with mutual transport by ice. sharing of ideas and understand- The rise of sea level and related ings. These deep collaborations effects of increased fall storm surges offer powerful alternatives to more are of particular concern, both to conventional research approaches. ocean scientists and coastal residents. Ann Fienup-Riordan, Calista Elders Council (CEC) Elders’ long-term observations of Mark John, CEC these changes may be particularly Alice Rearden, CEC valuable. The village of Newtok, 10 Visiting officials, including Senator Mark Begich, inspect the eroding shoreline at Newtok, spring feet above sea level, was established in 1950 on the low-lying tundra The Bering Sea Project is a partnership between 2010. the North Pacific Research Board’s Bering Sea north of Nelson Island. Men chose Integrated Ecosystem Research Program and the Another opportunity for col- the site because it was accessible National Science Foundation’s Bering Ecosystem laboration is provided by meteorolo- to barges bringing in lumber for Study. www.nprb.org/beringseaproject gist Uma Bhatt who used satellite the new school. Despite Newtok’s images to demonstrate the links marshy location, it doubled in size to between diminishing Arctic sea ice 350 today. At the same time Newtok Fig. 3 and changes in the Arctic terrestrial was growing, the land was sink- ecosystems. She and her colleagues ing and eroding at an alarming rate found that areas in the High Arctic (Figure 2). A move to relocate the have experienced the largest changes, village to a bedrock site on Nelson with some exceptions over land Island is already underway. regions along the eastern Bering. In discussions with Bhatt, elders pointed Why We Did It out both a decline in tundra berry Yup’ik coastal residents of all ages production and the timing of the are concerned by the unprecedented harvest in recent years, which they changes in climate and ecology they associate with a decrease in fall rain are witnessing along the Bering and snow cover. Winds during the Sea coast, including changes in Mark John and Alice Rearden looking over archi- growing season were another factor. the ranges and availability of fish, val photos with Maryann Andrews of Emmonak These observations point to the need mammals, and birds; coastal ero- and Barbara Joe of Alakanuk during an elders’ to look at changes in wind and pre- sion; later fall freeze-up and earlier gathering in Anchorage, April 2012. cipitation, as well as sea ice cover, to spring breakup; unusual weather explain changes in coastal ecosystems. patterns; and increased storminess. Elders also shared valuable Community members feel strongly observations about sediment-laden that elders’ perspectives on past ice--a common characteristic of the periods of resource scarcity, storm

The Role of Edgy Phytoplankton in the Bering Sea Ice Environment THEY DETERMINE WHO GETS WHAT AND WHERE

This project focused on the • How does the formation and • What is the relative importance growth of phytoplankton, the movement of ice influence the of dissolved nitrogen transported small photosynthetic organisms fertility of the region? onto the shelf from deeper waters, that bloom in the spring in high- • Are the patterns consistent from relative to organic nitrogen that is latitude seas throughout the world. year-to-year (or at least predict- recycled on the shelf? We wondered how sea ice impacts able from sea ice patterns)? continued on page 2 the pattern and intensity of pro- Fig. 1 duction in the eastern Bering Sea. One of the primary mechanisms we examined was the role of ice in determining the availability of nutrients, particularly nitrogen, required for the growth of all organisms. Some of the more detailed questions we addressed include:

The Big Picture Although we know that the extent of sea ice has varied in the eastern Bering Sea, the impact of these variations on the fish, birds, and marine mammals is not well understood. Some of this gap is due to the limited oceanographic sampling of ice that has taken place, particularly in relation to plankton growth and its relationship to nutrient levels. A detailed, mechanistic understanding of how physi- Nitrogen (N) productivity, surface nitrate concentrations, and ice extent in the eastern Bering Sea in A. 2007; B. 2008; cal environmental changes propagate C. 2009; and D. 2010. In each panel, the color map represents surface nitrate concentrations (nitrate is the preferred from the plankton to the populations form of nitrogen for phytoplankton growth). Note that the data in 2010 are from a smaller region of the shelf than in the other years. The vertical bars represent nitrogen productivity (a measure of the rate of phytoplankton growth). of upper food web levels is needed to For each N-productivity bar, purple represents the amount of nitrate productivity and gray represents the amount better manage stocks and predict future of ammonium productivity (the two different forms of nitrogen; ammonium is less preferred). The solid line is the ecological conditions. 200 m depth. The dashed lines represent the ice extent in March, April and May in each year, and together with the nitrogen productivity rates show the elevated productivity associated with the ice edge on the western shelf.

NITROGEN SUPPLY FOR NEW PRODUCTION AND ITS RELATION TO CLIMATIC CONDITIONS ON THE EASTERN BERING SEA SHELF A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. How We Did It previously produced organic mat- a combination of oceanographic We found that, when it comes ter. This pattern showed up clearly factors such as wind, ocean cur- to phytoplankton productivity, not in the phytoplankton incubations rents, and ice that control when and all ice edges are created equal— we did, as well as in isotopic mea- where light and nutrients provide some were associated with dense surements that were made on the suitable conditions for growth. phytoplankton blooms, but oth- nutrients themselves. Phytoplankton growth impacts the ers were not. An exception was upper food web levels in a bottom the region of the outer shelf, from Why We Did It up fashion, and as the spatial pat- just north of the Pribilof Islands to The fish, birds, and marine tern of phytoplankton productivity beyond Zhemchug Canyon, where mammals that were the focus of changes from year-to-year, the fish, we found heavy growths of phyto- the Bering Sea Project depend on birds, and marine mammals must plankton in each of the four years the food web, of which they are a deal with resulting variations in we sampled it (Figure 1). The ice part, to supply them with enough their food supply. appears to consistently create good resources at appropriate times in Raymond Sambrotto, Lamont-Doherty Earth Observatory of growth conditions here. Also, there their lives. Food for all organisms Columbia University was a cross-shelf pattern in the use can be traced to the initial forma- Daniel Sigman, Princeton University of nitrogen by phytoplankton in tion of organic material by photo- the spring. The outer shelf ice edge synthetic organisms; in the sea, this The Bering Sea Project is a partnership between blooms were fueled mainly by deep- mainly comes from phytoplankton. the North Pacific Research Board’s Bering Sea water nitrogen, while phytoplankton A challenge for marine animals, Integrated Ecosystem Research Program and the growth in shallower, inshore waters however, is the extreme variabil- National Science Foundation’s Bering Ecosystem www.nprb.org/beringseaproject had a much greater dependence ity of phytoplankton production. Study. on nitrogen that was recycled from Phytoplankton are dependent on R. Sambrotto Photograph of the ice edge in the Bering Sea showing the dense growth of algae that turns the bottom of the ice brown. The ice releases these algal cells as it melts, and these contribute to dense phytoplankton blooms at the ice edge.

Protists - Tiny Predators of Phytoplankton AN AMAZING ARRAY OF ONE-CELLED ORGANISMS FEED IN THE BERING SEA While copepods are small, about studied the importance of microzoo- microzooplankton. In these samples, the size of grains of rice, microzooplankton protists as consumers of we took photographs of protists that plankton protists are even tinier, less algae, predominately diatoms, in the were caught in the act of feeding on than 200 microns in size, smaller Bering Sea. Experiments were done diatoms. than poppy seeds. How are such in on-deck incubators to measure The results of the experimental miniscule predators able to feed on growth of algae in seawater with and incubations showed that the micro- diatoms, single-celled algae that are without the presence of microzoo- zooplankton protists were important often as big, or even bigger, in size plankton predators. The difference consumers of diatoms. We found than the protist themselves? It turns in algal growth, measured by change a positive relationship between the out that these protists have many, in amount of chlorophyll during the biomass of these predatory pro- often surprising, ways to prey on incubations, showed how much algal tists and phytoplankton stocks as diatoms. production was eaten by the micro- measured by the concentration of zooplankton. Using a microscope, chlorophyll in the water (Figure 1). How We Did It we inspected water samples that we At some sites, we measured a high During spring (March-June) had preserved at sea to determine amount of protist grazing on intense Bering Sea cruises in sea ice, we the biomass and types of predatory continued on page 2 Fig. 1 The Big Picture When ocean ecologists say, “all fish is diatom,” they mean that the annual blooms of these large, lipid-filled algae support major marine fisheries. Diatoms, which grow both in sea ice and in the water, are known to form the base of food webs in the Bering Sea. Yet, exactly how the production of diatoms moves through the food chain to fish, seabirds, seals, and whales is still debated. The standard concept is that diatoms are eaten by crustacean zooplankton such as copepods and krill, which are then consumed by higher predators. However, this view of a straight-line food chain is giving way to evidence that much of the diatom production is instead Log-log relationship between biomass of microzooplankton (μg C/liter) and concentration of phyto- consumed by unicellular predators, protists plankton (μg chl-a/liter) in the Bering Sea during spring (March-June). Chlorophyll ranged from 0.2 to in the microzooplankton. 38 μg/liter, and microzooplankton biomass from 2 to 72 μg C/liter.

MESOZOOPLANKTON-MICROBIAL FOOD WEB INTERACTIONS IN A CLIMATICALLY-CHANGING SEA ICE ENVIRONMENT A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. diatom blooms. Our most surpris- chain. The protoplasm surrounds in the Bering Sea; although parasitic ing finding was the varied ways that the diatoms, and then enzymes are flagellates have been reported to protists fed on diatoms. released to digest the algae and slurp crash a diatom bloom in a European The most common types of the food back into the dinoflagellate coastal system. protist predators of diatoms were cell (Figure 2-D). large-sized dinoflagellates. Common Some types of protists feeding on Why We Did It species of marine dinoflagellates use diatoms were unexpected. Shelled We hope that future studies will only organic materials as a source amoebae that sucked out the pro- discover the true significance of of food, and make their living by toplasm of centric diatoms (Figures these varied protists as consumers of feeding on other cells. Abundant 3-A, 3-B) were among the most diatoms, and whether their feed- Gyrodinium dinoflagellates in the curious predators. In one shipboard ing impact in Bering Sea food webs Bering Sea were able to engulf large experiment, these amoebae dra- might dramatically increase as a diatom cells and chains (Figures matically increased in abundance, result of global warming. 2-A, 2-B, 2-C). In some cases, the which showed that they were able to Evelyn Sherr, College of Oceanic and Atmospheric Sciences dinoflagellate cell was so distended rapidly grow on diatom food. Even (CEOAS), Oregon State University (OSU) to accommodate a long diatom smaller protists prey on diatoms Barry Sherr, CEOAS, OSU chain that it appeared about to pop by attaching to the silica shell and Carin Ashjian, Department of Biology, Woods Hole Oceanographic Institution (Figure 2-C). Other types of preda- injecting enzymes to digest the cell Robert Campbell, Graduate School of Oceanography, tory dinoflagellates are encased in contents. Parasitic flagellates have University of Rhode Island rigid armor plates, called a theca. been previously observed preying These thecate dinoflagellates cannot on centric diatoms during summer The Bering Sea Project is a partnership between change their shape to surround a in the Bering Sea. Similar flagellates the North Pacific Research Board’s Bering Sea diatom chain as do their Gyrodinium infested chains of pennate diatoms Integrated Ecosystem Research Project and the cousins. Instead, the dinoflagellates during our spring study (Figure National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject extrude an amoeba-like blob of 3-C). We don’t yet know how protoplasm that attaches to a diatom important these diatom parasites are

Fig. 2 Fig. 3 A, B, C: Evelyn Sherr; D: Celia Ross Sherr; D: Celia C: Evelyn A, B, Sherr Evelyn Dinoflagellates are well known as predators of large diatom cells and chains. Other types of protists found feeding on diatoms included shelled amoebae and A. Gyrodinium sp. dinoflagellate without ingested prey. B. Gyrodinium dinofla- parasitic flagellates. A. Two amoebae attached to one diatom cell. B. Single gellate cell distended with an engulfed single centric diatom cell. C. Gyrodinium amoeba feeding on a diatom cell. C. Fragillariopsis diatom chain infested with dinoflagellate cell grossly distended with an engulfed pennate diatom chain parasitic flagellates. about 40 cells in length. D. Thecate dinoflagellate feeding on a diatom chain by attachment of an extruded blob of protoplasm containing digestive enzymes. The brown color of the cells is from the iodine fixative used in preservation.

Organic Matter Mineralization in Bering Sea Sediments Fig. 1 WHAT THE SEAFLOOR TELLS US ABOUT CLIMATE-INDUCED CHANGES The seafloor is the Bering Sea’s The most energy-efficient mecha- recycling center. When food from nism of organic matter mineraliza- surface waters hits the bottom, it is tion is aerobic respiration. But, in consumed by scavengers and bacteria the absence of dissolved oxygen, that inhabit the seafloor. The bio- microbes use anaerobic respiration chemical processes (termed mineral- with different oxidants to break Deploying a multi-corer. ization) that break down this organic down organic carbon. The sequence carbon are much more complicated of oxidants that we would expect to Fig. 2 and more interesting than one might observe in Bering Sea sediments in imagine. Microbial communities order of decreasing efficiency is oxy- use a variety of biochemical pro- gen, nitrate, manganese oxide, iron cesses to oxidize the food that hits oxide, and sulfate. This sequence the bottom. The oxidative pathways produces vertical gradients in these taken by the organic matter likely chemicals within the sediment vary with the rate of food supply to column. The various organic matter the seafloor. Because much of the mineralization mechanisms follow organic matter export to the Bering different chemical reactions and Sea floor occurs during seasonal sea produce different by-products. ice melt, a warming climate would We hypothesized that organic- be expected to reduce the quan- carbon mineralization pathways tity of organic matter reaching the would vary with food supply to the sediment, and this change could be seafloor, and this would produce observed by studying organic matter observable regional variation in Red dots indicate locations of core samples taken mineralization processes in Bering the organic matter mineralization from across the eastern Bering Sea over a four- year period. Sea sediments. continued on page 2

The Big Picture Organic matter production fuels the highly productive Bering Sea food web. The amount of organic matter exported to the seafloor varies spatially across the Bering Shelf, and may reflect longer-term changes in productivity related to the seasonal melting of sea ice and the development of nutrient limitation on the shelf. But what happens to this material after it hits the bottom? We quantified the fate of organic matter in Bering Sea sediments and discovered that the processes vary regionally, reflecting the export of organic matter from the water column. Because sedimentary processes tend to filter out short-term fluctuations, the variation in organic-matter mineralization in the sediments may be a useful indicator of climate-induced changes in ecosystem productivity that fuels the important Bering Sea fishery.

DENITRIFICATION AND GLOBAL CHANGE IN BERING SEA SHELF SEDIMENTS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. pathways. Specifically, we expected and caps the tubes, and transports allowed us to determine how sedi- the proportion of organic matter the intact section of seafloor back mentary organic matter was remin- mineralization due to aerobic respira- to the ship for analysis and experi- eralized in different regions of the tion to increase from north to south mentation. We collected sediment Bering Sea, and to test our hypoth- on the Bering Shelf, and also increase cores from approximately 125 loca- esis regarding its regional variation. from onshore to offshore. This pat- tions on the Bering Shelf, slope and Indeed, our results confirmed that tern parallels independent estimates rise over four years (Figure 2). We food supplied from the overly- of organic matter export, which is incubated cores on the ship at near ing water takes different oxidative largest in the northern shelf and in situ temperatures and directly pathways in different regions of the drops to the south and offshore. We measured the rate of oxygen con- Bering Sea in a manner that is con- set out to test this hypothesis in order sumption and nitrogen gas produc- sistent with its variation in supply to better understand how organic tion (to quantify denitrification) in from the water column (Figure 3). matter is recycled in Bering Sea sedi- the sediments. We measured the rate ments and how it relates to the sup- of sediment mixing (bioturbation) Why We Did It ply of food from overlying water. using the chemical tracer 234Th and Sedimentary processes tend used that rate, along with profiles to reflect average conditions and How We Did It of manganese (Mn) and iron (Fe) filter out short-term fluctuations. We deployed a hydraulically- oxides, to estimate the rates of Mn Processes in overlying water are damped “multi-corer” (Figure 1) and Fe reduction. We also incubated subject to considerable variation, which drops to the bottom, slowly sediment amended with radioactive even at small spatial and temporal plunges eight sampling tubes into sulfate to determine the rates of sul- scales. Thus, variation in sedi- the sediment, carefully withdraws fate reduction. These measurements mentary processes such as organic matter mineralization may indicate Fig. 3 longer-term changes in conditions in the Bering Sea. Our results were consistent with our hypothesis that the proportion of organic matter mineralization due to aerobic respiration would increase from north to south on the Bering Shelf, and also increase from onshore to offshore. This pattern parallels independent estimates of organic matter export. As this ecosystem changes on decadal and longer time scales, these changes may be reflected in organic matter mineralization pathways revealed in the sediment.

David Shull, Department of Environmental Sciences, Western Washington University Allan Devol, School of Oceanography, University of Washington (UW) Rachel Horak, School of Oceanography, UW

The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Moving from the Northern Bering Shelf toward the south (that is, toward the middle shelf at similar water Integrated Ecosystem Research Program and the depths), and from the middle shelf to deeper water, the relative importance of aerobic respiration increases National Science Foundation’s Bering Ecosystem and anaerobic respiration, especially sulfate reduction, decreases in importance. Study. www.nprb.org/beringseaproject

Hot Spots in the Bering Sea 24-HOUR DINERS FOR SEABIRDS AND WHALES If you have to hunt for your large patches. But does the way Fig. 1 food in the cold and stormy these predators hunt, and whether 180˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” Bering Sea, finding predictably they are tied to a breeding site, dense patches, or persistent “hot affect their ability to respond to spots” of your favorite prey saves these dense patches of prey or 60˚0’0”N you time and energy, and may changes therein? 60˚0’0”N make the difference between survival and starvation. But what How We Did It happens if you are a seabird or a At sea, we examined distribu- 55˚0’0”N 55˚0’0”N fur seal and changes in the ocean tions of surface-feeding black- make these hot spots less predict- legged kittiwakes (Rissa tridactyla) 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” able during a time when you have and pursuit-diving thick-billed to regularly return to the place murres (Uria lomvia)) during their We found euphausiids all over the place, with where you nurture your young? summer nesting period when their persistent hot spots within specific 37 × 37 Would the change matter if you foraging range is limited. We also kilometer blocks are a migratory whale that is not looked at free-ranging humpback tied to one place, and is just in (Megaptera novaeangliae) and fin Fig. 2 Alaska to take advantage of the whales (Balaenoptera physalus). We ocean’s summer bounty? studied the distribution of all four 180˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” We know these ocean predators species in relation to two of their often exploit places where small key prey: age-1 walleye pollock fishes and zooplankton persist in continued on back side 60˚0’0”N 60˚0’0”N

The Big Picture In Alaska, seabirds, whales, fishes, and plankton are abundant on the Bering Sea shelf and slope, 55˚0’0”N a productive ecosystem supplying food for millions of seabirds and tens of thousands of marine 55˚0’0”N

mammals. In this study, we tackled the Bering Sea Project hypothesis that climate and ocean condi- 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” tions influencing circulation patterns and physical domain boundaries will affect the distribution, frequency and persistence of fronts and other oceanographic features that concentrate prey, and Age-1 pollock were patchier and their hot spots affect the foraging success of marine birds and mammals largely through bottom-up processes. persisted only on scales greater than 37 kilometers. We quantified the distributions of open-ocean prey species and determined that marine predators often associated with areas where these “hot spots of prey” persist at certain times of the year for several years. But we also wanted to determine whether the hunt for food differed among species that were tied to a colony or not, or between animals built to dive for their food versus those that can fly long distances but must feed on the surface. Our conclusion: being tied to a central place matters, as does the way you look for food.

TOP PREDATOR HOTSPOT PERSISTENCE A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. (Theragra chalcogramma) and euphausiids (zooplankton of the family Euphausiidae). We surveyed the prey once each year during 2004 and 2006-2010, and surveyed the seabirds in 2006- 2010 and the whales in 2008 and 2010. We compared the seabird and whale locations to where age-1 pollock and euphausiids were concentrated and considered how long these concentrations were present in time and space on an annual scale. This allowed us to compare this measure of prey persistence among

Alexandra Ravelo Alexandra annual surveys. Seabirds, whales and other ocean predators seek out persistent “hot spots” where small fishes and other prey gather in dense patches. Why We Did It The ability to remember the loca- 180˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” 180˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” tion of preferred prey is an important part of the foraging behavior of whales, seabirds, and other ocean 60˚0’0”N 60˚0’0”N 60˚0’0”N 60˚0’0”N predators. An important characteristic of these prey concentrations is their persistence in time and space,

55˚0’0”N 55˚0’0”N which allows predators to predict or 55˚0’0”N 55˚0’0”N Fig. 3 Fig. 4 remember their locations and concentrate search efforts accordingly. 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” Predictable prey locations reduce Both kittiwakes and murres, despite the difference in their feeding style, were consistently associated search time and thus the energetic with age-1 pollock but not consistently with hot spots of euphausiids, even though the euphausiids hot spots were more persistent than those of the small fish. The diving thick-billed murres, which have costs of foraging. Predators tied to greater travel costs than kittiwakes, foraged on prey concentrations nearer their island colonies than a location to incubate and rear their did the surface-feeding black-legged kittiwakes, which foraged widely. young face the additional challenge of locating prey close enough to

180˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” 180˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” their colony to frequently feed their young.

Mike Sigler, NOAA Alaska Fisheries Science Center

60˚0’0”N 60˚0’0”N Nancy Friday, NOAA National Marine Mammal Laboratory 60˚0’0”N 60˚0’0”N Kathy Kuletz, US Fish and Wildlife Service Patrick Ressler, NOAA Alaska Fisheries Science Center Chris Wilson, NOAA Alaska Fisheries Science Center Alex Zerbini, NOAA National Marine Mammal Laboratory

55˚0’0”N 55˚0’0”N 55˚0’0”N 55˚0’0”N Fig. 5 Fig. 6 The Bering Sea Project is a partnership between

175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” 175˚0’0”W 170˚0’0” 165˚0’0”W 160˚0’0” the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the Humpback and fin whales were not tied to a central place. We found humpback whales only where National Science Foundation’s Bering Ecosystem euphausiids were concentrated and where these concentrations were persistent. We observed fin Study. www.nprb.org/beringseaproject whales where age-1 pollock were more likely to occur, similar to black-legged kittiwakes and thick- billed murres, but their association with euphausiids was unclear.

Warm and Cold Years in the Southeastern Bering Sea WEATHER MATTERS

The Bering Sea is ice-free during persists in the south for many weeks summer, but beginning in November in early spring. The Big Picture or December, sea ice begins to form Sea ice plays an important role The transition between the Arctic along the coast. In January and in the physics and biology of the and subarctic occurs in the southern February, strong winds out of the eastern Bering Sea. It results in Bering Sea, and the boundary between the north push the ice southward 1,000 colder spring ocean temperatures, two is very sensitive to climate changes. km, covering much of the shelf. Air an early ice-associated phytoplank- Changes in the temporal patterns of temperature and the timing and per- ton bloom, a less saline water variability can also impact this system. sistence of these “arctic blasts” varies column and a summer cold pool Seasonally icy seas, like the Bering Sea, widely from year to year. While there where temperatures in the bottom respond differently to changes in ice is always ice on the northern shelf water layer remain below 2 °C. cover than Arctic seas that presently in winter and much of spring, the In February 2000, the south- have year-round ice. Recent multiple maximum southern extent of the eastern Bering Sea entered an consecutive years of warm conditions ice can vary by 100s of kilometers almost 6-year period of little ice with less ice in winter and spring yielded between years (Figure 1). In “warm” and “warm” conditions. After a fewer large zooplankton, an important years, there is little ice in March transition year in 2006, extensive prey species in this ecosystem, and led to lower pollock recruitment. and April south of latitude 57° sea ice returned to the southern 30’, whereas in “cold” years the ice shelf, and these cold conditions continued on page 2 Fig. 1

(A) Average number of days in which sea ice was present in March and April during 2001-2010. The anomalies of sea-ice coverage during March and April during (B) the cold years, 2007-2010, and (C) the warm years, 2001-2005.

BIOPHYSICAL MOORINGS A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. were still present in 2013. Scientists How We Did It ships maintaining the moorings originally hypothesized that warmer We utilized a wide range of provided data on prey availability. conditions would favor walleye data from cruises, moorings, pollock and other fish species that the National Snow and Ice Data Why We Did It prefer temperatures above 2°C; Center, and Alaska Fisheries Science The southern Bering Sea is a however, with warmer conditions, Center trawl surveys to examine the rich ecosystem that supports large there was a sharp decrease in the relationship between ice in March numbers of marine mammals and availability of key prey items for and April and depth-averaged tem- seabirds, and provides approxi- young-of-the-year pollock, limiting perature from long-term mooring mately 40% of the U.S. catch of the survival of fish during their first on the southeastern shelf, M2. The fish and shellfish. We now know winter (Figure 2). An interesting timing of the spring phytoplankton that changes in the weather patterns question that remains is “has the blooms was also obtained from the and ice extent over the southern Bering Sea shifted from strong year- chlorophyll fluorescence data on the shelf affect zooplankton abundance to-year variability to a multiyear moorings, showing that when ice and distribution patterns, which in pattern, which is more common in was present after mid March on the turn impacts the fishes, large baleen the Gulf of Alaska?” Such a change southern shelf, there was an increase whales, and seabirds that feed in would have important repercussions in fluorescence and a decrease in these waters. Climate models predict on this ecosystem. nutrients. Plankton net tows from that the southern Bering Sea will become warmer, with reduced sea Fig. 2 ice in the next couple of decades. If the warm period (2000-2005) is any indication of how the ecosystem will respond to warming, such a change will strongly affect the existing ecosystem. Understanding how shifts in climate impact this system will help scientists predict who the winners and losers could be, and provide the opportunity to help cushion the impact of the changes on humans who utilize this ecosystem.

Phyllis J. Stabeno, National Oceanic and Atmospheric Administration (NOAA) Pacific Marine Environmental Laboratory Jeffrey Napp, NOAA Alaska Fisheries Science Center Lisa Sheffield Guy, University of Washington Joint Institute for the Study of the Atmosphere and Ocean Mike Sigler, NOAA Alaska Fisheries Science Center

Literature cited Ianelli, J., T. Honkalehto, S. Barbeaux, S. Kotwicki, K. Aydin and N. Williamson (2012). Assessment of the walleye pollock stock in the eastern Bering Sea. 2012 North Pacific Groundfish Stock Assessment and Fishery Evaluation Reports for 2013.

Large crustacean prey and year class strength of walleye pollock. (Top) Abundance of copepods (Calanus The Bering Sea Project is a partnership between spp) and adult and juvenile euphausiids (krill) sampled during the summer. Copepods were sampled the North Pacific Research Board’s Bering Sea with plankton nets and euphausiid abundance was estimated with acoustics. (Bottom) Estimated num- Integrated Ecosystem Research Project and the ber of pollock surviving to age-1 for each year class. Estimates obtained from stock assessment models National Science Foundation’s Bering Ecosystem (Ianelli et al., 2012, Table 1.23). Study. www.nprb.org/beringseaproject

Summer Microzooplankton in the Bering Sea THEIR SURPRISING ROLE

In spring, large diatom blooms microzooplankton over phyto- occur in the Bering Sea that sup- plankton. Microzooplankton can ply food for large zooplankton that be an important link between phy- in turn are food for many small toplankton production and higher fish, seabirds and whales. In sum- trophic levels, especially when

mer, the big diatoms blooms are phytoplankton are scarce or small Diane Stoecker gone from the surface waters and in size. Our goal was to determine Microzooplankton include heterotrophic dinoflagel- most of the phytoplankton are too if microzooplankton are important lates, such as this Gyrodinium with ingested prey. small for zooplankton to eat. What as a potential food source for large do zooplankton eat in summer zooplankton in summer. The Big Picture in the Bering Sea? We thought We addressed a gap in knowledge of that part of the answer might be How We Did It planktonic food webs in the Bering Sea by microzooplankton, which are We went on month-long cruises examining the role of microzooplankton in microscopic, one-celled organisms in summers of 2008, 2009 and 2010 summer. We found that microzooplankton that eat small phytoplankton. The on which we collected water from were very abundant, particularly in shelf microzooplankton, although tiny different depths with Niskin bottles waters, and were a major food source avail- by our standards, are big enough on a a conductivity, temperature, able to larger zooplankton. By examining to be captured and ingested by and depth recorder (CTD) rosette differences in the role of microzooplankton large zooplankton. In fact, many (Figure 1). We used the water to in regions with different physical forc- large zooplankton prefer to eat continued on page 2 ing (i.e., stratification, mixing, and other oceanographic features) and comparing the Fig. 1 role of microzooplankton among years, our study specifically addressed the Bering Sea Project hypothesis that “climate-induced changes in physical forcing will modify the availability and partitioning of food for all trophic levels through bottom-up processes.” We found that there are important regional differences in the Eastern Bering Sea, with differences in physical forcing affecting the role of microzooplankton. Comparison of our data from three “cold” years to previously collected data from “warm” years show that microzooplankton are an important component of planktonic food webs in the Eastern Diane Stoecker Kristen Blattner removes water from a Niskin bottle on the CTD rosette to determine the quantity of Bering Sea in both “warm” and “cold” years. microzooplankton.

SUMMER MICROZOOPLANKTON IN THE BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Fig. 2 so that we would know how impor- the northern shelf there were high tant they were relative to phyto- concentrations of phytoplankton at plankton as a potential food source depth. These deep concentrations for zooplankton. We also wanted to of phytoplankton were probably determine how much of the phy- remains of the spring bloom that toplankton production was eaten had sunk out of the surface waters. by microzooplankton. Although Microzooplankton are an impor- phytoplankton stocks are low in tant component in these deep, cold summer, there are “hotspots” along layers of plankton that may serve the shelf edge and near the Pribilof as “refrigerators full of food” for Islands where there are more phyto- zooplankton in summer, forming plankton than on most of the shelf. an important link in food webs that We wanted to determine if the support higher trophic levels in importance of microzooplankton summer on the Eastern Bering Sea was greater in areas with lower phy- Shelf. toplankton stocks than in areas with Diane Stoecker, University of Maryland Center for higher phytoplankton stocks. Environmental Science, Horn Point Laboratory We observed that in summer the Alison Weigel, University of Maryland Center for presence of microzooplankton was Environmental Science Kristen Blattner, University of Maryland Center for very important in surface waters Environmental Science John Dolan over much of the Bering Sea Shelf Microzooplankton include planktonic ciliates, such because they dominated the size as this tintinnid from the Bering Sea. The Bering Sea Project is a partnership between class of plankton that is the right the North Pacific Research Board’s Bering Sea conduct grazing experiments at sea size food for large zooplankton. On Integrated Ecosystem Research Program and the to estimate the amount of phyto- the middle and inner shelf, where National Science Foundation’s Bering Ecosystem plankton eaten by microzooplank- phytoplankton are scarce in sum- Study. www.nprb.org/beringseaproject ton. We did this by incubating the mer, microzooplankton biomass was water in flowing seawater incubators higher than phytoplankton biomass! on-deck and measuring changes in This was a bit puzzling, because in chlorophyll a (green plant pigment food webs, there is usually a higher that is a proxy for phytoplankton biomass of “grass” than “cows.” Part biomass) in bottles with different of the answer to this puzzle is that concentrations of microzooplank- many of the microzooplankton were ton. Some of the water we preserved large, green ciliates (Figure 2) that and brought back to our labora- are grazers on small phytoplankton tory at the University of Maryland but are photosynthetic and can also

Center for Environmental Science, produce their own food. We found Diane Stoecker so that we could examine it under that in surface waters on the middle A thecate heterotrophic dinoflagellate, also a a microscope. We used our micro- shelf, ciliates sometimes contributed member of the microzooplankton. scopic observations to identify, over 50% of the chlorophyll! count, and estimate the biomass of Microzooplankton were also microzooplankton. important as grazers on phytoplankton; they consumed almost all Why We Did It of the daily phytoplankton pro- We set out to determine the duction on the middle and inner abundance and biomass of micro- shelf. Although phytoplankton zooplankton, and to compare their were generally low in abundance biomass to phytoplankton biomass in surface waters on the shelf, on

A Lasting Legacy of the Bering Sea Project ARCHIVAL AND PRESERVATION OF THE PROJECT DATA FOR CURRENT AND FUTURE RESEARCH

In a collaboration called the physical oceanography, plankton, How We Did It “Bering Sea Project,” the National fishes, seabirds, marine mammals, The data management support Science Foundation (NSF) sup- humans, traditional knowledge and for the BEST and BSIERP Programs ported the Bering Ecosystem Study economic outcomes. The resulting developed independently during the (BEST) and the North Pacific 356 datasets established a new para- initial years of data collection. Later Research Board (NPRB) developed digm for critical information needed the support was consolidated, but the and supported the Bering Sea Inte- to answer key questions about these principles of support for investigator grated Ecosystem Research Program changes. The Earth Observing datasets remained firm as NSF and (BSIERP) to address changes in this Laboratory (EOL) of the National NPRB worked together to develop a critical marine ecosystem. More Center for Atmospheric Research single archive at EOL for the Bering than 100 scientists engaged in field (NCAR) brought 25 years of expe- Sea Project. The Bering Sea Data data collection, original research rience to provide all facets of data Archive at http://beringsea.eol.ucar. and ecosystem modeling during the management support to the Bering edu (Figure 1) is the single source for Bering Sea Project to link climate, Sea Project. all data from this collaborative effort. continued on page 2 Fig. 1 The Big Picture The Bering Sea project data archive developed by and housed at the NCAR/EOL will remain the long-term legacy of the Bering Sea Project. More than 100 investigators deployed during different seasons to document the ecosystem and related oceanography and meteorology of the region. Not only was the volume of available data from the region significantly increased during the Bering Sea Project, the data coverage in space and time extended into previously unsampled domains. While much has been learned from the initial and ongoing analyses of these data, they will continue to provide fodder for future analyses in response to unanticipated and serendipitous observations, to serve as model forcing and validation resources, and to define baselines against which future ecosystem changes may be evaluated. EOL developed the archive, uploaded datasets and documentation from users, provided web Bering Sea Project Database example entry. It is possible to search 356 access to the data and has assumed long-term stewardship of this combined BEST and BSIERP datasets by project, cruise and science subject. The unique data archive for the benefit of science and society as they resulting table provides a direct link to the dataset and documentation for easy download and access. seek to better understand the Bering Sea ecosystem.

BERING SEA PROJECT DATA MANAGEMENT A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. Our comprehensive data man- current ship track and position, efforts and accomplishments of agement strategy included early all ship–based sampling stations 100 investigators over more than involvement with the science team from the current or any previous six years. The analyses of those data- to determine their requirements and cruise (critical in the repeat loca- sets are already revealing important establish priorities based on available tion sampling strategy used during information about the make-up of resources, resulting in a clear specifi- the project) and any operational this unique ecosystem in the Bering cation for metadata and documenta- products (e.g. satellite, sea ice edge, Sea. Ongoing analysis efforts are tion to accompany all datasets. The currents) used for real time cruise enhanced by a high-quality data easily accessible database cross-refer- track selection. The BEST Field archive that assures consistent access ences each unique investigator data- Catalogs for each cruise remain to all of the valuable data. EOL set (Figure 1). Researchers can peruse active via the EOL website for provides the long-term stewardship data inventory through a search future reference. of the Bering Sea Project data using tool to access data listed in tables by EOL also worked closely with a the established capabilities of the cruise, subject category or investiga- Bering Sea Project ethnographer to NCAR archive system. tor’s name. Ongoing maintenance develop a Geographic Information Don Stott, Earth Observing Laboratory, National Center for of the metadata assures its long-term System (Figure 3) user tool for Atmospheric Research accuracy, thus allowing future consis- displaying detailed data and infor- Janet Scannell, Earth Observing Laboratory, National Center for tent access and data discovery. mation collected during the Nelson Atmospheric Research James Moore, Earth Observing Laboratory, National Center for EOL provided specialized sup- Island project, including place Atmospheric Research port to the 10 NSF-sponsored names and links to stories and pho- Steve Williams, Earth Observing Laboratory, National Center for Bering Sea Project cruises on tos by location. Atmospheric Research Amanda Orin, Earth Observing Laboratory, National Center for multiple ships (2007-2010) that Atmospheric Research included implementation of a Why We Did It BEST Project Field Catalog (Figure After all of the data collection The Bering Sea Project is a partnership between 2) for use aboard ship. It allowed is a distant memory, there will be the North Pacific Research Board’s Bering Sea real-time documentation of data a single, unified data legacy for the Integrated Ecosystem Research Program and the collection to be uploaded by the Bering Sea Project. The 356 datasets National Science Foundation’s Bering Ecosystem science team, heads-up displays of will help mark the extraordinary Study. www.nprb.org/beringseaproject

Fig. 2 Fig. 3

The EOL BEST Field Catalog deployed on various ships during 10 cruises from EOL-developed LTK GIS mapserver interface for Nelson Island. The image 2007-10. Image shown is the cruise track and stations during the R/V Knorr here shows a sampling of the place names acquired during the multi-year summer 2009 cruise ((KN195-10). The catalog also provides operational study. Each place name is color or symbol coded to identify the specific type and research data products, station reports and preliminary research of site (e.g. burial, hunting, historical). Each site is an active link to related analysis products. photos and stories about that specific location.

A Moveable Feast The Big Picture SEABIRDS TRACK PREY IN THE SOUTHEAST BERING SEA Few studies of seabirds in Alaska have occurred outside of summer months. Seabirds have to find enough food How We Did It We found that seabirds appeared to be responding to broad-scale changes in to raise their chicks at the colony We measured the abundance and seasonal prey distribution, with a greater in summer, which restricts how far distribution of predators and prey in dependence on forage fish over the middle they can search for food. Afterwards, the southeast Bering Sea in summer and inner shelf in fall. Fall may be as as they prepare for migration and and fall (2008–2010) by conducting crucial as summer to the health of seabird winter survival, birds can more freely seabird surveys from the same fisher- populations, and changes in forage fish search for prey. The breeding and ies research vessels used to estimate distribution in fall due to climate change post-breeding periods thus pose krill and fish abundance. could be bad for seabirds. At the least, it different challenges for seabirds. We In summer, seabirds tended to be could affect the species composition of looked at seabird response to prey ‘clustered’ and occurred near breeding the seabird community and where or how distribution in summer and fall, as colonies, where they foraged primar- much seabirds aggregate. Understanding an indication of how seabirds might ily on the outer shelf. In fall, seabirds mechanisms affecting seabird-prey respond to changes occurring in the were more abundant overall, with a relationships is critical to ecosystem-based management in a changing climate. Bering Sea. continued on page 2

Fig. 1

Distribution of seabirds and key prey in summer (top panels) and fall (bottom panels) in the study years 2008-2010. Seabird numbers are represented by scaled circles, and prey density is shown as light to dark shading.

SEABIRD BROAD-SCALE DISTRIBUTION A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. USFWS Sophie Webb DeCicco Luke Millions of short-tailed shearwaters sometimes A newly fledged black-legged kittiwake. Age-0 juvenile walleye pollock, a key prey for aggregate at Unimak Pass in July, along with seabirds in the Bering Sea. humpback whales. greater variety of species, as southern key driver of seabird distribution in these energetic demands, we studied migrants moved in for the feast of summer was colony location, whereas what factors influenced their distribu- available fish and euphausiids (also in fall, the distribution of forage fish, tion, and whether they used different known as krill). In addition, newly including capelin and juvenile pol- areas of the Bering Sea in summer fledged juveniles joined adults at sea. lock, was also important. than in fall. Seabirds were also less clustered in Kathy J. Kuletz. U.S. Fish and Wildlife Service fall, and they dispersed throughout Why We Did It Robert Suryan, Oregon State University the outer and middle shelf domains Summer and fall can be energeti- Sandra Parker-Stetter, University of Washington (Figure 1). Even shearwaters, which cally demanding periods for seabirds. Patrick H. Ressler, Alaska Fisheries Science Center Martin Renner, Tern Again Consulting, Homer, AK do not breed in Alaska and thus In summer, breeding seabirds need John Horne, University of Washington were not tied to colonies in summer, to acquire enough prey close to their Ed Farley, Auke Bay Lab, Alaska Fisheries Science Center clustered near the Alaska Peninsula in breeding colony to feed rapidly grow- Elizabeth A. Labunski, U.S. Fish and Wildlife Service summer where krill were abundant, ing chicks. Fall is a different type of but in fall, they dispersed throughout energy ‘bottleneck,’ because seabirds The Bering Sea Project is a partnership between the shelf and farther north. need to replenish depleted fat reserves, the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the There were inter-annual variations undergo feather molt and replace- National Science Foundation’s Bering Ecosystem in seabird abundance and response ment, and prepare for migration. To Study. www.nprb.org/beringseaproject to prey distribution, but overall, a better understand how seabirds meet Sandra Parker-Stetter Euphausiids, or ‘krill’, are also important prey for many seabird species; these krill were sampled during a Bering Sea Project trawl.

Understanding Bering Sea Groundfish Populations USING MODELS TO SHED LIGHT ON PATTERNS AND TRENDS

We developed simulation models pool of cold bottom water after cold We developed and evaluated two of predator-prey relationships that winters may provide refuge for juve- alternative multispecies fish models allowed us to reproduce observed niles, reducing cannibalism by adult for the eastern Bering Sea that con- changes in populations of pollock, pollock. Because of the dominant sider the interactions among walleye cod, and flatfish in the eastern abundance of pollock, the net effect pollock (separated into juvenile and Bering Sea since the 1980s. We of warmer temperatures is increased adult groups), Pacific cod, arrow- learned that, in warm years, age-1 juvenile mortality, resulting in fewer tooth flounder, and a small-mouth juvenile pollock were more heav- survivors to grow to adults to sup- flatfish group comprising yellowfin ily eaten by arrowtooth flounder port future fisheries. We continue sole, flathead sole, northern rock and cannibaliz ed by adult pol- to explore ways that environmental sole, and Alaska plaice. One type, lock, whereas fewer age-2 pollock conditions alter these relationships, called a biomass dynamics model, were eaten by cod. These different and to evaluate their implica- generally performed better than the temperature responses likely reflect tions on fishery management and continued on page 2 different thermal preferences by expected future fishery yields. species, which may change with life stage. For instance, a lingering How We Did It The Big Picture Fig. 1 To understand variability of multiple species in the ocean, scientists often develop whole ecosystem models that attempt to explain the flow of energy from phytoplankton throughout the marine ecosystem. Such ecosystem models tend to be very complicated and require large quantities of data, many assumptions, and large teams of modelers and other researchers. Instead, we developed simpler multispecies models, informed by routinely collected assessment and ecological data, to better understand patterns and trends of the most commercially important fish species in the eastern Bering Sea. Results are intended to foster an improvement in Predator-prey relationships among eastern Bering Sea fish species included in this study. Arrows represent the directions of predator prey. Predator-prey relationships were inferred from the collective sustainable management the contents of fish stomachs sampled during the eastern Bering Sea bottom trawl surveys by the Alaska of these important fishery resources. Fisheries Science Center.

CORRELATIVE BIOMASS DYNAMICS MODEL A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. other type, called a delay difference abundance are not independent. Fig. 3a model. Both models describe the Good years for pollock and cod predator-prey interactions among reproductive success tend to coin- these five groups of fish. Because cide, and display patterns opposite juvenile pollock serve as prey for those for flatfish. Are these trends a all species, juvenile pollock were result of see-saw patterns in preda- modeled separately from adults. tors and prey, or due to species’ Our model was developed based on responses to environmental varia- many years of fish stomach samples tions, or perhaps a result of com- collected by the National Marine mercial fishing? Our study intends Fisheries Service, Alaska Fisheries to develop a deeper understand- Fig. 3b Science Center. Once we worked ing of interactions among major out the predator-prey interactions, groundfish species in the Bering we used our models to examine the Sea, thereby fostering a joint man- effects of fishing and environmental agement approach that acknowl- factors on these groundfish species edges ecological interactions of based on findings from companion these species and the combined studies. effects of climate and fishing.

Tadayasu Uchiyama, School of Fisheries and Ocean Sciences, Why We Did It University of Alaska Fairbanks (UAF) Landings of pollock, cod and Gordon H. Kruse, School of Fisheries and Ocean Sciences, UAF Franz J. Mueter, School of Fisheries and Ocean Sciences, UAF flatfish account for more than half (a) Model-estimated biomass of age-1 pollock of all U.S. commercial fishery land- and age-1 pollock biomass lost to predation by ings. Annual catch limits are set for The Bering Sea Project is a partnership between adult pollock, arrowtooth flounder, Pacific cod, each species individually, based on the North Pacific Research Board’s Bering Sea and small-mouth flatfishes (values in 1000 metric assessments of their abundance and Integrated Ecosystem Research Program and the tons); and (b) Effect of bottom temperature on predation of age-1 juvenile pollock by these same productivity. Yet, patterns in fish National Science Foundation’s Bering Ecosystem Study. www.nprb.org/beringseaproject predators. The x axis is bottom temperature in °C, whereas the y axis shows estimated biomass of Fig. 2 juvenile pollock lost to predation, expressed as a percentage relative to the biomass lost to predation at the mean bottom temperature of 2.25°C

Schematic diagrams showing alternative hypotheses on how cold climate may affect distribution of fish on the eastern Bering Sea shelf and predation on young pollock. The cold pool (blue) is a pool of cold water (< 2°C) on the Bering Sea shelf formed by melting sea ice. In cold years, the cold pool covers a large portion of the middle shelf region. Most fish species are driven to the outer shelf region by the cold water, where predation is intensified by increased prey and predator density (a). However, there is some evidence that young pollock (major prey for other fish, including adult pollock) are more tolerant to cold water, in which they are protected from predators (b). If this is the case, predation on young pollock would decline under cold climate and increase under warm climate.

The Contribution of Dissolved Iron from Melting Ice in the Bering Sea SEA ICE AND IRON – ESSENTIAL SPRINGTIME ROLES Like humans, algae require its formation. When ice melts the trace metal iron for healthy in the spring, these Fe-rich min- growth. We tested the hypoth- eral particles are released into esis that although the initial algal the water column, and a portion growth in spring depletes available of the particulate Fe becomes iron (Fe) in the winter-mixed sur- dissolved, contributing to the face water of the Bering Sea shelf, available Fe flux to the stratified resulting in limited algal growth, surface water. This additional Fe Nomura Jennifer the input of Fe from melting ice source is especially important to relieves this limitation. the spring bloom, as vertical mix- The Big Picture Sea ice can be an important ing of iron-rich subsurface waters We explored the role of sea ice in source of available Fe to the is inhibited by the strong water delivering dissolved iron (DFe), essential surface ocean (Figure 1). Fe-rich column stratification brought to the health of phytoplankton, and particles derived from eolian about by the creation of a surface found that areas of the Bering Sea outer deposition, fresh water runoff, low-density layer of water when shelf not influenced by ice contain insuf- and sediment suspension can be the ice melts. ficient DFe for the complete assimilation incorporated into sea ice during continued on page 2 of available nitrate by algae. In contrast, outer shelf areas influenced by melt- Fig. 1 ing sea ice contained sufficient DFe concentrations to support complete biological utilization of nitrate. In addition to providing water column stability, melting sea ice provides a source of DFe to the outer shelf that is important in maintaining ice-edge algal blooms. In the absence of this input, diatom productivity over the outer shelf and shelf break may become limited by iron during spring. Variability in sea ice extent is likely to translate into a varying supply of DFe to the Bering Sea outer shelf and shelf break in early spring, and thereby contribute to changes in the timing and community composition of the spring phytoplankton bloom. Pathways of iron supply from melting sea ice to the water column.

THE ROLE OF ICE MELTING IN PROVIDING AVAILABLE IRON TO THE SURFACE WATER OF THE EASTERN BERING SEA SHELF A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. How We Did It magnitude higher than the DFe in Our dissolved iron (DFe) mea- the ice core. If only a portion of this surements from both water column is bioavailable, it represents a further samples and ice cores collected substantial source of Fe from the during the 2007- Bering Sea Project melting ice. cruise indicate that the melting ice provided substantial DFe to Why We Did It the water column. This additional The Bering Sea is one of the DFe input, while highly variable most productive regions in the (Figure 2), was particularly impor- world. It exhibits a band of excep- tant to shelf break surface waters. tionally high productivity along In the absence of this Fe source, the the shelf break during spring and concentration of DFe in the surface summer. Previous observational water would not be sufficient to and modeling studies indicate that Erika Acuna allow algae to utilize fully the high changes in the seasonal ice cover The timing of the spring bloom nutrient concentrations observed in influence open-water productivity also affects the transfer of energy to the outer shelf, and the productivity (the timing of the spring bloom upper trophic levels. Because the of this area would be limited below and the composition of the phy- outer shelf contains much higher its full potential. The particulate Fe toplankton community) over the concentrations of macronutrients, in the ice cores was 1 – 2 orders of Bering Sea shelf and shelf break. particularly nitrogen, than the middle and inner shelf, an ice edge Fig. 2 that reaches the outer shelf in spring has the potential to support a larger 65 ˚N ice edge phytoplankton bloom as 6 3 2 ice begins to melt. However, high macronutrient concentrations in the 1 4 60 ˚N 7 outer shelf can only be fully assimi- 5 lated when enough iron is available. We determined the acutely

55 ˚N important influence of ice melt on the distribution of DFe, compared to available macronutrients in this area, and its possible implica- 50 ˚N 175 ˚E 180 ˚E 175 ˚W 170 ˚W 165 ˚W 160 ˚W 155 ˚W tions for the spring phytoplankton bloom.

Jingfeng Wu, Rosenstiel School of Marine and Atmospheric Science, University of Miami Ana Aguilar-Islas, Institute of Marine Science, University of Alaska Fairbanks Rob Rember, International Arctic Research Center, University of Alaska Fairbanks

The Bering Sea Project is a partnership between the North Pacific Research Board’s Bering Sea Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem Dissolved iron concentrations from replicate samples in the sea ice of the Eastern Bering Sea shelf in Study. www.nprb.org/beringseaproject April/May 2007, with data from the Arctic and Antarctic Oceans for comparison.

Aging Murres in a Warming Sea OLD AGE AND EXPERIENCE BEAT YOUTH IN POOR CONDITIONS Thick-billed murres (Uria lom- adapting to dramatic manmade and via) are a common seabird in the climate-induced changes in their Bering Sea. Better understanding environments. of their demography and life history is crucial to predicting their role as How We Did It an indicator of a changing ecosys- We measured telomere length tem and how they may respond (a DNA marker) as an indicator to worsening conditions or future of biological age, and compared it changes in the dominant climate pat- among three murre colonies in the terns of the Bering Sea. Our project southeastern Bering Sea that have explored the relationship between contrasting environmental condi- biological age and environmental tions and population trajectories. conditions in thick-billed murres Providing a more accurate picture with breeding grounds in the Bering of an organism’s true aging process, Sea. Murres are long-lived animals, biological age is a measure of aging and in the wild may survive up to a that integrates chronological age PAPISH RAM venerable age of 30 years old, while continued on page 2

Fig. 1 The Big Picture The Bering Sea system is characterized by ice covered winters and complex interactions of food webs and water masses during the summers. Seabirds are top predators, and they act as land-based indicators of various changing marine signals: fish stocks, timing of annual marine food web changes, and climate-related fluctuations in the environment. Some long-lived seabirds, with lifespans easily reaching 30 years, may have witnessed two or more radical regime shifts in the environment. Our work has demonstrated that longevity is an important factor in how seabird populations respond to their environment: age and Nutritional stress (measured by corticosterone levels) increases with age on high quality colonies (left: Bogoslof and environmental conditions interact to explain St. George) and decreases on poor quality colonies (St. Paul). Note that since telomere quantity decreases with age, the stress levels that affect reproduction and the x-axes run from large quantities to small so as to run from young to old, as a chronological age axis would. survival of breeding murres.

PATCH DYNAMICS STUDY, PRIBILOFS AND NORTH BERING SEA A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. (time since birth) with the effects fit in obtaining food than younger their quality and ability, based on of stress, reproduction, and indi- individuals. When food is plenti- biological age. If the environment vidual variability. As organisms ful, prior experience in finding food becomes less predictable, is it better age, their physiological deteriora- is less important, but as conditions to have populations comprising tion may lessen their ability to meet worsen, the experience of older younger or older birds responding environmental challenges. On the individuals becomes beneficial. to that situation? Is a colony of old other hand, older birds may be Under the worst conditions, all birds in trouble, or will it be more more experienced and be better at birds become food limited; here likely to weather poor foraging responding to the environment: older birds outperform younger conditions successfully? Knowing they know where to forage when ones, as their experience in finding that younger birds are poor foragers conditions have changed. food and weathering tough years or that older birds are more stressed Our results demonstrate that becomes more important than their could help explain why some colo- stress levels of breeding thick-billed failing physiology. nies do well and others decline. murres depend on an interaction Rebecca Young, Institute of Arctic Biology (IAB), Department of of colony conditions and biological Why We Did It Biology and Wildlife, University of Alaska Fairbanks (UAF) age (Figure 1). When conditions Although population modeling Chris Barger, IAB, Department of Biology and Wildlife, UAF are favorable, such as on Bogoslof estimates demographic parameters, Ine Dorresteijn, IAB, UAF Island, or relatively stable (e.g., St. the age structures of wild popula- Sasha Kitaysky, IAB, Department of Biology and Wildlife, UAF George), biologically older birds tions are often unknown. However, have higher stress levels, likely knowing the makeup of populations The Bering Sea Project is a partnership between due to the effects of aging. When and how different age classes per- the North Pacific Research Board’s Bering Sea conditions are poor, such as on form in the environment is crucial Integrated Ecosystem Research Program and the National Science Foundation’s Bering Ecosystem St. Paul Island, biologically older to our understanding of animals’ Study. www.nprb.org/beringseaproject birds have lower stress levels. We responses to that environment. concluded that older birds are more Especially in long-lived organisms, experienced, but also might be less like seabirds, adults can vary in

Murres breed in dense colonies on sea cliffs in the circumpolar north. RAM PAPISH RAM

Steps Toward Predicting the Future of the Bering Sea Fish Catch HOW COMPUTER MODELS HELP FISHERMEN FIND THE “COLD POOL” ... AND DINNER

With the Bering Sea bringing in how this cold pool will change and over 50% of the US fish catch, there where it will be found? We found The Big Picture are some obvious advantages to fish- that, using a good computer simula- Although the Bering Sea is small ermen, fish resource managers, and tion, to some extent, we can. compared to the world ocean, it is still markets if we can predict, even just too big to measure all of it at one time, by some months, which fish stocks How We Did It other than by satellites (which can only will do well. One clue to this is to We used a state-of-the-art measure the surface). Thus, to study the understand the creation and fate of computer model focused on simu- whole system, we created a virtual reality, a computer model of the system from the eastern Bering Sea “cold pool,” a lating the ocean and sea ice of the seafloor to the sea ice surface. This region on the Bering Sea shelf about the Bering Sea. To enable it to be virtual reality is based on our under- the size of California below 2°C near real-time, this BESTMAS standing of the physics, chemistry and (~36°F). This “cold pool” (which (Bering Ecosystem STudy ice–ocean biology of the real world and (crucially) changes through the year and from Modeling and Assimilation System) is tested against measurements we can year to year) is important for crab model is driven by atmospheric make. Researchers, managers and fish- and bottom-fish distributions. forcings from the weather forecast- ermen can then use this model as a tool For example, it acts as a barrier to ing models of the National Centers for understanding, as a framework for northward migration of some types for Environmental Prediction their measurements, and (given enough of fish, e.g., walleye pollock, one of (NCEP) and the National Center model skill) as a predictor of the system the largest and most valuable fisher- for Atmospheric Research (NCAR). and where, for example, to find the cold ies in the world. Can we predict pool ... and hence dinner. continued on page 2 Fig. 1

Eastern Bering Sea shelf bottom water temperatures (ºC) from ship-based surveys (right column) and corresponding BESTMAS model results (left column) for 2003 (a warm year, upper panels) and 2009 (a cold year, lower panels). The cold pool is marked by purple color. This figure shows that BESTMAS is able to capture reasonably well the spatial patterns of observed spring-summer bottom layer temperature fields and the distribution and extent of the cold pool (purple region) for both cold and warm years.

THE IMPACT OF CHANGES IN SEA ICE ON THE PHYSICAL FORCINGS OF THE EASTERN BERING ECOSYSTEM - RETROSPECTIVE INVESTIGATION AND FUTURE PROJECTION A component of the BEST-BSIERP Bering Sea Project, funded by the National Science Foundation and the North Pacific Research Board with in-kind support from participants. This model runs on a computer to predict what the spring and sum- cluster (connected computers that mer will be like some months in the work together) at the University of future. Washington, where simulating one year of the Bering Sea ice-ocean sys- Why We Did It tem takes about four hours of com- Quantifying the cold pool is a puter time. We found the model’s key part of predicting the balance of sea ice—one of the drivers of the fisheries in the Bering Sea. The loca- cold pool formation—matched well tion and duration of the cold pool with data from satellites. Moreover, changes a lot during the year and we found that the extent and loca- from year to year, dependent on tion of the cold pool in the model atmosphere, ocean and sea ice con- agreed well with ship data from the ditions during the previous winter. region, in the years where ship data All these preconditions interact, but was available (Figure 1). a coupled ice-ocean model such as So then we can use the model to BESTMAS allows us to combine all study what the cold pool looked like these effects in a physically consis- in years when there wasn’t ship data, tent manner, and make predictions and, most importantly, to consider of the cold pool location and extent why and how the cold pool forms, months in advance. and how the cold pool changes over Jinlun Zhang, University of Washington the season. From this, we found out, Rebecca Woodgate, University of Washington for example, that the simulated field Armchair Oceanography. The University of of bottom layer temperature on the Washington computer cluster MIZ (Marginal Ice Bering Sea shelf at the end of May is The Bering Sea Project is a partnership between Zone) used to run the BESTMAS model, alongside the North Pacific Research Board’s Bering Sea the creator of the BESTMAS model, Jinlun Zhang. a good predictor of the distribution Integrated Ecosystem Research Program and the and extent of cold bottom waters National Science Foundation’s Bering Ecosystem throughout late spring and summer Study. www.nprb.org/beringseaproject (Figure 2). Thus, we can use the model results from the end of May

Fig. 2

Simulated May 31 daily mean and June, July, and August monthly mean fields of bottom water temperature (˚C) for 2003 (a warm year) and 2009 (a cold year). Black line represents the 200 m depth contour. Purple shows areas of bottom temperatures below 2˚C, representing, on the Bering Sea shelf (i.e., between the 200 m contour and Alaska), the cold pool extent. This figure shows that the simulated field of bottom layer temperature on the Bering Sea shelf on May 31 is a good predictor of the distribution and extent of the cold pool throughout late spring and summer, for both cold and warm years.