Prince William Sound Herring Forage Contingency EVOSTC Project Thomas C. Kline, Jr. Ph. D., P. I. Robert W. Campbell Ph. D., Post-Doc Kevin Siwicke Tech. Prince William Sound Science Center Cordova, AK Alaska Marine Science Symposium 2009

Outline Context for project Questions & Objectives Field components Lab results from 2007 samples Zooplankton Chemistry (zooplankton -> fish) What is missing?

Context: SEA & GLOBEC SEA River-Lake Hypothesis: Bottom-up forcing – subsidies, larval drift SEA Prey Switching Hypothesis SEA Herring Over-wintering Hypothesis – age-0 herring fail to acquire sufficient energy to survive first winter Conceptual model from GLOBEC implementation plan (left) shows 3 dominant pelagic fish taxa -Interact at different LH stages -Interaction mediated by zooplankton What is not shown is the potential for herring to move in and out of PWS (question asked by D. Hay)

Questions & Objectives Zooplankton concentration, composition, & energy content SIA for oceanic subsidies in PWS: temporal & spatial variation, relationship to herring performance (WBEC, eventually recruitment) Herring foraging relationships with other fishes Project focus: over-wintering at age 0

Herring age 0+ samples to capture over-wintering- Fall (Nov) and Spring (Mar), (Thorne) Zooplankton samples- Spring bloom (May) – Fall ‘bloom’ (Sept) Two field components

Sampling 3 strata: Bay, Sound, Gulf Simpson Whale Zaikof Eaglek

Detailed lab results from 2007 samples : ZOOPLANKTON Energy content of zooplankton Euphausiids in PWS Herring as plankton Plankton community structure (see poster by Dr. Rob Campbell)

Plankton energy May (top) and September-October (bottom) Energy in terms of density with respect to water volume - measure of habitat quality; in terms of energy density per unit mass (ash-free dry weight = AFDW) of zooplankton- measure of food quality. There was approximately one order of magnitude less energy available per unit volume of water in the fall compared to spring, however, food quality was generally comparable. Spring samples with a number of euphausiids are depicted as pentagons; this may have been important as winter forage for herring (as it has in Sitka Sound)

Whale Bay was the most euphuasiid rich area in May Thysanoessa inermis dominated the euphausiid species composition in Whale Bay. Contributed a smaller fraction of those found elsewhere in the Sound and less still on the slope. May have been a contributing factor to the higher energetic content of herring during the preceding winter. T. inermis is a shelf species that was more abundant in colder years (Pinchuk et. al 2008). Euphausiids (krill)

Herring larvae and other ichthyoplankton in PWS, May There was about one fish larva per 1 to 10 m 3 of water. Two to three week old herring larvae contributed to significant portions of the ichthyoplankton composition in two of the nursery bays. A similar pattern is expected for the 2008 survey based upon samples picked for SIA. Herring as plankton

Hierarchical clustering analysis of plankton taxa in PWS September-October Map of the stations indicating the cluster identity of each station. Colors correspond to the identified clusters (Green = Bay cluster; Light blue = Eastern PWS cluster; Dark blue: Central PWS cluster; Red = Whale Bay cluster). Plankton community structure (see poster by Dr. Rob Campbell )

Detailed lab results from 2007 samples: CHEMISTRY (Zooplankton -> fish) Oceanic subsidies rationale (from GLOBEC) Detection of oceanic subsidies method (SIA, zooplankton) Relationships between herring SIA and WBEC in 2006 – 2007 WBEC  in time (during 2007, vs. SEA) Relationships between herring and other fishes using SIA

Why might oceanic subsidies be important ? Marine survival rate of PWS hatchery pink salmon was inversely correlated to their early marine stable carbon isotope value. Diagnostic value of carbon SIA– need zooplankton data too Marine survival rate was proportional to oceanic subsidies inferred from carbon SIA; 50% of marine survival rate explained (from Kline et al. 2008).

Low  13 C’ values reflect oceanic carbon in PWS in 2007 Bulk sample vs. single species analysis: same result PWS > -20 Significance of oceanic species in bays Bulk samples

2007 observations fit well into historical data Carbon SIA of Neocalanus cristatus (+ 1.0) from 1998 to 2004; from Kline et al (GLOBEC) Carbon SIA of northern lampfish, Stenobrachius leucopsarus (Myctophidae) from 1997 to 2005; from Kline in progress

Inter-annual differences in carbon sourcing: Fall 2006 and 2007 herring energy in relation to SIA Modeled whole body energy content (WBEC) of age-0 herring (Y-axis) to their  13 C’ values (x- axis) observed during November 2006 (limited samples) and November 2007 as a reflection of pre-winter condition. Data are presented as ‘convex hulls’ that show the range for each parameter by data group. Herring in 2007 were relatively less dependent on GOA then A number herring from St. Mathews had relatively greater GOA carbon (values more negative than -19) and these tended towards higher WBEC.

Spring 2007 herring energy in relation to SIA Whole body energy content (WBEC) of age-0 herring in relation to  13 C’ values observed during March to April 2007 among PWS herring nursery bays as a reflection of post-winter condition. A fortuitous second set of samples from April provided by Steve Moffitt (ADFG) lacked the lowest WBEC values observed in March. The  13 C’ values overlap in part with those observed in the fall of 2006 so are more negative than those observed later in 2007 (previous figure). The East arm of Whale Bay tended to have herring with higher energy content as well as lower  13 C’ values that may have resulted from feeding on euphausiids.

WBEC per fish -1 (2007 data; fall-winter not sequential) April attrition of small low WBEC herring vs. March

WBEC per fish -2 (2007 data vs. SEA data from AJ Paul)

Interaction with other fishes - 1 SIA of age 0 and 1 walleye pollock (Theragra chalcogramma) sampled incidentally with herring in March 2007 (sympatric over-winter dietary overlap). Data are not inconsistent with predation on herring by a few of the pollock (these have higher  15 N values). Lower  13 C’ values of herring are similar to the values observed in November 2006, which is consistent with little feeding during winter, the intervening period. Much higher  13 C’ values in pollock suggest feeding on PWS carbon during winter Pollock are thus able to maintain or increase energy content during the winter while herring cannot.

Interaction with other fishes - 2 Center of the  13 C’ value distribution near - 19 suggests that mostly PWS carbon was important for fall 2007 herring food webs, which contrasts with 2006 observations and some of those from the 1990’s There was far less overlap between herring and pollock in March compared to November (re. previous slide)

What is missing? We have only have observed poor year classes of herring. What do herring that become a strong year class look like? A strong year class for PWS is about 1 billion age-3 herring PWS herring recruitment history (from Funk 2008) Uncommon strong recruitments sustain fisheries; 1904 year class of Atlantic herring, from Hardy (data from Hjort)

Review Plankton energetics (Bay & system carrying capacity questions) Distinctive zooplankton communities (strong recruitment?) Plankton isotope gradients comparable to past Herring and other fish received little GOA subsidies in 2007 (strong recruitment?) Herring in March 2007 were worse off then during SEA Pollock find food in the winter whereas herring do not (strong recruitment?) Herring and other Fish larval distribution (more bays = strong recruitment?) What do herring destined to become a billion fish year class look like at age-0?