Page 1
REVIEWS
Hypotheses concerning the decline and poor recoveryof Pacific herring in Prince William Sound, Alaska
Walter H. Pearson • Richard B. Deriso •
Ralph A. Elston • Sharon E. Hook •
Keith R. Parker • Jack W. Anderson
Received: 1 October 2009 / Accepted: 10 May 2011 / Published online: 29 May 2011
� The Author(s) 2011. This article is published with open access at Springerlink.com
Abstract This paper updates previous reviews of
the 1993 stock decline of Pacific herring (Clupea
pallasi) in Prince William Sound, Alaska, and
focuses on hypotheses about subsequent poor recov-
ery. Recent age structured assessment modeling with
covariate analysis indicates that the population
dynamics of the sound’s herring are influenced by
oceanic factors, nutrition, and, most substantially,
hatchery releases of juvenile pink salmon. For the
1993 decline, poor nutrition remains the most
probable cause with disease a secondary response.
Concerning poor recovery, we examined 16 potential
factors and found three to be causal: oceanic factors,
poor nutrition, and hatchery releases of juvenile pink
salmon. Absences of strong year classes at both
Sitka and Prince William Sound after 1993 indicate
the action of large-scale ocean processes. Beyond
regional-scale environmental factors, two factors
specific to the sound influence the population
dynamics of herring and are likely impeding recov-
ery. First, pink salmon fry releases have increased to
about 600 million annually and may disrupt feeding
in young herring, which require adequate nutrition
for growth and overwintering survival. Juvenile pink
salmon and age-1 herring co-occur in nearshore
areas of bays in late spring and summer, and
available data on dietary overlap indicates potential
competition between the age-1 juvenile herring and
juvenile pink salmon. Field studies demonstrate that
juvenile herring reduce food intake substantially in
the presence of juvenile pink salmon. Second,
overwintering humpback whales may consume
potentially large amounts of adult herring, but
further studies must confirm to what extent whale
predation reduces herring biomass.
Keywords Pacific herring � Clupea pallasi �Fisheries collapse � Fisheries recovery �Prince William Sound � Alaska �Exxon Valdez oil spill
Jack W. Anderson—Deceased.
W. H. Pearson (&)
Peapod Research, 7335 Watermark Drive, Allendale,
MI 49401, USA
e-mail: [email protected]
R. B. Deriso
Inter-American Tropical Tuna Commission (IATTC),
Scripps Institution of Oceanography, La Jolla,
CA 92093-0203, USA
R. A. Elston
AquaTechnics, Inc., PO Box 687, Carlsborg,
WA 98324, USA
S. E. Hook � J. W. Anderson
Battelle Marine Sciences Laboratory, 159 West Sequim
Bay Road, Sequim, WA 98382, USA
K. R. Parker
Data Analysis Group, 5100 Cherry Creek Road,
Coverdale, CA 95425, USA
123
Rev Fish Biol Fisheries (2012) 22:95–135
DOI 10.1007/s11160-011-9225-7
Page 2
Introduction
Because of the commercial and ecological importance
of the Pacific herring (Clupea pallasi) population in
Prince William Sound (PWS) Alaska, there is concern
for the present and future states of PWS herring. In a
comprehensive review, Hay et al. (2001) found that
during the last century many herring stocks around the
world have collapsed. Intense fishing has often been
associated with such collapses, but habitat loss and
environmental change have also led to collapses of
herring and other species (Hilborn 1997). Recovery of
most collapsed herring stocks usually took a decade or
more (Hay et al. 2001). For example, after 15 years of
no fishing, the Georges Bank herring returned to stock
levels above those before the 1977 collapse. Some
stocks, however, have not returned to their former
abundance even after several decades. For example,
the Hokkaido-Sakhalin stock has remained extremely
depressed since its decline in the late 1950s. Most
collapsed stocks appear to recover, but the present
status of some stocks is ‘‘worrisome.’’ One stock
considered worrisome is that in PWS (Hay et al. 2001).
PWS herring biomass increased steadily through
the 1980s (Fig. 1). Funk (1993, 1994) used ASA (age
structured assessment) modeling to predict that the
1993 PWS herring spawning biomass would be a
record high of over 121,000 metric tons. However,
the predicted biomass failed to materialize (Brown
et al. 1994, 1996; Carls et al. 2002). Brown et al.
(1994) report that the observed PWS spawning
biomass was about 27,000 metric tons in spring
1993 or about 20% of expectation.
After the decline, the fishery was closed from 1994
through 1996. The PWS herring stock increased to
the point where modest harvests of 3,629 to 4,536
metric tons were permitted in 1997 and 1998. Then,
in 1999, only half the predicted PWS biomass was
found (Marty et al. 2003). In 1999, the PWS herring
fishery was closed and remains closed because of
insufficient biomass.
Before the early 1990s, the PWS herring popula-
tion and fishery were sustained by recruitment of
strong year classes about every 4 years (Funk and
Sandone 1990; Funk 1994; Williams and Quinn
2000a, b). Recovery of PWS herring would be greatly
aided by the recruitment of a highly successful year
class. However, no strong year class has emerged in
PWS since the 1988 year class recruited in 1991 and
1992 (Fig. 2).
The two aims of this paper are (1) to briefly update
the review of Pearson et al. (1999) concerning the
decline of PWS Pacific herring (Clupea pallasi) stock
in 1992–1993 (Table 1) and (2) to assess hypotheses
concerning the subsequent poor recovery of the stock.
For the first aim, the decline, we focus on information
that has become available since the late 1990s,
particularly work subsequent to Pearson et al. (1999),
which includes studies by Marty et al. (2003, 2004,
2007, 2010) on disease, a comprehensive disease
Calendar Year
1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004
Bio
mas
s (m
etri
c to
ns)
0
20000
40000
60000
80000
100000
120000
140000
160000
Kilo
met
ers
of S
paw
n/K
ilom
eter
-Day
s of
Spa
wn
0
50
100
150
200
250
300
350
400
Peak Aerial Biomass (metric tons) ASA Biomass (metric tons) Hydroacoustic Biomass (metric tons) Spawn Egg Deposition (metric tons)Kilometers of SpawnKilometer-Days of Spawn
Fig. 1 Herring biomass
indices for Prince William
Sound by calendar year.
Data from: ADFG
(unpublished data, See
Appendix), Ashe et al.
(2005), Donaldson et al.
(1993), Moffitt (2006)
96 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 3
review by Elston and Meyers (2009), and ASA
modeling focused on PWS herring (Quinn et al. 2001;
Deriso et al. 2007, 2008; Hulson et al. 2008). Deriso
et al. (2008) used ASA modeling with covariate
analysis to quantitatively assess 19 decline and
recovery hypotheses. ASA models have been well
described in the literature (Quinn and Deriso 1999;
Deriso et al. 2007, 2008) and will not be discussed in
detail here. An important aspect of these models is
that their statistical nature enables them to be used
appropriately to provide estimates of uncertainty and
to test hypotheses. Deriso et al. (2008) used 19
different time series of data generally for the period
from the early 1970s to 2005 as single and multi-
factor covariates in their ASA modeling. Modeling
results revealed the need to assess predation and
competition in greater detail and led to substantial
effort and analysis in those areas in this paper.
Identifying the hypotheses
We have identified hypotheses from literature
reviews and have distinguished the hypotheses about
the decline from those about poor recovery (Table 2).
Herring have supported important fisheries world-
wide for centuries (Hay et al. 2001; Stephenson
2001), and high variability in abundance has been
known for decades to be a common and inherent
feature of herring stocks (Hjort 1914; Smith 1988;
Rothschild 1986; Southward et al. 1988; Hay et al.
2001). Prominent among the various causes proposed
for herring stock fluctuations have been the influence
of ocean climate on recruitment processes and the
effects of intense fisheries (Hay et al. 2001; Stephen-
son 2001). Unlike other declines that involve recruit-
ment failure (Hilborn 1997), the PWS 1992–1993
decline was a one-time event with an acute increase
in adult mortality, not a recruitment failure (Pearson
et al. 1999; Quinn et al. 2001; Carls et al. 2002;
Deriso et al. 2008; Hulson et al. 2008). Hypotheses
concerning the decline must provide mechanisms that
increase mortality in all adult age classes between the
spring of 1992 and the spring of 1993. Mechanisms
governing poor recovery cannot be assumed to be the
same as those for the decline. Hypotheses for poor
recovery must provide mechanisms for (1) no
recruitment of strong year classes after the 1988 year
class recruited and (2) continued low herring
biomass.
Calendar Year
1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005
Num
ber
of A
ge-T
hree
Her
ring
(m
illio
ns o
f fi
sh)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
Prince Willliam Sound Sitka
Fig. 2 Number of age-3 herring recruits for Prince William Sound and Sitka, Alaska, estimated by ADFG ASA models by calendar
year (Dressel 2006, Moffitt 2006)
Rev Fish Biol Fisheries (2012) 22:95–135 97
123
Page 4
Hypotheses for the decline of PWS herring
The hypotheses concerning the decline and accep-
tance and rejection of certain hypotheses have
evolved substantially since 1993. The ASA models
of Quinn et al. (2001), Hulson et al. (2008), and
Deriso et al. (2008) all require an abrupt increase in
mortality in 1992–1993 to achieve reasonable model
fits. Any hypothesis for the decline must explain this
abrupt increase in adult mortality between 1992 and
1993. We briefly describe the evolution of the
hypotheses below and assess the two hypotheses,
oil exposure from the 1989 Exxon Valdez oil spill
(EVOS) and disease, about which some debate
continues.
At the time of the decline, the major hypotheses
concerning its cause(s) centered on two factors: oil
exposure and disease. After identifying viral hemor-
rhagic septicemia virus (VHSV) in Pacific herring,
Meyers et al. (1994) concluded that the role of VHSV
in the 1992–1993 decline could not be established
with the 1993 data. Brown et al. (1996) conjectured
that the decline was due to an epizootic, and Marty
et al. (1998) implicated VHSV disease as the cause of
the 1992–1993 decline.
Later, Pearson et al. (1999) reviewed 13 hypoth-
eses concerning the 1993 decline (Table 1). These 13
hypotheses fell into groups related to the EVOS,
harvest, and natural processes, especially disease.
Pearson et al. rejected hypotheses concerning oil-spill
effects based on the high population levels following
the spill, the lack of change from the expected age-
class structure, and the low level of oil exposure
following the spill. They also found no evidence to
support hypotheses about harvesting effects and
judged hypotheses concerning water temperature,
predation, and competition to have insufficient infor-
mation to accept or reject them. Condition, as weight
at specific length, is a measure of nutritional state of
fish (Winters and Wheeler 1994; Cardinale et al.
2003). The 1993 mortality was preceded by sharply
declining condition (weight at specific length) in
PWS herring from about 1986 (Fig. 3) and by lower
zooplankton abundance (Fig. 4). Pearson et al.
Table 1 Hypotheses assessed by Pearson et al. (1999)
ID Hypothesis
(i) Three hypotheses postulate that the 1989 oil spill caused the 1993 decline
H1 Acute toxic effects of oil exposure in 1989 caused the decline
H2 Toxic effects of oil lingering in PWS through 1992 or 1993 produced the decline
H2aLingering oil affected the herring directly
H2bLingering oil affected the herring indirectly by reducing their food supply
H3 Delayed effects of the 1989 oil spill induced a disease that caused the decline
(ii) Two hypotheses postulate that a harvesting effect caused the decline
H4 Overfishing led to the decline of the PWS herring stock
H5 Underfishing (1989 Fisheries Closure) allowed the herring biomass to increase to the point that density-dependent mechanisms
led to the decline
(iii) Four hypotheses postulate that a series or combination of natural factors caused the decline
H6 Poor nutrition alone or in conjunction with a disease caused by an opportunistic pathogen produced substantially lower herring
survival that in turn caused the decline
H6aRecently increasing herring biomass increased intraspecific competition for prey (a density dependent response)
H6bRecently increasing biomass of other fish increased interspecific competition for prey
H6cLarge-scale oceanic processes reduced food supply that in turn lowered the nutritional status of PWS herring
H7 An infectious disease caused the decline
H7aDisease alone caused the decline
H7aLow temperatures triggered a disease outbreak that in turn reduced survival
H8 Increased predation reduced the population level
H9 Natural chaotic population dynamics produced the decline, perhaps by a stochastic process in which small changes in the
physical or biological environment produced a large, abrupt change in the PWS herring population
98 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 5
Table 2 List of hypotheses concerning the decline of PWS herring and subsequent poor recovery
Decline hypotheses Narrative
A
1 Population dynamics Natural chaotic population dynamics produced the decline, perhaps by a stochastic
process in which small changes in the physical or biological environment produced a
large, abrupt change in the PWS herring population
2 Harvest—overfishing Overfishing of PWS herring led to the decline
3 Harvest—fisheries closure The 1989 fishing closure led to high biomasses that exceeded the carrying capacity of
PWS for herring
4 Spawning habitat Loss Loss of spawning habitat led to the decline
5 Disease and I. hoferi Mortality from disease from parasite I. hoferi caused the decline
6 Disease and VHSV Mortality from disease from pathogen VHSV caused the decline
7 Disease and EOK pound fishery Confinement in net pens for EOK fishery induced VHS and disease mortality led to the
decline
8 Oil Exposure and VHSV Oil exposure induced VHS and disease mortality led to the decline
9 Oil Exposure Alone Oil exposure alone directly caused mortality and the decline
10Competition or predation by pink salmonIncreased releases of hatchery pink salmon fry increased predation and/or competition
with adult herring and led to the decline
11Predation by sea lions Mortality from predation by sea lions led to the decline
12Poor nutrition Poor nutritional state of adult herring caused the decline
13Ocean factors Changes in ocean factors caused the decline
Recovery hypotheses
B
1 Population dynamics Natural chaotic population dynamics are keeping PWS herring at a low level, perhaps
by a stochastic process in which small changes in the physical or biological
environment moved the PWS herring population to a new set point
2 Harvest—overfishing Overfishing of PWS herring impairs recruitment and recovery
3 Harvesting—fisheries closure Fishing closures impair recruitment and recovery
4 Spawning habitat loss Loss of spawning habitat impairs recruitment and recovery
5 Behavioral conservatism Behavioral conservatism impairs adjustment to environmental changes and reduces
recruitment and recovery
6 Disease and VHSV Disease mortality from VHSV impairs recruitment and recovery
7 Disease and I. hoferi Disease mortality from I. hoferi impairs recruitment and recovery
8 Disease and EOK Pound Fishery Disease mortality from VHSV induced by net pen confinement in EOK fishery impairs
recruitment and recovery
9 Oil Exposure and VHSV Exposure to residual oil in PWS induces VHS, which impairs recruitment and recovery
11 Whales as predator in predator pit Predation by humpback whales has PWS herring in a predator pit
12 Predation by adult coho salmon Mortality from predation by adult coho salmon impairs recruitment and recovery
13 Predation by adult pink salmon Mortality from predation by adult pink salmon impairs recruitment and recovery
14 Predation by juvenile pink salmon Mortality from predation by pink salmon fry released from hatcheries impairs
recruitment and recovery
15 Poor nutrition Poor nutritional conditions in PWS larvae and juveniles impair recruitment and
recovery
16 Ocean factors Changes in ocean factors have changed the recruitment processes and their outcomes
17 Competition with juvenile
pink salmon
Competition with pink salmon fry released from hatcheries produced food limitation
that impairs growth and survival of juvenile herring and impairs recruitment and
recovery
Rev Fish Biol Fisheries (2012) 22:95–135 99
123
Page 6
concluded that poor nutrition, either alone or in
combination with disease or other natural factors,
probably caused the decline.
Since the earlier review (Pearson et al. 1999), new
information is available about the decline relative to
oil exposure, harvest, disease, poor nutrition, and
oceanic factors. Hypotheses relating the decline to
oil-spill effects, harvest effects, spawning-habitat
loss, or the spawn-on-kelp fishery continue to be
unsupported by the available evidence (Table 3). The
declining herring condition and low zooplankton
abundance (Figs. 3 and 4) described in Pearson et al.
(1999) still support the conclusion that poor nutrition
was a major factor in the 1993 decline. The results of
the ASA modeling of Deriso et al. (2008) offer
evidence for both oceanic factors (winter sea-surface
temperature) and poor nutrition in the 1993 decline
(Table 4). Since the 1999 review, other investigators
also report that the 1993 decline was most likely
derived from poor nutrition with disease as a
secondary factor (Carls et al. 2002; Marty et al.
2003; Hulson et al. 2008; Rice and Carls 2007; Elston
and Meyers 2009). Poor nutrition remains the best
explanation of the 1992–1993 decline, although some
debate remains about potential influence from the oil
spill and about the role of disease agents.
Oil exposure
Recent evidence continues to support the earlier
conclusions of Pearson et al. (1999) that oil exposure
during or after the EVOS did not play a role in the
1992–1993 decline. Boehm et al. (2007) synthesized
the available information on the water-column con-
centrations of total polycyclic aromatic hydrocarbons
(TPAH) with emphasis on nearshore waters in the
spill path and concluded that ‘‘water column concen-
trations of TPAH resulting for the spill returned to
Calendar Year
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006
Wei
ght a
t Spe
cifi
c L
engt
h (g
)
85
90
95
100
105
110
115
120
125
Prince William SoundSitka
Fig. 3 Herring condition as
weight at specific length in
spring for Prince William
Sound and Sitka by
calendar year. Weight at
specific length was
calculated following
Winters and Wheeler
(1994). Data from: ADFG
(unpublished data, See
Appendix), Biggs et al.
(1992), Brady et al. (1987,
1988, 1990, 1991a, b),
Donaldson et al. (1992),
Dressel (2006), Sandone
(1988)
Calendar Year
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998
Plan
kton
Set
tled
Vol
ume
(mL
/m3)
0
1
2
3
4
5
6
7
Fig. 4 Settled volume of spring zooplankton in Prince
William Sound. Data from: Cooney et al. (2001)
100 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 7
Ta
ble
3S
um
mar
yo
fev
iden
cefo
rh
yp
oth
eses
con
cern
ing
the
dec
lin
eo
fP
WS
her
rin
gan
dsu
bse
qu
ent
po
or
reco
ver
y
Dec
line
Hypoth
esis
Corr
elat
ion
or
asso
ciat
ion?
Use
din
AS
A
model
(Der
iso
etal
.2008)?
Sig
nifi
cant
single
fact
or
in
random
izat
ion
test
s
(Der
iso
etal
.2008)?
Fac
tor
inm
ult
ifac
tor
model
(Der
iso
etal
.
2008)?
Mec
han
ism
bio
logic
ally
pla
usi
ble
?
Support
ing
evid
ence
Contr
ary
evid
ence
Role
indec
line?
1P
opula
tion
Dynam
ics
NA
NN
AN
AN
ot
test
able
,
subsu
med
under
oce
anre
gim
e
NA
NA
Subsu
med
under
oce
anfa
ctors
2H
arves
t—
over
fish
ing
Wea
k,
hig
hhar
ves
tsfo
r
thre
eyea
rspre
cedin
g
the
dec
line
NN
AN
AY
Som
eover
fish
ing
occ
urr
edin
earl
y
1990s
Am
ount
of
over
fish
ing
not
suffi
cien
tto
acco
unt
for
dec
line
Not
likel
y
3H
arves
t—
fish
erie
s
closu
re
Wea
k,
Som
e
under
fish
ing
inth
e
1980s
and
no
har
ves
t
in1989
NN
AN
AY
Som
eunder
har
ves
ting
occ
urr
edin
1980s
Tim
ing
of
under
har
ves
ting
and
dec
line
do
not
alig
n.
Am
ount
insu
ffici
ent
Not
likel
y
4S
paw
nin
g
hab
itat
loss
Wea
k,
Spaw
nin
ghab
itat
ingood
condit
ion
and
not
lim
itin
g
N,
Indir
ectl
y
see
H9
NA
NA
YS
om
esp
awnin
gst
opped
inso
me
PW
Sar
eas
Expan
sion
and
contr
acti
on
of
spaw
nin
glo
cati
ons
asex
pec
ted
wit
hch
anges
inst
ock
level
s
Not
likel
y
5D
isea
sean
d
I.hofe
ri
Wea
k,
dis
ease
was
obse
rved
,pre
val
ence
low
and
not
corr
elat
ed
wit
hbio
mas
s
Y,
age
3an
d
old
erplu
s
age
5an
d
old
er
Y,
when
addit
ional
mort
alit
yfa
ctor
in
model
Y,
intw
ofa
ctor
model
when
addit
ional
mort
alit
yfa
ctor
inm
odel
YI.
h.c
anle
adto
mort
alit
y
inold
erfi
sh
Loss
of
old
erag
ecl
asse
sw
asnot
evid
ent
indec
line
soth
at
obse
rved
age
stru
cture
does
not
support
role
indec
line.
Afa
ctor
wit
hm
ort
alit
ybut
not
wit
hout
mort
alit
yfa
ctor
sugges
tsno
role
indec
line.
Pat
tern
in
pre
val
ence
sco
ntr
ary
to
epiz
ooti
c
Not
likel
y
6D
isea
sean
d
VH
SV
Wea
k,
Dis
ease
was
obse
rved
,pre
val
ence
low
and
not
corr
elat
ed
wit
hbio
mas
s.N
o
esti
mat
eof
VH
SV
pre
val
ence
for
1993
and
bef
ore
Y,
age
3an
d
old
er
N,
not
for
age
3an
d
old
erfi
sh
N,
not
for
age
3
and
old
erfi
sh
YV
HS
dis
ease
can
occ
ur
and
lead
tom
ort
alit
y
when
fish
are
stre
ssed
VH
SV
affe
cts
pri
mar
ily
young
fish
.F
ish
that
surv
ive
vir
al
exposu
reat
earl
yag
eac
quir
e
imm
unit
y.
Mort
alit
yin
1992/1
993
was
not
pri
mar
ily
on
young
fish
but
even
among
age
clas
ses
3to
9?
.O
bse
rved
age
stru
cture
not
consi
sten
tw
ith
a
role
indec
line.
Not
asi
gnifi
cant
fact
or
inA
SA
model
of
Der
iso
etal
.(2
008)
Not
aca
use
of
the
dec
line.
Pro
bab
ly
seco
ndar
y
resp
onse
afte
r
nutr
itio
nal
or
oth
erst
ress
Rev Fish Biol Fisheries (2012) 22:95–135 101
123
Page 8
Ta
ble
3co
nti
nu
ed
Dec
line
Hypoth
esis
Corr
elat
ion
or
asso
ciat
ion?
Use
din
AS
A
model
(Der
iso
etal
.2008)?
Sig
nifi
cant
single
fact
or
in
random
izat
ion
test
s
(Der
iso
etal
.2008)?
Fac
tor
inm
ult
ifac
tor
model
(Der
iso
etal
.
2008
)?
Mec
han
ism
bio
logic
ally
pla
usi
ble
?
Support
ing
evid
ence
Contr
ary
evid
ence
Role
indec
line?
7D
isea
sean
d
EO
Kpound
fish
ery
Moder
ate,
EO
Kpound
fish
ery
incr
ease
dto
pea
kin
1992
and
wit
h
hig
hhar
ves
tin
1993
Y,
Ages
3an
d
old
er,
Ages
3an
d4
NN
YP
oundin
gdem
onst
rate
d
toin
crea
sepre
val
ence
of
VH
SV
and
dis
ease
VH
SV
mort
alit
ym
ainly
on
younger
fish
.M
ort
alit
yin
1992/1
993
was
even
among
age
clas
ses
3to
9?
.O
bse
rved
age
stru
cture
not
consi
sten
tw
ith
a
role
indec
line
Not
likel
y
8O
ilex
posu
re
and
VH
SV
Wea
k,
VH
SV
obse
rved
in1993.
No
signifi
cant
corr
elat
ion
bet
wee
n
PW
Sher
ring
bio
mas
s
and
PA
Hs
innea
rshore
wat
ers
from
shore
line
muss
els
Y,
Age
3an
d
old
er
NN
YO
ne
exper
imen
tal
study
sugges
tsoil
exposu
re
may
induce
VH
S
dis
ease
Four
subse
quen
tex
per
imen
tal
studie
sfa
ilto
show
VH
SV
induce
dby
oil
.L
abora
tory
exper
imen
tsfa
iled
to
dem
onst
rate
ali
nk
bet
wee
noil
exposu
rean
dV
HS
V.
Even
if
link
exis
ts,
oil
exposu
rew
asnot
hig
hen
ough
toin
duce
effe
cts
Not
likel
y
9O
ilex
posu
re
alone
Wea
k,
no
signifi
cant
corr
elat
ion
bet
wee
n
PW
Sher
ring
bio
mas
s
and
PA
Hs
innea
rshore
wat
ers
from
shore
line
muss
els
Y,
age
3an
d
old
er
NN
May
be
Thorn
ean
dT
hom
as
(2008
)ar
gue
that
dec
line
star
ted
in1989
and
was
due
inpar
tto
adult
mort
alit
yfr
om
exposu
reto
wat
er
surf
ace
duri
ng
nig
htl
y
gulp
ing
Wat
erco
lum
nco
nce
ntr
atio
ns
not
suffi
cien
tin
1991
and
afte
rwar
d
toca
use
adult
mort
alit
y.
Exposu
reth
rough
gulp
ing
too
inte
rmit
tent
inti
me
and
spac
eto
induce
mort
alit
y.
Mec
han
ism
not
dem
onst
rate
d.
Avoid
ance
of
oil
slic
ks
and
shee
ns
would
reduce
exposu
re.
Tim
ing
of
exposu
rean
ddec
line
not
alig
ned
Not
likel
y
9C
om
pet
itio
n
or
pre
dat
ion
by
pin
k
salm
on
Str
ong,
Pin
ksa
lmon
fry
rele
ases
and
asso
ciat
ed
adult
retu
rns
incr
ease
d
dra
mat
ical
lybef
ore
the
dec
line
NN
AN
AN
Dra
mat
icin
crea
sein
adult
retu
rns
of
hat
cher
ypin
ksa
lmon
pre
cedin
gth
edec
line
No
mec
han
ism
toin
duce
adult
her
ring
mort
alit
y.
Pin
ksa
lmon
inte
ract
ions
asi
gnifi
cant
var
iable
inD
eris
oet
al.
(2008
)
AS
Am
odel
for
recr
uit
men
t
Not
likel
yfo
r
dec
line;
Lik
ely
for
poor
reco
ver
y
102 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 9
Ta
ble
3co
nti
nu
ed
Dec
line
Hypoth
esis
Corr
elat
ion
or
asso
ciat
ion?
Use
din
AS
A
model
(Der
iso
etal
.2008)?
Sig
nifi
cant
single
fact
or
in
random
izat
ion
test
s
(Der
iso
etal
.2008)?
Fac
tor
inm
ult
ifac
tor
model
(Der
iso
etal
.
2008)?
Mec
han
ism
bio
logic
ally
pla
usi
ble
?
Support
ing
evid
ence
Contr
ary
evid
ence
Role
indec
line?
10
Pre
dat
ion
by
sea
lions
Wea
k,
Sea
lion
abundan
cean
d
dis
trib
uti
on
appea
rsto
posi
tivel
yco
rrel
ate
wit
hher
ring
bio
mas
s
Y,
Ages
3an
d
old
er
N,
Eff
ect
inw
rong
dir
ecti
on
NY
Sea
lions
pre
yon
subad
ult
and
adult
her
ring
inP
WS
Sea
lion
dis
trib
uti
on
appea
rsto
be
dri
ven
by
her
ring
bio
mas
s
rath
erth
anvic
ever
sa.
Sea
lion
abundan
ceto
osm
all
tobe
signifi
cant
sourc
eof
mort
alit
y
by
pre
dat
ion.
Sea
lion
dec
line
inP
WS
par
alle
lsA
lask
a-w
ide
dec
lines
star
ting
in1980s
Not
likel
y
11
Poor
nutr
itio
nS
trong,
wei
ght
atle
ngth
pro
gre
ssiv
ely
dec
reas
edbef
ore
the
dec
line
and
reco
ver
ed
afte
rth
edec
line
YY
,sp
ring
condit
ion
Y,
spri
ng
condit
ion
YD
ecre
ases
inw
eight
and
condit
ion
pre
cede
the
dec
line.
Condit
ion
in
PW
Slo
wer
than
at
Sit
ka
Sourc
eof
mort
alit
yst
ill
not
dir
ectl
ydem
onst
rate
d
Lik
ely
12
Oce
anfa
ctors
Moder
ate,
regim
esh
ift
occ
urr
ed,
Bio
mas
s
incr
ease
inher
ring
afte
r1976
asso
ciat
ed
wit
hre
gim
esh
ift.
Dec
reas
ing
zoopla
nkto
n
abundan
ceaf
ter
late
1980s
YY
,G
OA
win
ter
wat
erte
mper
ature
wit
hout
Addit
ional
Mort
alit
yF
acto
r
Y,
GO
Aw
inte
r
wat
erte
mper
ature
wit
hout
addit
ional
mort
alit
yfa
ctor
YP
aral
lel
chan
ges
in
bio
mas
san
dyea
r-cl
ass
stre
ngth
atP
WS
and
sitk
ain
dic
ate
resp
onse
togulf
-wid
e
envir
onm
enta
lch
anges
Sourc
eof
mort
alit
yst
ill
not
dir
ectl
ydem
onst
rate
d
Lik
ely
Rec
over
y
Hypoth
esis
Corr
elat
ion
or
asso
ciat
ion?
Use
din
AS
A
Model
(Der
iso
etal
.2008
)?
Sig
nifi
cant
single
fact
or
in
random
izat
ion
test
s
(Der
iso
etal
.2008)?
Fac
tor
inm
ult
ifac
tor
model
(Der
iso
etal
.
2008
)?
Bio
logic
ally
pla
usi
ble
mec
han
ism
?
Support
ing
evid
ence
Contr
adic
tory
evid
ence
Role
inpoor
reco
ver
y?
1P
opula
tion
dynam
ics
NA
NN
AN
AN
AN
AN
AL
ack
of
stro
ng
yea
rcl
asse
s
atP
WS
and
Sit
ka
indic
ate
dom
ain
shif
tat
regio
nal
scal
e;su
bsu
med
under
oce
anre
gim
e
Rev Fish Biol Fisheries (2012) 22:95–135 103
123
Page 10
Ta
ble
3co
nti
nu
ed
Rec
over
y
Hypoth
esis
Corr
elat
ion
or
asso
ciat
ion?
Use
din
AS
A
Model
(Der
iso
etal
.2008)?
Sig
nifi
cant
single
fact
or
in
random
izat
ion
test
s
(Der
iso
etal
.2008
)?
Fac
tor
inm
ult
ifac
tor
model
(Der
iso
etal
.
2008
)?
Bio
logic
ally
pla
usi
ble
mec
han
ism
?
Support
ing
evid
ence
Contr
adic
tory
evid
ence
Role
inpoor
reco
ver
y?
2H
arves
t—
over
fish
ing
No,
litt
leor
no
fish
ing
afte
r
1993
NN
AN
AN
Som
eover
fish
ing
occ
urr
edin
earl
y1990s
Sit
ka
asw
ell
asP
WS
show
edla
ckof
stro
ng
yea
rcl
asse
saf
ter
earl
y1990s
but
Sit
ka
has
har
ves
t
Not
likel
y
3H
arves
ting—
fish
erie
scl
osu
re
No,
litt
leor
no
fish
ing
afte
r
1993
NN
AN
AN
Som
eunder
har
ves
ting
occ
urr
edin
1980s
Bec
ause
of
non-a
ligned
tim
ing,
can
not
be
afa
ctor
inla
ckof
stro
ng
yea
r
clas
ses
Not
likel
y
4S
paw
nin
ghab
itat
loss
No
NN
AN
AY
Som
esp
awnin
gst
opped
inso
me
oil
edar
eas
post
dec
line
Spaw
nin
ghab
itat
ingood
condit
ion
and
not
lim
itin
g;
Spaw
nin
gch
anges
not
rela
ted
tooil
ed/u
noil
ed
condit
ions;
chan
ges
insp
awnin
g
loca
tion
isex
pec
ted
wit
hch
anges
inst
ock
level
s
Not
likel
y
5B
ehav
iora
l
conse
rvat
ism
No in
form
atio
n
NN
AN
AM
aybe
Lit
tle
info
rmat
ion
Lit
tle
info
rmat
ion.
Ifan
y
conse
rvat
ism
occ
urr
ed,
too
short
-
lived
toex
pla
inti
me
cours
eof
lack
of
stro
ng
yea
rcl
asse
sover
10
yea
rs
Not
likel
y
6D
isea
sean
d
VH
SV
Yan
dN
Y,
ages
3
and
4
NY
YV
HS
dis
ease
affe
cts
younger
fish
.V
HS
can
occ
ur
and
lead
tom
ort
alit
yw
hen
fish
are
stre
ssed
Not
aca
use
of
the
lack
of
stro
ng
yea
r
clas
ses.
Not
support
edby
dis
ease
anal
ysi
sof
Els
ton
and
Mey
ers
(2009).
Som
ein
fluen
cebut
not
a
maj
or
sourc
eof
mort
alit
yin
AS
A
model
ing
of
Huls
on
etal
.(2
008).
VH
Sef
fect
on
juven
ile
her
ring
mort
alit
yen
ters
AS
Am
odel
of
Der
iso
etal
.(2
008)
afte
rpin
k
salm
on
inte
ract
ions,
poor
nutr
itio
n,
and
GO
Aw
inte
rte
mper
ature
in
mult
ifac
tor
AS
Am
odel
and
has
infl
uen
cein
wro
ng
dir
ecti
on
Not
likel
y
7D
isea
sean
d
I.hofe
ri
NY
,ag
es3
and
4
NN
NI.
hofe
rica
nle
adto
mort
alit
y
inold
erfi
sh.
Som
e
dis
rupti
on
of
age
clas
s
stru
cture
inla
te1990s.
infl
uen
tial
inA
SA
model
s
of
Huls
on
etal
.(2
008)
and
Mar
tyet
al.
(2010)
Loss
of
old
erag
ecl
asse
snot
consi
sten
tpost
dec
line.
Age
stru
cture
asex
pec
ted
in2000s.
Loss
of
old
erfi
shnot
adir
ect
effe
cton
recr
uit
men
t
Lik
ely
anin
fluen
ceon
old
erfi
shbut
not
aca
use
of
poor
recr
uit
men
t.
May
be
seco
ndar
y
resp
onse
topoor
nutr
itio
nan
din
fluen
ce
on
adult
bio
mas
s
104 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 11
Ta
ble
3co
nti
nu
ed
Rec
over
y
Hy
poth
esis
Co
rrel
atio
no
r
asso
ciat
ion?
Use
din
AS
A
Mo
del
(Der
iso
etal
.2
00
8)?
Sig
nifi
can
tsi
ng
le
fact
or
in
ran
do
miz
atio
nte
sts
(Der
iso
etal
.2
00
8)?
Fac
tor
in
mult
ifac
tor
mod
el(D
eris
o
etal
.2
00
8)?
Bio
log
ical
ly
pla
usi
ble
mec
han
ism
?
Support
ing
evid
ence
Contr
adic
tory
evid
ence
Role
inpoor
reco
ver
y?
8D
isea
sean
dE
OK
po
un
dfi
sher
y
No
EO
Kfi
sher
yh
as
occ
urr
edsi
nce
the
dec
lin
e
Y,
ages
3
and
4
NN
NP
ou
nd
ing
can
incr
ease
pre
val
ence
of
VH
SV
Wit
hP
WS
fish
erie
scl
ose
d,
the
EO
Kp
ou
nd
fish
ery
can
no
t
adv
erse
lyaf
fect
reco
ver
y
No
tL
ikel
y
9O
ilex
po
sure
and
VH
SV
No
,P
AH
sat
bac
kg
rou
nd
afte
r
mid
-199
0s
Y,
Ag
es3
and
4
NN
NS
eed
ecli
ne
See
dec
lin
e.V
HS
Vn
ot
lin
ked
to
oil
exp
osu
re.
Th
ere
isn
ot
suffi
cien
to
ilex
po
sure
No
tli
kel
y
10
Wh
ales
as
pre
dat
or
in
pre
dat
or
pit
Hu
mp
bac
kw
hal
es
co-o
ccu
rw
ith
her
rin
go
nh
erri
ng
ov
erw
inte
rin
g
gro
un
ds.
Whal
es
targ
etsc
ho
oli
ng
fora
ge
fish
NN
AN
AY
Ov
erw
inte
rin
gh
um
pbac
k
wh
ales
ob
serv
edin
her
rin
go
ver
win
teri
ng
area
inP
WS
.W
hal
e
abu
nd
ance
incr
easi
ng
.
Ov
erw
inte
rin
gd
iet
no
w
app
ears
tob
ep
rim
aril
y
her
rin
g
Rec
ent
pre
lim
inar
yd
ata
is
sup
po
rtiv
eb
ut
esti
mat
eso
f
con
sum
pti
on
var
yb
road
ly.
Mo
re
stu
dy
on
ov
erw
inte
rin
gw
hal
e
abu
nd
ance
and
die
tis
nee
ded
tore
fin
eco
nsu
mp
tio
nes
tim
ates
Lik
ely
asco
ntr
ibu
tin
g
fact
or
red
uci
ng
adu
lt
her
rin
gb
iom
ass.
Infl
uen
ceo
n
recr
uit
men
tu
nli
kel
y
11
Pre
dat
ion
by
adu
lt
coho
salm
on
Wea
k,
adu
ltco
ho
retu
rns
hav
e
incr
ease
din
rece
nt
yea
rs
Y,
ages
1
and
2
NN
YIn
crea
ses
inco
ho
salm
on
inre
cen
ty
ears
may
hav
ein
crea
sed
exte
nt
of
pre
dat
ion
on
yo
un
g
her
rin
g
No
ne
avai
lab
leN
ot
lik
ely
12
Pre
dat
ion
by
adu
lt
pin
ksa
lmo
n
Str
on
g,
Pin
ksa
lmo
n
retu
rnin
gad
ult
s
incr
ease
d
dra
mat
ical
lyb
efo
re
the
dec
lin
ean
d
hav
ere
mai
ned
hig
h
afte
rd
ecli
ne
Y,
Ag
es0
,1
,
and
1
Y,
Pin
kS
alm
on
Ret
urn
ing
Ad
ult
s
NA
NA
SA
mo
del
fin
ds
pin
k
salm
on
adu
lts
a
sig
nifi
can
tsi
ng
lefa
cto
r
No
mec
han
ism
for
adu
ltp
ink
salm
on
toaf
fect
juv
enil
e
her
rin
g.
Ad
ult
pin
ksa
lmo
nd
o
no
tfe
edw
hen
retu
rnin
gto
PW
S.
Eff
ect
isth
rou
gh
pin
ksa
lmo
nfr
y
No
tL
ikel
y
13
Pre
dat
ion
by
juv
enil
ep
ink
salm
on
Str
on
g,
Pin
k
salm
on
fry
rele
ases
incr
ease
d
dra
mat
ical
lyb
efo
re
the
dec
lin
ean
d
hav
ere
mai
ned
hig
h
afte
rd
ecli
ne
Y,
ages
0,
1,
and
2
Y,
Pin
kS
alm
on
Fry
Y,
Pin
kS
alm
on
Fry
YA
SA
mo
del
ing
fin
ds
juv
enil
ep
ink
salm
on
sig
nifi
can
tin
sin
gle
fact
or
and
mu
ltif
acto
rm
od
els.
Pre
dat
ion
by
juv
enil
ep
ink
salm
on
on
age-
0h
erri
ng
inJu
lyan
dA
ug
ust
lik
ely
No
tem
pir
ical
lyd
emo
nst
rate
dP
oss
ibly
.P
red
atio
nb
y
juv
enil
ep
ink
salm
on
and
juv
enil
eh
erri
ng
lik
ely
,b
ut
exte
nt
of
pre
dat
ion
isu
ncl
ear
Rev Fish Biol Fisheries (2012) 22:95–135 105
123
Page 12
Ta
ble
3co
nti
nu
ed
Rec
ov
ery
Hy
po
thes
isC
orr
elat
ion
or
asso
ciat
ion?
Use
din
AS
A
Mo
del
(Der
iso
etal
.2
00
8)?
Sig
nifi
can
tsi
ng
le
fact
or
in
ran
do
miz
atio
nte
sts
(Der
iso
etal
.2
00
8)?
Fac
tor
inm
ult
ifac
tor
mo
del
(Der
iso
etal
.
20
08
)?
Bio
log
ical
ly
pla
usi
ble
mec
han
ism
?
Su
pp
ort
ing
evid
ence
Co
ntr
adic
tory
evid
ence
Ro
lein
po
or
reco
ver
y?
14
Po
or
nu
trit
ion
Mo
der
ate,
po
or
nu
trit
ion
inP
WS
in1
99
0.
Rec
over
edin
Sit
ka
bu
tn
ot
PW
S
Y,
ages
3
and
old
er
Y,
spri
ng
con
dit
ion
wit
ho
ut
add
itio
nal
Mo
rtal
ity
Fac
tor
Y,
spri
ng
con
dit
ion
wit
ho
ut
add
itio
nal
mo
rtal
ity
fact
or
YC
on
tinu
edlo
wzo
op
lan
kto
n
abu
nd
ance
inP
WS
.
Co
nd
itio
nin
PW
Slo
wer
than
atS
itk
a.N
utr
itio
nfo
r
over
win
teri
ng
age
0an
d
age
1h
erri
ng
isk
eyto
yea
rcl
ass
stre
ng
th
No
tem
pir
ical
ly
dem
onst
rate
d
Lik
ely
inas
soci
atio
nw
ith
oce
anfa
cto
rsan
dp
ink
salm
on
fry
rele
ases
15
Oce
anF
acto
rsS
tro
ng
,n
atu
ral
log
of
nu
mb
ers
and
bio
mas
so
fag
e4
her
rin
gsi
gn
ifica
nt
qu
adra
tic
rela
tio
nsh
ipw
ith
GO
Aw
inte
r
tem
per
atu
re
Y,
Ag
e3
and
old
er
Y,
GO
AW
inte
r
Wat
er
Tem
per
ature
Y,
GO
AW
inte
r
Wat
erT
emp
erat
ure
YC
han
ges
atP
WS
and
Sit
ka
ind
icat
ere
spo
nse
toG
ulf
-
wid
een
vir
on
men
tal
chan
ges
.N
ost
ron
gy
ear
clas
ses
atei
ther
PW
So
r
Sit
ka
afte
r1
99
3
No
tem
pir
ical
ly
dem
onst
rate
d
Lik
ely
inas
soci
atio
nw
ith
oce
anfa
cto
rsan
dp
ink
salm
on
fry
rele
ases
;L
ikel
y
for
lack
of
stro
ng
yea
r
clas
ses
16
Co
mp
etit
ion
Str
on
g,
pin
ksa
lmo
n
fry
rele
ases
incr
ease
d
dra
mat
ical
ly
bef
ore
the
dec
lin
e
and
hav
ere
mai
ned
hig
haf
ter
dec
lin
e
Y,
ages
0,
1,an
d
2
Y,
Pin
kS
alm
on
Y,
Pin
kS
alm
on
YA
SA
mo
del
ing
fin
ds
juv
enil
ep
ink
salm
on
sig
nifi
can
tin
sin
gle
fact
or
and
mu
ltif
acto
rm
od
els.
Juv
enil
ep
ink
salm
on
and
age
1h
erri
ng
ov
erla
pin
spac
ean
dti
me
and
shar
e
pre
y.
Fie
ldo
bse
rvat
ion
s
inP
WS
sho
wsi
gn
ifica
nt
dis
rup
tio
no
fh
erri
ng
feed
ing
wh
enju
ven
ile
pin
ksa
lmo
nco
-occ
ur
No
tem
pir
ical
ly
dem
onst
rate
d
Lik
ely
inas
soci
atio
nw
ith
oce
anfa
cto
rsan
dp
oo
r
nu
trit
ion;
Mo
reli
kel
yan
d
mo
resu
bst
anti
alth
ano
cean
fact
ors
106 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 13
Ta
ble
4S
um
mar
yo
fo
utc
om
esfr
om
AS
Am
od
elin
go
fD
eris
oet
al.
(20
08
)so
rtin
gb
yra
nk
inra
nd
om
izat
ion
test
wit
ou
tan
add
itio
nal
mo
rtal
ity
fact
or.
NS
ind
icat
esa
no
n-
sig
nifi
can
tfa
cto
r
IDfr
om
Der
iso
etal
.(2
00
8)
Fac
tor
Fac
tor
affe
cts
ages
Ran
kb
ased
on
Ch
i-S
qu
are
Ran
kb
ased
on
ran
do
miz
atio
nte
st
Ran
kin
mu
ltif
acto
r
mo
del
w/m
ort
alit
yfa
cto
rw
/om
ort
alit
yfa
cto
r
w/p
arti
al
recr
uit
men
t
w/o
par
tial
recr
uit
men
t
w/p
arti
al
recr
uit
men
t
w/o
par
tial
recr
uit
men
t
w/m
ort
alit
y
fact
or
w/o
mo
rtal
ity
fact
or
w/m
ort
alit
y
fact
or
w/o
mo
rtal
ity
fact
or
15
PW
Sp
ink
salm
on
hat
cher
yfr
y
rele
ases
0an
d1
13
13
11
.51
and
21
,2
,an
d3
16
PW
Sp
ink
salm
on
hat
cher
yad
ult
retu
rns
0,
1,
and
22
42
42
1.5
NA
NA
2G
OA
win
ter
wat
erte
mp
erat
ure
3an
do
lder
NS
NS
3N
SN
S3
NO
2an
d3
1P
oo
rn
utr
itio
n3
and
old
erN
SN
S4
NS
NS
4N
O3
8V
HS
V-F
SR
sev
erit
yin
dex
3an
d4
NS
NS
6N
SN
S5
NO
4
9Ic
hth
yop
ho
nu
sh
ofe
riin
fect
ion
rate
All
36
NS
53
NS
2N
A
3S
pri
ng
–su
mm
erP
DO
3an
do
lder
NS
1N
S1
NS
NS
NA
NA
4F
all–
win
ter
PD
O3
and
old
erN
S2
52
NS
NS
NA
NA
5S
pri
ng
–su
mm
erN
OI
3an
do
lder
NS
NS
NS
NS
NS
NS
NA
NA
6F
all–
win
ter
NO
I3
and
old
erN
S5
NS
6N
SN
SN
AN
A
7V
HS
V-F
SR
sev
erit
yin
dex
3an
do
lder
NS
NS
NS
NS
NS
NS
NA
NA
10
Ich
thyo
ph
on
us
ho
feri
infe
ctio
nra
te5
and
old
erN
SN
SN
SN
SN
SN
SN
AN
A
11
Oil
exp
osu
rein
dex
,P
AH
in
nea
rsh
ore
mu
ssel
s
3an
do
lder
NS
NS
NS
NS
NS
NS
NA
NA
12
Oil
exp
osu
rein
dex
,P
AH
in
nea
rsh
ore
mu
ssel
s
Ag
e0
NS
NS
NS
NS
NS
NS
NA
NA
13
Sp
awn
-on
-kel
pcl
ose
dp
ou
nd
fish
ery
3an
do
lder
NS
NS
NS
NS
NS
NS
NA
NA
14
Sp
awn
-on
-kel
pcl
ose
dp
ou
nd
fish
ery
3an
d4
NS
NS
NS
NS
NS
NS
NA
NA
17
PW
Sto
tal
pin
ksa
lmo
nre
turn
s(w
ild
and
hat
cher
y)
0,
1,
and
2N
SN
SN
SN
SN
SN
SN
AN
A
18
PW
Sco
ho
salm
on
adu
ltre
turn
sp
lus
recr
eati
on
alca
tch
1an
d2
NS
NS
NS
NS
NS
NS
NA
NA
19
Sea
lio
nab
un
dan
ce
(no
tin
clu
din
gp
up
s)
3an
do
lder
NS
NS
NS
NS
NS
NS
NA
NA
Rev Fish Biol Fisheries (2012) 22:95–135 107
123
Page 14
background levels by 1990, ranging from 0.001 to
0.10 ppb TPAH.’’ An oil-exposure index derived
from the TPAH burdens in mussels confirms that oil
exposure was not a significant factor affecting adult
mortality in the single-factor or multi-factor ASA
models of Deriso et al. (2008) (Table 4). Carls et al.
(2002), Marty et al. (2003), Rice and Carls (2007)
and Hulson et al. (2008) find factors other than the
EVOS (e.g., poor nutrition) to be the likely direct
cause of the 1992–1993 decline.
Thomas and Thorne (2003) and Thorne and
Thomas (2008) suggest that the decline actually
began in 1989 and was associated with the EVOS. As
a mechanism for exposure to oil and to subsequent
mortality, they postulate direct contact with oil when
herring break the water surface during ‘‘gulping’’ of
air to fill their swim bladders. At dusk, herring
schools ascend from deeper waters to surface waters
to feed (Blaxter 1985). While at the surface, herring
may gulp air to fill their swim bladders and regain
buoyancy. The extent of gulping is unknown.
Gulping is a plausible way for herring to contact
oil on the water surface, but such contact would be
brief and intermittent. Behavioral avoidance of slicks
and sheens is also plausible and would prevent or
minimize exposure. Herring avoid suspended dredged
materials (Johnston and Wildish 1981), but the extent
to which herring avoid oil is unknown.
Assuming herring were unable to avoid slicks and
sheens, the window for significant direct contact with
EVOS oil in slicks and sheens would have been
limited primarily to spring 1989 and would have
ended when EVOS sheens decreased to background
by the fall of 1990. The oil slicks from the Exxon
Valdez began to exit PWS about one week after the
beginning of the spill (Wolfe et al. 1994). By April
1989, floating oil within PWS had fallen to primarily
surface sheens (Wolfe et al. 1994).
At their greatest extent, sheens associated with
EVOS covered only a very small area of PWS on a
few days. The occurrence, volume, and source of
PWS sheens were surveyed from summer 1989 until
fall 1990 (Taft et al. 1995). The majority of the
sheens (75%) were small (\1 l) with only 19 of the
total 827 sheens being greater than 1,000 l in oil
volume. A rainbow-colored sheen has a thickness of
0.0003 mm (Taft et al. 1995) so that a rainbow sheen
of 1,000 l would cover an area of about 3,000 m2. Of
these 19 sheens, 13 large-volume sheens associated
with the EVOS occurred on October 24, 1989, after a
storm had released oil from beach sediments. These
13 sheens had volumes from 1,000 to 54,000 l and
covered areas from 0.003 km2 about 0.34 km2. If all
13 were of the maximum observed extent, these
sheens would have covered about 2.2 km2 or about
0.049% of PWS total area of 9,059 km2. After March
1990, no sheens from EVOS were greater than 65 l in
oil volume. By September of 1990, PWS sheens from
EVOS had decreased to smaller volumes than those
associated with normal vessel traffic.
If direct and lethal contact between herring and oil
had been widespread in 1989, then large, observable
herring kills would be expected, but none was
observed. In contrast, mass mortalities of herring have
been observed during epizootics (Tibbo and Graham
1963; Patterson 1996; Meyers et al. 1999) and an
industrial spill of elemental phosphorus (Winters et al.
1986). However, even with the massive boat traffic
throughout PWS during the 1989 and 1990 cleanup—
and with scientists specifically looking for fish kills—
no herring (or other fish) kills were observed.
If herring had direct and lethal contact with spilled
oil, adult herring survival would have been lowered
in 1989. In fact, adult herring survival was highest in
1989 and lowest in 1992 (Hulson et al. 2008)opposite
to the expectation if direct contact with slicks was
inducing substantial adult herring mortality.
Viral hemorrhagic septicemia virus
Viral hemorrhagic septicemia virus (VHSV) infects
Pacific herring and other marine fish species along the
west coast of North America (Meyers et al. 1994;
Hedrick et al. 2003). Recent studies suggest that VHS
is primarily a disease of age-0 herring and that the
disease is either rapidly resolved or fatal within
2 weeks following infection (Kocan et al. 2001;
Hershberger et al. 2007; Elston and Meyers 2009).
Fish surviving VHS acquire immunity. VHSV is an
opportunistic pathogen in adult fish that produces
disease when fish are stressed by environmental or
other factors (Meyers et al. 1994; Elston and Meyers
2009).
Based on an analysis of VHSV-positive herring
samples in 1993 from PWS and other areas, Meyers
et al. (1994) concluded that the role of VHSV in the
1992–1993 decline could not be established with the
1993 data. A critical and comprehensive review of
108 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 15
the role of disease in PWS herring population
dynamics from 1989 to 2005 led Elston and Meyers
(2009) to conclude that VHS was not a primary cause
of the 1992–1993 decline for several reasons: (1)
VHSV is principally a pathogen of juvenile herring
(Kocan et al. 1999a); (2) the mortality in 1992–1993
was distributed across all age classes including those
age classes resistant to the disease (Elston et al.
1997); (3) no mass mortality was observed; (4) the
poor condition of the herring itself was sufficient to
cause the decline; and (5) most herring simply did not
return to spawning grounds in 1993 and could not be
examined for cause of death. The designation of the
herring population crash of 1992–1993 as an ‘‘epi-
demic,’’ implying a primary infectious cause (e.g.,
Marty et al. 2003; Marty 2007) is without basis.
Elston and Meyers (2009) evaluated all evidence
regarding VHS in PWS herring from 1989 to 2005
and found no evidence to support any assignment of
significant effects of VHSV on PWS herring biomass
or recruitment.
Additionally, hypothesis tests using the ASA
modeling by Deriso et al. (2008) did not find VHSV
to be a significant factor when tested in the model as a
single factor. The VHSV indices used for younger
fish were selected by the hypothesis test procedure as
the fourth factor in a four-factor model, however the
regression coefficient had the incorrect (i.e., positive)
sign.
Viral hemorrhagic septicemia virus and oil
exposure
Pearson et al. (1999) found no evidence that either
acute or chronic oil exposure from the EVOS
contributed to the 1993 decline. Carls et al. (2002)
stated that the consequences of high biomass, disease,
and poor nutrition are the more likely explanations of
the 1993 decline but indicated that indirect links to
the EVOS could not be entirely ruled out. Carls et al.
(2002) and Marty et al. (2007) find a link from oil
exposure to the decline unlikely, in part because of
the time delay between oil exposure and the decline.
An indirect link from oil exposure through disease
is suggested by one experiment of Carls et al. (1998)
but is not supported by subsequent experiments. In
reviewing experiments that attempted to induce or
exacerbate VHSV in herring by weathered or fresh
crude oil exposure (Carls et al. 1998; Kennedy et al.
1999; Kocan et al. 1999b; Sanders 2005), Elston and
Meyers (2009) found that none of the experiments
with juvenile and adult herring demonstrated a causal
link between any oil effect and VHSV at oil-exposure
concentrations as high as 178 to 328 ppb TPAH
(range of highest treatments used) for between 14 and
28 days. Rice and Carls (2007) now also consider the
EVOS not to be a causal or contributing factor in the
decline.
Pathogen Ichthyophonus hoferi
The pathogen Ichthyophonus hoferi has long been
associated with epizootics leading to mass mortalities
and low population levels in Atlantic herring (Tibbo
and Graham 1963; Patterson 1996; Kramer-Schadt
et al. 2010). An alternative or additional causative
agent, I. hoferi, was identified by Meyers et al. (1994)
and Marty et al. (1998) but was believed an incidental
finding at the time and subsequently (Quinn et al.
2001; Marty et al. 2003).
Recent work by Kramer-Schadt et al. (2010),
studying Norwegian spring spawning herring
(NSSH), and a reiteration of the ASA model (Marty
et al. 2010) suggest that I. hoferi can be an important
factor in herring population declines. However,
Kramer-Schadt et al. (2010) demonstrate that repre-
sentative sampling of herring populations to estimate
prevalence, severity, and impact of I. hoferi requires
extensive effort to account for broad geographic
migration patterns, differential distribution of the
disease by age class, and clustering of infected fish in
schools more vulnerable to sampling.
Sampling of PWS herring for the presence of
I. hoferi prior to the decline was limited but two points
derived from all available data are contrary to the
I. hoferi as the causative agent in the decline. First,
mass mortalities with large numbers of moribund and
dead fish in shallow areas have been associated with
I. hoferi in Atlantic herring (Sindermann 1958), but the
geographically limited mortality observed in PWS
herring in 1993 (Meyers et al. 1994) showed a very low
prevalence of I. hoferi. Removal of moribund fish by
predation, as observed by Kramer-Schadt et al. (2010)
for Atlantic herring, remains a possibility, but it also
must remain an unknown for which there is no
evidence in PWS in 1993. Second, the prevalence of
I. hoferi infection in PWS (based on very limited
sampling, and determined by histology, Table 5)
Rev Fish Biol Fisheries (2012) 22:95–135 109
123
Page 16
before and during the decline was low, i.e., 5.7% in
spring 1992 and 5.1% in spring 1993 but increased to
29% in 1994 after the decline (Table 5). This pattern of
low levels before the decline and high levels after the
decline is the reverse of the expectation of high
prevalence before and lower prevalence after the
decline if I. hoferi was causal in the decline. The
expectation would be that the prevalence decreases
after the epizootic because infected fish succumb to
disease. The prevalence data is contrary to the notion
that the I. hoferi infection was causative in the PWS
decline (Table 5).
Hypotheses concerning poor recovery
Hypotheses for poor recovery must provide mecha-
nisms for (1) no recruitment of strong year classes in
PWS after recruitment of the 1988 year class and (2)
low herring biomass. Recovery hypotheses examined
(Table 3) include oil exposure, harvesting, disease,
oceanic factors, predation, and competition.
Oil exposure
Among the post-decline exposure pathways that could
limit development of strong year classes, and hence
recruitment, the exposure of most concern is that of
juvenile (age-0? and 1?) herring to water-column
TPAH concentrations. Nearshore areas of PWS bays
and fjords provide nursery areas for juvenile herring
(Stokesbury et al. 1999, 2000; Norcross et al. 2001). In
such locations, mussels provide composite water
sampling of intertidal and shallow subtidal areas.
The use of bivalves for monitoring the marine
environment dates back to the 1970s (Farrington and
Quinn 1973; Dunn and Stich 1976; Dunn and Young
1976). Mussel tissues have been used extensively in
determining levels of total TPAH in the waters of PWS
(Short and Harris 2006; Boehm et al. 2004, 2007). For
Table 5 Prevalence of Ichthyophonus hoferi in PWS Pacific herring samples and mean age of fish sampled
Year I. hoferiprevalencea
Method usedb Mean age of
herring (years)c
1989 13.0 Histology ND
1990 ND ND ND
1991 3.4 Histology 7.3
1992 5.7 Histology 4.7
1993 5.1 Histology ND
1994 23.1 Histology 6.2
1995 21.7 Histology 6.2
1996 20.4 Histology 5.8
1997 16.2 Histology 5.4
1998 17.6 Histology 4.7
1999 24.3 Histology 6.1
2000 21.7 Histology 6.2
2001 38.0 Histology 5.7
2002 15.0 Histology 4.0
2003 20.8 Gross/Calc 4.2
2004 28.4 Gross/Calc 5.1
2005 50.6 Gross/Calc 5.7
2006 40.0 Gross/Calc 6.2
a Values for I. hoferi prevalence for 1989–2006 from Marty et al. (2010). ND no data availableb Prevalence of I. hoferi for 1994–2002 is the prevalence of cases with sum-I. hoferi score[0 by histology. Prevalence of I. hoferi for
2003 through 2006 was calculated using values for I. hoferi prevalence determined by gross examination of the heart and the first
order linear regression of gross I. hoferi prevalence (x) versus prevalence determined by sum-I. hoferi score[0 (y) for samples from
1997 through 2002 (y = 1.81x ? 2.05; r2 = 0.89) from Marty et al. (2007) and as cited in Marty et al. (2010)c Mean age values from Marty et al. (2010). ND no data available
110 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 17
PWS, Neff and Burns (1996) applied an equilibrium-
partitioning technique to relate tissue burdens of
TPAH in mussels to contaminant levels in the water
column. Direct water-column sampling was discon-
tinued several years after the EVOS because TPAH
concentrations were at background or below method
detection limits. Boehm et al. (2004, 2007) produced
an extensive database of the concentrations of TPAH
in mussel samples collected from 1989 to 2005 from
oiled and unoiled shorelines including sites of past
human activity (e.g., mines, canneries) in PWS.
Figure 5 shows the arithmetic means of TPAH
water concentrations derived from analyses of mussel
tissues from oiled and unoiled shorelines using the
conversion of Neff and Burns (1996). The profile of
exposure concentrations in Fig. 5 shows the peak of
3.3 ppb in 1989 (the year of the spill), followed by
exponential decay to less than 0.1 ppb in 1990 and to
low background levels (\0.02 ppb) since 1995.
The TPAH levels were too low to cause toxic
effects. Even in 1989, the TPAH levels (3.3 ppb)
were below the 28 ppb that Carls et al. (1998) found
associated with VHSV in one laboratory experiment.
For all years except 1989, water concentrations were
also below the value of 0.4 ppb, the value which
Carls et al. (1999) postulated as the threshold level
for the development of abnormal larvae in herring
eggs exposed to oil. While evidence of localized
effects on herring eggs were observed by Pearson
et al. (1995) in 1989, there was no evidence of effects
on eggs in 1990 (Brown et al. 1996; Pearson et al.
1995). In 1995, Johnson et al. (1997) incubated eggs
from mature herring collected from oiled and unoiled
areas of PWS and found no regional differences in
fertilization, hatching times, hatching success, or
larval viability or abnormalities. The water TPAH
data and experimental studies provide no evidence
that exposure to oil in the water column has affected
herring or prevented their recovery in PWS.
Subsurface oil residues are not a source of
exposure for herring. Juvenile herring forage in the
water column and do not dig in sediments. Other
factors also limit the potential of subsurface oil
residues as an exposure source for herring. First, there
was little overlap between herring spawning grounds
and oiling at the time of the EVOS (Fig. 6) (Pearson
et al. 1995, 1999) and much less between the residual
oiling and spawning areas (Rice and Carls 2007).
Second, the bioavailability of subsurface oil residues
is low. In 2002, Neff et al. (2006) analyzed beach
1989
1990
1991
1992
1993
1994
1995
1998
1999
2000
2001
2002
2003
2004
TPA
H (
ppb)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
1990
1991
1992
1993
1994
1995
1998
1999
2000
2001
2002
2003
2004
TPA
H (
ppb)
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07Oiled Unoiled
Fig. 5 Arithmetic mean TPAH concentrations (ppb) in Prince
William Sound seawater. Mussel samples were collected from
coastal waters oiled in 1989 and unoiled areas. Tissue values
from Boehm et al. (2007) were converted to seawater
concentrations using the method of Neff and Burns (1996)
Rev Fish Biol Fisheries (2012) 22:95–135 111
123
Page 18
sediment, beach pits, and intertidal biota for 17 sites
heavily oiled in 1989 and found low TPAH concen-
trations in biota (e.g., means of 22.4 ± 59 (SD) ppb
in fish from oiled sites, 6.4 ± 4.9 ppb in unoiled
sites, and 108 ± 165 ppb in unoiled sites with human
activity). In 2002, TPAH concentrations in the tissue
of the cockscomb gunnel, which inhabits the inter-
tidal and shallow subtidal zones of PWS, and in the
tissues of clams, mussels, and worms, did not differ
significantly between previously oiled sites and
unoiled reference sites and are predominantly from
combustion sources—not from crude oil (Neff et al.
2006).
The ASA modeling results do not support an effect
of oil exposure on recruitment. Deriso et al. (2008)
used TPAH concentrations in PWS nearshore areas as a
covariate in their ASA modeling. The TPAH levels fell
during the spill and have been at background levels for
over two decades. The oil-exposure index derived from
the TPAH burdens in mussels was not a significant
factor affecting recruitment in the single-factor or
multi-factor ASA models (Deriso et al. 2008).
Fig. 6 Herring spawning
areas and oiled shorelines in
PWS during 1989. Base
map from Pearson et al.
(1999) reproduced with
permission. The boundaries
of the Herring Spawning
Areas designated by ADFG
have changed slightly over
the year, and the 1989
boundaries given by Biggs
et al. (1992) appear as an
overlay on the base map
112 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 19
Harvest
The evidence that harvest effects are not involved in
the poor recovery of PWS herring derives from the lack
of strong year classes after 1993 in both PWS and Sitka
(Fig. 2). Fishing has continued at Sitka but not PWS.
This parallelism between PWS and Sitka indicates that
the lack of strong year classes in PWS since 1993 is due
to some regional-scale environmental change rather
than a harvest effect. Also, Deriso et al. (2008) found
that the commercial harvest of PWS herring was
commensurate with harvests being actively managed
to be below 20% of the spawning biomass.
Spawning habitat loss
Loss of spawning habitat is one of the three main
general causes of fishery declines listed by Hilborn
(1997) and is an impediment to stock success (Ste-
phenson 1997). Spawning habitat, however, is not
limited and has not been lost in PWS. Spawning sites in
PWS are not subject to shoreline urbanization or other
widespread human disturbances that degrade or frag-
ment herring spawning habitat in other regions.
Shifts in PWS herring spawning sites since the
1970s (Fig. 7) are typical of those in other herring
stocks. Spawning sites change over the years, to be
abandoned and then reoccupied (McQuinn 1997; Hay
et al. 2001). Both Atlantic and Pacific herring are
repeat spawners at particular sites but with varying
degrees of site fidelity (McQuinn 1997; Hay et al.
2001). This repeat spawning is not natal homing
because herring do not necessarily return to the site
where the fish were hatched. Herring spawning habitat
in PWS appears to be in excess of what herring require,
and, in any event, little if any spawning habitat was
impacted or lost as a consequence of the 1989 spill. The
shifts in spawning locations (Fig. 7) are not related to
oiled or unoiled condition following EVOS. For these
reasons, we reject the hypothesis that the poor recovery
of herring can be attributed to habitat loss from the
1989 oil spill.
Viral hemorrhagic septicemia virus
VHSV primarily affects juvenile herring so that the
potential role of VHSV in reducing recruitment and
limiting recovery needs to be examined. Quinn et al.
Calendar Year
1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Spaw
n K
ilom
eter
-Day
s
0
100
200
300
400
Naked Island North Northeast Southeast Montague Island
Fig. 7 Kilometer-days of herring spawn in Prince William Sound by subarea by calendar year. The 1989 boundaries for the ADFG
Herring Areas appear in Fig. 6. Data from: Dillingham et al. (2004)
Rev Fish Biol Fisheries (2012) 22:95–135 113
123
Page 20
(2001) incorporated disease indices in their ASA
model of PWS herring and found that negative
correlations between VHSV prevalence and recruit-
ment or the logarithm of recruitment at lags of 0 year
and 1 year were not significant. Lagging the VHSV
indices by 1 year improved the negative correlation
with recruitment, but the correlation was still not
significant. Quinn et al. (2001) suggest that disease
may still be a factor in recruitment, the effect of
which was masked by other factors in their study.
Based on a later version of the ASA model, Marty
et al. (2003) conclude that both VHSV and Ichthy-
ophonus hoferi may have affected the PWS herring
population since the 1993 decline. Elston and Meyers
(2009) reviewed all VHSV data from PWS herring
from 1989 through 2005, including correlation anal-
ysis with recruitment lags of 0, 1, 2, and 3 years and
concluded that there was no evidence that VHSV
affected recruitment processes after the decline.
These authors also provided corrected prevalence
data showing that the prevalence of VHSV in PWS
herring never exceeded 7.0% from 1994 to 2002 in
contrast to Marty et al. (2003). The ASA modeling of
Deriso et al. (2008) does not find VHSV adversely
influencing ages-3 and -4 herring recruitment. The
ASA modeling of Hulson et al. (2008) finds only
small influence of VHSV prevalence on herring
mortality in PWS and supports Elston and Meyers
(2009). As discussed in more detail in Elston and
Meyers (2009), the evidence is not sufficient to assign
VHSV as a major factor limiting recovery.
Viral hemorrhagic septicemia virus and oil
exposure
As with the decline, the experimental evidence does
not support any causal relationship between oil
exposure and VHSV in the poor recovery of PWS
herring. With the oil levels in PWS at background
levels throughout the period of poor recovery and
lack of evidence for a significant effect of VHSV on
herring with or without oil present (Elston and
Meyers 2009), a linkage between oil exposure,
VHSV, and poor recovery is not warranted.
Pathogen Ichthyophonus hoferi
The marked and progressive increase in observed
infection rate of herring by of I. hoferi after the
decline (Table 5) suggests some involvement of
disease in the recovery process. Ichthyophonus hoferi
is a pathogen producing chronic disease that pro-
gressively affects older age classes of herring more
severely (Marty et al. 1998; Hershberger et al. 2002).
I. hoferi is associated with adult Atlantic herring
mortality, in some cases on a massive scale (Tibbo
and Graham 1963; Patterson 1996). Marty et al.
(2003) concluded from ASA modeling that the PWS
prevalence of I. hoferi did not affect adult herring
survival from 1994 to 2000. The available PWS
disease data suggest that the I. hoferi infection rate is
more influential in mortality of adult PWS herring
than VHSV, especially since 2000 (Marty et al.
2010).
This hypothesis may ultimately be proven correct,
but the findings from Kramer-Schadt et al. (2010)
indicate the I. hoferi prevalence data for PWS suffer
in two aspects that preclude drawing strong conclu-
sions from the prevalence data. First, the field
observational technique used to identify infected fish
needs to be corrected for sensitivity to achieve an
estimated match with histology diagnosis (a high
significant correlation was shown between field
observations and histology [Marty et al. 2010]).
Measurements with the field technique underestimate
infection rate (Table 5; Kocan et al. 2011; Kramer-
Schadt et al. 2010) as determined by the more
accurate tissue explant method. The time series
produced by correction from field observation to
histology appears reasonably consistent, but total
infection rate likely remains underestimated (Kocan
et al. 1999a, b, 2011). There is also a lack of
correlation between prevalence by histology or visual
examination and prevalence as determined by tissue
explant culture, so that the extent to which prevalence
was underestimated by the first two techniques likely
varied substantially between samples. Kocan et al.
(2011) visualized tissue sections from infected hosts
three-dimensionally, and found non-random distribu-
tion of the parasite, establishing a mechanism for the
variability in sensitivity encountered when using
histology for the detection of I. hoferi in infected
hosts. Second, sampling error for the PWS data is
likely large because Holst (1996) and Kramer-Schadt
et al. (2010) found infection rates were up to 3 times
higher in small compared to large catches. Swimming
impairment in infected fish appears to separate them
from the larger schools and can also be expected to
114 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 21
make infected fish more vulnerable to sampling.
Finally, the consequences of infection severities and
the time course of the disease under various infection
scenarios remain poorly understood. The observed
spatial and seasonal patchiness of infected fish
suggests that sampling herring for prevalence without
taking such variation into account could yield sam-
pling bias and erroneous conclusions.
Multiple disease factors
Several authors have introduced the hypothesis that
filamentous bacteria played a role in the decline and
lack of recovery of PWS herring populations (Hulson
et al. 2008; Marty 2007; Marty et al. 2007). In the
only reported, confirmed infection of this type in
Pacific herring, Carls et al. (1998) reported a low
mortality rate (3%), scale loss, and raised white
cutaneous foci associated with filamentous bacterial
rods in captive herring from near Juneau, Alaska.
However, culture of kidneys from PWS herring
sampled from 1994 to 2002 failed to produce any
filamentous bacterial isolates. Filamentous marine
bacteria (such as Tenacibaculum maritimum, for-
merly Flexibacter maritimus) are a part of the normal
external flora of marine fish and often, if not always,
require an injury to the tissues as a portal of entry
(Avendano-Herrera et al. 2006). However, once
established, T. maritimum can cause an aggressive
infection and contribute to mortality. Herring are
susceptible to stress, which can cause blood vessel
dilation, hemorrhaging, and external injuries to skin
and fins, opening the portals for secondary pathogens.
There is no hard evidence to show that T. maritimum
was responsible for the skin ulcers observed in PWS
herring between 1993 and 2005 (Elston and Meyers
2009), but it could have played a role in exacerbating
lesions in stressed fish. Given the low prevalence of
observed ulcers in PWS herring and the complete
absence of any data confirming infection by filamen-
tous bacteria, the assertion of a primary role of
filamentous bacteria in controlling PWS herring
populations is not supported.
Predation
The synthesis of Spies (2007) provides substantial
detail on trophic interactions in PWS and GOA (Gulf
of Alaska) and a background for assessing effects of
predation on herring. Forage fish in PWS include
juvenile Pacific herring, Pacific sand lance, capelin,
eulachon, juvenile salmon, juvenile pollock, and
some small mesopelagic fishes. Predators utilizing
this complex of forage fish include marine mammals,
marine birds, and larger fish. The high fat content of
sand lance, capelin, and juvenile herring make these
fish attractive prey to predators; for example, Pacific
herring, sand lance, and capelin comprise 58% of the
diet by weight of 14 seabirds in PWS (Spies 2007).
The ASA modeling of Deriso et al. (2008) finds
PWS hatchery releases of juvenile pink salmon to
consistently be the single factor affecting recruitment
(Table 4) but could not discern whether the mecha-
nism was predation or competition; the paper states
‘‘The salmon factors always were highly significant
and reduced the spawning biomass to 20% of its size
as calculated by removal of the hatchery effect in the
ASA [model].’’ In contrast, the ASA modeling did
not find adult coho predation applied to age-1 and
age-2 herring to be a significant factor influencing
recruitment. Similarly, the modeling did not find
adult sea lion predation to be a significant factor
influencing adult herring mortality.
Whale predation on herring can be significant.
Minke whale predation on Norwegian spring-spawn-
ing herring varies annually but can account for 45%
of the mortality (Tjelmeland and Lindstrom 2005).
For Atlantic herring on Georges Bank, both fin and
humpback whale predation removes substantial por-
tions of the herring biomass while seabird predation
removed about 100 times less biomass (Overholtz
and Link 2007). This section examines recovery
hypotheses concerning predation on herring by pink
salmon and humpback whales.
Pink salmon predation
Unlike all other herring fishery areas, PWS has pink
salmon hatcheries that contribute substantial numbers
of fry above and beyond wild fry entering the Sound.
Juvenile pink salmon releases from PWS hatcheries
began in the mid-1970s. The annual releases first
reached 600 million in 1989 and have been about 600
million since the mid-1990s (Fig. 8). Here we
examine the potential for predation by juvenile pink
salmon on juvenile herring. We do not consider
predation by returning adult pink salmon because
previous studies (Okey and Pauly 1999) report that
Rev Fish Biol Fisheries (2012) 22:95–135 115
123
Page 22
returning pink salmon ‘‘feed little while in [PWS], if
at all.’’
In contrast to adult pink salmon, which feed
primarily on zooplankton and micronekton (Rugger-
one and Nielsen 2004), juvenile pink salmon in PWS
consume substantial amounts of fish (Sturdevant
1999; Willette et al. 1999a). Stomach analyses of
PWS juvenile pink salmon by Sturdevant (1999)
reveal that fish are 28% of the diet by weight over all
months but are as high as 80% in July. Juvenile pink
salmon and age-0 herring overlap in the nearshore
nursery areas from July through at least August.
Fish in general select prey that are 20–30% of their
body size, but they can select prey that are up to 50%
of their size when they commence piscivory (Willette
et al. 1999b). Juvenile herring are smaller than pink
salmon when the herring enter PWS. Age-0 herring
enter the Sound in May at less than 0.2 g average
body weight (Okey and Pauly 1999) and grow to
6.9 g the first summer (Sturdevant et al. 2001). Pink
salmon, by contrast, enter the Sound in mid-May at
0.26 and 0.53-g average body weights for wild and
hatchery fish, respectively, and grow to 99 g by
October (Cross et al. 2005). In May, age-1? herring
weigh 12 g and are much larger than any of the other
0? forage fish. Juvenile pink salmon grow more
quickly than the age-1? herring but do not outweigh
them until August. Predation by juvenile pink salmon
most likely affects age-0 rather than age-1 herring.
Humpback whale predation
The humpback whale (Megaptera novaeangliae) is a
predator that targets herring aggregations (Hjort and
Ruud 1929). Such targeted predation leads Brown
(2004) to suggest that PWS herring are now in ‘‘a
predator pit,’’ where recruitment cannot exceed losses
from predation.
Humpback whales visiting PWS are part of the
Central North Pacific Stock of humpback whales,
which has increased in recent years, primarily from
increases in Southeast Alaska (Angliss and Outlaw
2005). A recent estimate for the Central North Pacific
Stock has 2,036 whales with 147 comprising the PWS
population. Whales observed in PWS are transient,
being observed elsewhere in the GOA during the
same summer (Von Ziegesar et al. 1994; Angliss and
Outlaw 2005) and in other parts of the range in other
seasons (Calambokidis and Steiger 2001).
Whale abundance in PWS peaks in July and falls
to low levels by mid-September. The number of
humpbacks identified in PWS from 1988 to 1993
averaged 49 with a range from 26 to 66 (Von
Ziegesar et al. 1994; Calambokidis and Steiger 2001).
Von Ziegesar (pers. comm.) estimates that PWS has
50 to 70 whales during the summer and 10 to 20
whales during the winter. Von Ziegesar et al. (1994)
estimated summer residence time for individual
whales in PWS to be 49 days for summer resident
Calendar Year
1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006
Bio
mas
s (m
etri
c to
ns)
0
20000
40000
60000
80000
100000
120000
140000
160000
Juve
nile
Pin
k Sa
lmon
Rel
ease
d (m
illio
ns)
0
100
200
300
400
500
600
700
Prince William Sound ASA Biomass (metric tons) (Moffitt, 2006)Prince William Sound ASA Biomass (metric tons) (Donaldson et al. 1995)Prince William Sound Pink Salmon Fry Release (millions) (Ashe et al. 2005; Farrington 2004; White 2005)
Fig. 8 Prince William
Sound herring ASA
biomass and pink salmon
fry released from hatcheries
by calendar year. Data
from: Ashe et al. (2005),
Donaldson et al. (1995),
Farrington (2004), Moffitt
(2006), White (2005)
116 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 23
whales and 27 days for other more transient whales.
During summer feeding in PWS, humpback distribu-
tion in 1988 to 1990 centered on the southern end of
Knight Island and Montague Strait (Von Ziegesar
et al. 1994).
Preliminary results from ongoing studies of hump-
back whales in PWS indicate that the overwintering
population is higher in recent years (2008, 2009) than
that reported by Von Ziegesar et al. (1994) for the
1990’s. PWS-wide surveys estimated the numbers of
whales overwintering in PWS to be 67 in September
2008 and with a peak of 112 in October 2008 (Rice
2009). Slightly over 50 whales were estimated to be
overwintering in November and December 2008, and
only 8 whales, in January through March 2009. The
total PWS whale population also appears to have
increased to 160 (Rice 2009).
The diet of humpback whales in their North Pacific
feeding grounds consists mainly of zooplankton and
small schooling fish (Angliss and Outlaw 2005).
Reconstruction of current humpback diets near
Kodiak Island estimated that the bulk of the diet
was comprised of walleye pollock (37%), capelin
(29%), and euphausiids (22%) with herring being
only 0.03% of the diet (Witteveen et al. 2006).
Stomach analyses of humpback whales conducted in
the 1920s near Kodiak Island indicated that surf smelt
and euphausiids were the major dietary items (Wit-
teveen et al. 2006). Based on sonar surveys and scat
analysis, the majority of the potential prey in PWS
during the summer were euphausiids, but herring and
sand lance also were present (Matkin and Hobbs
1999). Based on observations of the winter distribu-
tion of herring and humpbacks, Matkin and Hobbs
(1999) and Brown (2004) have inferred that the
overwintering whales are feeding on herring. How-
ever, studies have shown that high euphausiid
abundance leads to aggregations of humpback and
other whales (Hjort and Ruud 1929 cited in Sverdrup
et al. 1942; Everson 2000). Also plausible is the
inference that the observed winter co-occurrence
results from pursuit by both herring and humpbacks
of euphausiids, probably a major, if not the dominant,
zooplankton in winter. Preliminary results of ongoing
visual and acoustic observations along with fatty acid
analyses indicate that overwintering whales in PWS
frequently feed on herring (Rice 2009).
Tables 6 and 7 present estimates of the consump-
tion of herring by humpbacks in PWS during the
summer feeding season and the overwintering
period. Median value for the standing stock of
humpbacks of 60 whales in the summer is taken
from the range given by Von Ziegesar et al. (2006
pers. comm.). The monthly values for overwintering
whales are taken from Rice (2009). A range of
values for the daily consumption rates was used to
estimate consumption. The information on the diet
composition for summering whales leads us to use a
worst-case high estimate of consumption by assum-
ing that 50% of the diet is fish and that 95% of the
fish are herring. Similarly, we estimate predation by
overwintering whales, assuming that the winter diet
is 80% herring.
In the summer, the whales would feed primarily on
age-2 herring because the younger age classes are in
nearshore nursery areas inaccessible to whales and
adult age classes are outside PWS or in eastern parts
of PWS where whales do not aggregate in summer
(Stokesbury et al. 1999; Brown et al. 2002). Assum-
ing an estimated weight of 74 g for individual age-2
herring, estimated consumption by 60 summering
whales would range from 0.011 to 44.9 million sub-
adult herring with an average of about 19.6 million
sub-adult herring (Table 6). Assuming an average
weight of 158.4 g for individual adult herring,
estimated consumption by overwintering whales
ranged from 17 to 102 million herring with an
average of about 51.3 million adult herring (Table 7).
The abundance of the PWS adult herring popula-
tion was about 900 million per year from 1980
through 1993 (estimated from Fig. 1) and about 200
million per year from 1994 to 2005 (Fig. 2). The
overwintering whales appear to be consuming about
26% of the recent adult population. The abundance of
age-3 herring in PWS was about 81 million fish per
year from 1994 to 2005. We have no estimates of the
age-2 population. Assuming the age-2 population to
be generally greater than the age-3 population, the
summering whales would then be consuming less
than 24% of the age-2 population.
Poor nutrition and oceanic factors
Modeling by Rose et al. (2007) illustrates how
temperature and other oceanic factors can influence
weight at age and recruitment in herring populations.
Oceanic factors have been found to influence herring
recruitment in Alaska (Collie 1991; Williams and
Rev Fish Biol Fisheries (2012) 22:95–135 117
123
Page 24
Quinn 2000a, b; Zebdi and Collie 1995). The analysis
of 14 herring stocks by Williams and Quinn (2000a,
b) show that the GOA stocks, which include the PWS
and Sitka stocks, demonstrated similar interannual
patterns in recruitment and weight at age that set
them apart from herring stocks in the Bering Sea or
British Columbia. The similarities among the GOA
stocks indicate that they respond in a similar way to
regional-scale environmental factors occurring over
the GOA (Williams and Quinn 2000b).
Evidence for operation of oceanic factors in PWS
and GOA comes from a comparison of recruitment
(the numbers of herring at age-3 and age-4) that
reveals that emergence of strong years occurred at the
same time at PWS and Sitka from 1980 to 1993
(Fig. 2, Table 8). More importantly, strong year
classes emerged at the same time in both PWS and
Sitka until 1993; after 1993, strong year classes have
not emerged in either PWS or Sitka (Fig. 2). Auto-
correlation analysis supports such synchrony. The
4-year lag provides the highest positive autocorrela-
tion for the numbers of both age-3 and age-4 herring
at PWS and Sitka from 1980 through 1993 (Table 9).
After 1993, the high positive autocorrelations at
4-year lag are no longer evident at either PWS or
Sitka. The parallel trends are strong evidence that
population dynamics of both PWS and Sitka herring
are being influenced at the same time by GOA-wide
ocean processes.
Further, the nutritional state of the PWS herring
after the decline remains much poorer than that at
Sitka (Table 10). With the data available, we may
not be able to separate the relative roles of poor
nutrition and other oceanic factors, but the field
observations of Foy and Norcross (2001) and the
analysis by Norcross and Brown (2001) suggest how
these factors play a role in determining year-class
strength in PWS herring. Elsewhere, changes in the
ocean environment have apparently changed the
zooplankton prey field and reduced overwintering
larval survival in Atlantic herring, in turn leading to
six sequential years of poor recruitment (Payne et al.
2009).
Poor nutrition associated with oceanic factors is a
plausible hypothesis for poor recovery in PWS
herring and has empirical evidence for its occurrence.
However, some additional process beyond environ-
mentally-derived poor nutrition must be occurring in
PWS but not at Sitka. The ASA modeling outcomes
suggest that the effects of the hatchery releases of
pink salmon fry operate as well and are of great
influence on recruitment and recovery.
Table 6 Estimated consumption of herring by Humpback whales in PWS during summer
Element/
source
Feeding rate
(kg per day)
Consumption
(kg per season)
assuming season
(days) = 122
Proportion of
diet as fish
Number of herring
consumed per whale
assuming weight
(g) of individual
herring = 74.4
Number of herring
consumed in PWS
assuming number
of whales = 60
Number of herring
consumed as % of
subadult population
= 81,000,000 (%)
Matkin and
Hobbs
(1999)
890 108,580 0.48 700,516 42,030,968 52
Matkin and
Hobbs
(1999)
890 108,580 0.10 145,941 8,756,452 11
Tynan (2004) 951 116,022 0.48 748,529 44,911,742 55
Tynan (2004) 2,200 268,400 0.02 72,151 4,329,032 5
Witteveen
et al. (2006)
370 45,140 0.48 291,226 17,473,548 22
Witteveen
et al. (2006)
370 45,140 0.0003 182 10,921 0.013
Mean 945 115,310 0.26 326,424 19,585,444 24
Minimum 0.03 182 10,921 0.013
Maximum 48.00 748,529 44,911,742 55
118 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 25
Competition with juvenile pink salmon
Results from the ASA modeling with covariates by
Deriso et al. (2008) appear in Table 4. These results
led us to evaluate mechanisms of competition and
predation by which juvenile pink salmon could
influence population dynamics in PWS herring. We
assessed the competition hypotheses on the basis of
both the ASA modeling results and the available
information from field studies.
Juvenile pink salmon releases into PWS now
average about 600 million annually (Fig. 8). ASA
modeling by Deriso et al. (2008) found releases of
hatchery pink salmon fry to consistently be the most
influential single factor on PWS herring recruitment
and the common factor in all their multi-factor
models (Table 4). The impact analysis of Deriso et al.
(2008) reveals that the pink salmon fry releases have
the most substantial impact, reducing the spawning
herring biomass to an estimated 20% of the level
without the releases. However, Deriso et al. (2008)
could did not address whether the mechanism behind
these results was predation, competition, or a com-
bination of the two.
Here we examine potential competition mecha-
nisms by which the pink salmon fry releases influence
PWS herring recruitment. To assess competition
hypotheses, we examine several questions: (1) Do
the competing species overlap in space and time?
(spatio-temporal overlap), (2) Do the diets of the
competitors include shared prey? (dietary overlap),
(3) Does consumptive demand exceed food avail-
ability (consumptive demand), (4) Is the feeding of
one competitor substantially reduced in the presence
of the other? (feeding disruption), and (5) DoesTa
ble
7E
stim
ated
con
sum
pti
on
of
her
rin
gb
yH
um
pb
ack
wh
ales
inP
WS
du
rin
gfa
llan
dw
inte
r
Ele
men
t/so
urc
eF
eed
ing
rate
(mt
per
day
)
Pro
po
rtio
no
f
die
tas
fish
Wei
gh
t(m
t)o
fh
erri
ng
con
sum
edb
y
nu
mb
ero
fw
hal
eso
bse
rved
,2
00
8,
20
09
To
tal
wei
gh
t(m
t)o
f
her
rin
gco
nsu
med
for
the
ov
erw
inte
rin
g
per
iod
Nu
mb
ero
fh
erri
ng
con
sum
edp
erw
hal
e
assu
min
gw
eig
ht
(kg
)
of
ind
ivid
ual
her
rin
g=
0.1
58
4
Nu
mb
ero
fh
erri
ng
con
sum
edas
%o
f
adu
ltp
op
ula
tio
n=
20
0,0
00
,00
0(%
)S
EP
67
OC
T
11
2
NO
V
53
DE
C
51
JAN
8
FE
B
8
MA
R
8
Mat
kin
and
Ho
bb
s(1
99
9)
0.8
90
0.8
01
,43
12
,39
21
,13
21
,08
91
71
17
11
71
6,5
58
41
,39
8,4
85
21
Ty
nan
(20
04
)0
.95
10
.80
1,5
29
2,5
56
1,2
10
1,1
64
18
31
83
18
37
,00
74
4,2
35
,90
92
2
Ty
nan
(20
04
)2
.20
00
.80
3,5
38
5,9
14
2,7
98
2,6
93
42
24
22
42
21
6,2
10
10
2,3
33
,33
35
1
Wit
tev
een
etal
.(2
00
6)
0.3
70
0.8
05
95
99
54
71
45
37
17
17
12
,72
61
7,2
10
,60
69
Wit
tev
een
etal
.(2
00
6)
0.3
70
0.8
05
95
99
54
71
45
37
17
17
12
,72
61
7,2
10
,60
69
Mea
n1
.10
30
.80
1,7
73
2,9
64
1,4
03
1,3
50
21
22
12
21
28
,12
55
1,2
94
,58
32
6
Min
imu
m0
.37
0–
59
59
95
47
14
53
71
71
71
2,7
26
17
,21
0,6
06
9
Max
imu
m2
.20
0–
3,5
38
5,9
14
2,7
98
2,6
93
42
24
22
42
21
6,2
10
10
2,3
33
,33
35
1
Nu
mb
ero
fw
hal
eso
bse
rved
fro
mR
ice
(20
09
)
Table 8 Coefficients of determination for the numbers of
herring at age 3 and age 4 between PWS and Sitka Sound, 1980
to 2005
Time period Age R-sq P-value
All years 3 0.779 \0.001
4 0.727 \0.001
1993 and before 3 0.833 \0.001
4 0.823 \0.001
After 1993 3 0.449 0.017
4 0.282 0.067
Data from ADFG, Dressel (2006), Moffitt (2006)
Rev Fish Biol Fisheries (2012) 22:95–135 119
123
Page 26
lowered prey availability or feeding disruption
adversely affect growth and survival (growth and
survival). To affirm a competition hypothesis, we
must find evidence for all of the following: spatio-
temporal overlap, dietary overlap, consumptive
demand or feeding disruption, and effects of food
limitation on herring growth and survival.
Spatio-temporal overlap
There is evidence for spatial and temporal overlap
between juvenile pink salmon and juvenile herring.
The degree and timing of overlap between pink
salmon and age-0 herring differs from that between
pink salmon and age-1 herring. Some previous
discussions of the overlap (e.g., Cooney et al. 2001)
have only considered the limited spatial overlap
between age-0 herring and pink salmon in May and
June.
Wild pink salmon fry enter the Sound in late April
or May (Boldt and Haldorson 2003), and release of
hatchery pink salmon fry is timed to optimally
correspond with the spring zooplankton bloom,
usually mid-May (Cross et al. 2005). Once salmon
fry have entered the Sound, they initially aggregate in
shallow, nearshore bays (Boldt and Haldorson 2003).
Young-of-the-year salmon stay within the 20-m
isobath (Okey and Pauly 1999). Furthermore, most
of the hatchery-released pink salmon inhabit the
corridor running from Port Valdez to Esther Island
and to Knight Island Passage, roughly one-half of the
Sound (Okey and Pauly 1999; Willette et al. 2001).
Juvenile pink salmon can inhabit PWS through
October but begin to migrate out of PWS into the
coastal waters of the GOA in July (Armstrong et al.
2005). The vast majority of juvenile pink salmon
have emigrated from the Sound by September.
The geographic distributions of age-0 and age-1
herring overlap substantially with the distribution of
hatchery juvenile pink salmon within PWS. In Fig. 9,
the one-dimensional migration route for hatchery
pink salmon fry taken from Willette et al. (2001) is
superimposed on the distribution of age-0 and age-1
herring in July of 1995 and 1996 taken from Norcross
Table 9 Autocorrelations in time series of the numbers of
herring at age 3 and age 4 in PWS and Sitka Sound, 1980–2005
All years 1993 and before After 1993
Lag Autocorr Lag Autocorr Lag Autocorr
Sitka age 3 1 -0.06 1 -0.09 1 0.16
2 -0.3 2 -0.31 2 -0.18
3 -0.14 3 -0.07 3 -0.37
4 0.61 4 0.61 4 -0.13
5 -0.07 5 -0.09 5 0.37
6 -0.16 6 -0.24 6 0.17
Sitka age 4 1 -0.04 1 -0.08 1 0.31
2 -0.28 2 -0.22 2 -0.2
3 -0.13 3 -0.07 3 -0.23
4 0.59 4 0.61 4 -0.24
5 -0.06 5 -0.08 5 0.17
6 -0.16 6 -0.16 6 0.14
PWS age 3 1 -0.08 1 -0.28 1 0.08
2 -0.11 2 -0.33 2 -0.13
3 0.06 3 -0.02 3 -0.38
4 0.57 4 0.53 4 -0.18
5 -0.11 5 -0.3 5 0.18
6 -0.07 6 -0.26 6 0.07
PWS Age 4 1 -0.04 1 -0.29 1 0.09
2 -0.12 2 -0.28 2 -0.2
3 0.04 3 -0.09 3 -0.42
4 0.57 4 0.48 4 -0.24
5 -0.03 5 -0.29 5 0.17
6 -0.08 6 -0.16 6 0.14
Data from ADFG, Dressel (2006), Moffitt (2006)
Italics indicates significant difference at P B 0.05
Table 10 Comparisons of weight at age and length at age of Pacific herring between PWS and Sitka Sound from 1983 to 1993 and
from 1994 to 2005
Variable 1993 and before P value 1994 and after P value
PWS Sitka PWS Sitka
Weight at age (g) 119.5 116.7 0.6 123.1 133.2 0.007
Length at age (mm) 205.7 202.2 0.02 210.6 209.8 0.59
Data from ADFG, Bruce (2006), Dressel (2006), Moffitt (2006)
120 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 27
et al. (2001). Age-0 herring are in the larval drift in
May and June and, after metamorphosis, enter PWS
bays and fjords in June and July (Norcross et al.
2001; Stokesberry et al. 2000). From distribution
studies over several years, Stokesbury et al. (1999,
2000) and Norcross et al. (2001) conclude that the
age-0 and age-1 herring use the nearshore areas of
PWS bays and fjords as nursery areas throughout the
year. In western PWS, juvenile pink salmon overlap
with age-0 herring from July onwards and with age-1
herring from May onwards. July is probably the time
of maximum overlap when there are substantial
numbers of juvenile pink salmon still within the PWS
nearshore and both age-0 and age-1 herring are in
PWS bays.
Several lines of evidence indicate that the extent of
spatial and temporal overlap is biologically
meaningful. First, Wertheimer and Celewycz (1996)
state that their beach seining captured ‘‘large num-
bers’’ of Pacific herring and other fish ‘‘coincident with
juvenile [pink] salmon’’ during April to June of 1989
and 1990 in PWS. Second, Willette et al. (1997)
indicate that in August and September of 1994, at a low
stock level for PWS herring, co-occurrence of juvenile
pink salmon and juvenile herring in the same net hauls
was 27%. Willette et al. (1997) used an anchovy purse
seine that was 20-m deep; therefore, these are not
beach samples. The mean length of the herring
indicates that they were age-1 and with perhaps some
age-2. Third, of the 17 stations described by Sturde-
vant (1999) where either allopatric or sympatric
herring and pink salmon occurred, about 25% met a
rigorous definition of sympatry, i.e., both species being
captured in the same net haul (Fig. 9). Sturdevant’s
Fig. 9 The locations of Prince William Sound hatcheries,
distribution of herring nursery areas and migration route of
pink salmon. Migration route is redrawn from Willette et al.
(2001); sympatric and allopatric aggregations, from Sturdevant
(1999); herring distributions, from Norcross et al. (2001)
Rev Fish Biol Fisheries (2012) 22:95–135 121
123
Page 28
(1999) study was not intended to be a survey of all the
allopatric and sympatric conditions for juvenile her-
ring and other species. In her discussion, Sturdevant
suggests that sympatry among the forage fishes may be
greater than her collection data suggest. Fourth,
Moulton (1997) finds substantial numbers of juvenile
pink salmon and juvenile herring occurring in the same
tow net hauls in northern Cook Inlet. The tow nets
were conducted away from the beaches and sampled to
a depth of 3.1 m. Analysis of Moulton’s raw data
shows that the sympatry between juvenile pink salmon
and age-1 herring in tow net sets in the surface waters
of northern Cook Inlet during the months of June and
July is 81 and 58% (Table 11). Age-0 herring, but not
juvenile pink salmon, are caught in September. The
available field evidence indicates that the rate of co-
occurrence of juvenile pink salmon and juvenile
herring ranges from 25% to about 80% depending on
timing.
Dietary overlap
Both herring and pink salmon juveniles are highly
omnivorous, with both species consuming small and
large copepods, larvaceans, and euphausiids (Sturd-
evant 1999; Boldt and Haldorson 2003). Both species
prefer large copepods in May (Foy and Norcross
1999; Willette 2001). In general, pink salmon avoid
small copepods (Cross et al. 2005), although small
copepods may be underrepresented in diet studies
because of their small size (Boldt and Haldorson
2003). The degree of dietary overlap varies, depend-
ing on the time of year, the date of sampling, and the
study, but overlap in food resources appears greatest
at the beginning of the summer, when both species
are consuming the seasonally abundant calanoid
copepods, and both are occupying nearshore habitats.
Later in the summer, herring primarily consume
smaller prey than that preferred by pink salmon but
eat larger prey items such as larvaceans as a fraction
of their diet. Larger prey items are not numerically
abundant in the diets of juvenile herring, but these
prey items contribute disproportionately to the over-
all amount of calories ingested and hence increase the
energy content of juvenile herring (Sturdevant 1999).
Consumptive demand and prey availability
Cross et al. (2005) suggest that the consumption
demands of pink salmon may exceed food availabil-
ity in PWS. They used a bioenergetics model (Hansen
et al. 1997) to predict the consumption demand of
juvenile pink salmon, and compared the demand for
certain prey with the availability of these taxa within
the prey fields. Estimated consumptive demand of
high-energy-content prey, such as large copepods,
larvaceans, and hyperiid amphipods, exceeded supply
in PWS. One advantage of the Cross study is that they
were able to take zooplankton tows at the same sites
where they collected fish and thereby were able to
compare demand and regional availability. Sturde-
vant (1999) also found locally decreased zooplankton
populations in areas with high densities of sympatric
juvenile salmon and herring.
Disruption of feeding during sympatry
Sturdevant’s (1999) examination of dietary overlap of
14 species of forage fish collected from PWS includes
field observations of feeding disruption that provide
strong evidence of food competition between juvenile
pink salmon and herring. When Pacific herring occur
sympatrically with pink salmon, Sturdevant observed
significant and substantial decreases for herring in
Table 11 Montlhy catch of tow nets for juvenile Pacific herring and pink salmon in Cook Inlet, Alaska,1993, by month from raw
data for Moulton (1997) provided by Moulton (pers. comm. 2006)
Category Number of sets Percent of sets with w/pink salmon and/or herring
June July September June (%) July (%) September (%)
Total number of sets with pinks and/or herring 103 55 15 100.0 100.0 100.0
Sets with herring and pinks (co-occurrence) 83 32 0 80.6 58.2 0.0
Sets with herring and no pinks 6 12 15 5.8 21.8 100.0
Sets with pinks and no herring 14 11 0 13.6 20.0 0.0
122 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 29
four feeding indices: median stomach fullness, prey
weight in stomach as a percentage of body weight,
median prey number, and median prey biomass
(Table 12). These observations indicate that when
herring are food-limited in the presence of salmon,
they do not shift diet composition, but instead
decrease feeding. The observations of decreased
feeding indices under sympatry led Sturdevant
(1999) to conclude that if these fishes regularly co-
occur, the carrying capacity of PWS may be
exceeded. The spatio-temporal overlap observations
discussed above indicate that co-occurrence is
common.
Effects of food limitation on growth and survival
of juvenile herring
Field observations in PWS indicate that food limita-
tion influences growth and survival of juvenile
herring; the ability of herring juveniles to gain
sufficient fat reserves to live through the winter
determines their overall recruitment (Cooney et al.
2001; Norcross et al. 2001).
Paul and Paul (1998) experimentally starved
young-of-the-year herring to determine the critical
energy content of herring juveniles to survive winter.
They found that fish that died during laboratory
starvation had a whole-body energy content (WBEC)
of 3.2 kJ g-1, whereas the WBEC of young-of-the-
year herring in PWS was 3.4 kJ g-1, suggesting that
juvenile herring use most of their reserves during the
winter (Paul and Paul 1998). Fish with low WBEC
are inferred to be at greater risk of predation, because
they probably leave predator refuges in search of food
(Paul and Paul 1998).
Field, laboratory, and modeling studies have led
Norcross et al. (2001) to consider adequate nutrition
(stored energy) to be essential for overwintering
survival of age-0 herring, which in turn is considered
important to year-class strength. Foy and Norcross
(2001) reported that warmer water temperatures in
nursery bays in 1997 were associated with lower
zooplankton densities and taxonomic diversity, and
that the frequency of empty stomachs among juvenile
herring (a mixture of age-0 and age-1 in unknown
proportion) was three times higher in October of the
year with warmer water temperatures. In the past,
strong year classes have typically emerged in PWS
and other GOA herring populations every 4 years
(Funk and Sandone 1990; Williams and Quinn 2000a;
Hay et al. 2001). The 1996 year class would have
been expected to be a strong one in the four-year
cycle. However, the 1996 year class (age-1 at the
time of the warmer water temperatures) had one of
the lowest recruitments (less than one million age-3
herring) in the PWS recruitment time series (Fig. 2).
Norcross and Brown (2001) and Norcross et al.
(2001) suggest that no one early life stage is the
‘‘key’’ to year class strength in PWS herring, and the
above observations suggest that, at least in the case at
hand, nutritional state, low zooplankton densities, and
warmer water temperatures at age-1 were associated
with an extremely weak year class (Brown 2004;
Brown and Norcross 2001).
Discussion
Based on Platt’s (1964) guidance for making infer-
ences, we organize the summary information in
Table 12 Feeding indices for allopatric and sympatric pink salmon and Pacific herring in PWS from Sturdevant (1999)
Species Condition Median stomach
fullness (%)
Median prey
as % body Wt
Median prey
number
Median prey
biomass (mg)
Herring Allopatric 75 1.5 383.5 19.97
Sympatric Trace 0.4 24 1.68
Significance (P=) \0.0001 \0.0001 0.0001 0.0035
Pink Salmon Allopatric 75 1.6 288.5 24.81
Sympatric 50 0.8 123 25.1
Significance (P=) 0.0138 \0.0001 0.255 0.0371
Note that 0.0371 may be typographic error. A value of 0.371 would match the conclusion in the text in Sturdevant (1999)
Rev Fish Biol Fisheries (2012) 22:95–135 123
123
Page 30
Table 3 under these four questions: (1) Is there
correlation or constant association between the
hypothesized factor and herring mortality, biomass,
or recruitment? (constant association test), (2) Is there
a biologically plausible mechanism for the hypothe-
sized factor to produce the observed effect? (mech-
anism test), (3) Is there evidence for the occurrence of
such a mechanism? (test of supporting evidence for
mechanism), and (4) Is there contrary evidence? (test
for contrary evidence).
Decline
We find no evidence that oil exposure, harvest
effects, spawning habitat loss, or the spawn-on-kelp
fishery caused or contributed to the 1993 decline
(Table 3). The evidence supports the contention that
at least a portion of the PWS herring population was
stressed by poor nutrition. Further evidence supports
the argument that infectious disease effects during the
decline were a secondary response to poor nutrition.
Most investigators now agree with the conclusion
of Pearson et al. (1999) that the decline was not
related to the oil spill. The link between the decline
and the oil spill postulated by Thomas and Thorne
(2003) and Thorne and Thomas (2008) has a
plausible mechanism for exposure to oil during the
initial year of the spill but lacks empirical evidence
for its occurrence in PWS. The limited distribution of
oil slicks over the large area of PWS and intermittent
nature of gulping at the water surface indicates a
limited exposure via gulping during the spill year and
essentially negligible exposure after the initial year.
Further, the results of ASA modeling (Deriso et al.
2008; Hulson et al. 2008) and other available
evidence is contrary to the notion of a mortality
event in adult herring during the spill year or year
other than 1992 or 1993. Consequently, we reject the
hypothesis that the decline started in 1989 and was
associated with the 1989 EVOS.
The sum of the observations is not sufficient to
assign VHSV as a cause of the 1992–1993 decline.
Recent detailed analysis (Elston and Meyers 2009)
demonstrates that the preponderance of evidence does
not support any primary role of infectious disease in
the decline. VHSV appears to have been a secondary
response in an undetermined portion of the PWS
herring population when the herring had already been
stressed by poor nutrition.
The prevalence patterns of I. hoferi are contrary to
the notion that this pathogen caused or contributed to
the decline, although there is evidence that I. hoferi
may play a role in PWS herring population dynamics
some years after the decline. Because of the potential
for sampling bias and changing measurement tech-
niques, the data on the prevalence of I. hoferi requires
cautious use. Despite these limitations, the preva-
lence patterns during the decline do not match the
expected pattern if I. hoferi was a causative agent in
the decline.
Poor nutrition preceding severe I. hoferi disease is
a plausible mechanism for the recent emergence of
the disease in PWS. Kramer-Schadt et al. (2010)
hypothesize that the interactions of several factors
drive the progression from I. hoferi infection to
disease and point out that a dominant causal factor
did not emerge from their study of herring epizootics.
They suggest that high fish density facilitates disease
transmission and that poor condition precedes disease
expression. Laboratory experiments by Kocan et al.
(1999a, b) showed that I. hoferi prevalence can
increase with fish age without an associated increase
in mortality. This finding suggested to Kramer-
Schadt et al. (2010) that Pacific herring in good
condition appear able to endure high infection
intensities, although the fish may be at high risk if
external conditions become unfavorable. Similar to
VHSV, disease expression and morbidity caused by
I. hoferi infection may result from, and perhaps
exacerbate, the effects of poor nutrition and
decreased condition, but there is insufficient knowl-
edge to fully understand the apparent interactive
effects of these factors. Certainly, more focused
research may reveal the possible synergistic mecha-
nisms by which these factors affect herring health and
condition, but the available evidence continues to
indicate that poor nutrition preceded disease expres-
sion for both VHSV and I. hoferi during the decline.
Our re-examination of available information and
recent modeling outcomes supports earlier conclu-
sions (Pearson et al. 1999) that poor nutrition is the
probable cause of the 1993 decline. Nutritional status
of PWS herring clearly began to decline in the mid-
1980s and reached a low in 1993. Zooplankton
abundance in PWS was low before, during, and
following the decline. In recent years, other investi-
gators have come to conclude that the 1993 decline
most likely derived from poor nutrition with disease
124 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 31
as a secondary factor (Carls et al. 2002; Marty et al.
2003; Hulson et al. 2008; Rice and Carls 2007; Elston
and Meyers 2009).
Recovery
We find no evidence that oil spill effects or oil
exposure, harvest effects, spawning habitat loss, or
VHSV caused or contributed to poor recovery of
PWS herring after 1993 (Table 3). Oceanic factors,
poor nutrition, and interactions with hatchery releases
of juvenile pink salmon are supported by the
evidence. Available information about disease from
I. hoferi and predation by overwintering whales
suggest roles for these factors in limiting recovery but
more information on these potential factors is needed
to confirm them as causal.
We attribute the lack of strong year classes in both
PWS and Sitka principally to oceanic factors oper-
ating at the scale of the GOA, and the nutritional state
of the adult PWS herring remains poorer than that at
Sitka and probably influences biomass levels of adult
herring in PWS. The condition of adult herring in
PWS would have been expected to improve under the
low biomass levels after the decline. Some factor
specifically related to PWS appears to be acting
beyond the regional-scale oceanic factors.
ASA modeling clearly and strongly indicates
600-million-per-year juvenile pink salmon releases
as a substantial factor affecting PWS herring recruit-
ment. Evidence presented above supports the position
that juvenile pink salmon releases affect juvenile
herring through predation on age-0 herring and food
competition with the age-1 herring.
Predation by juvenile pink salmon on age-0?
herring may be intense in mid-summer. Because age-
0 herring are the most abundant forage fish in PWS
waters (Okey and Pauly 1999), a substantial fraction
of juvenile pink salmon predation likely occurs on
herring. Also, the extent to which avoidance of such
predation limits feeding by age-0 herring is not
known. More specific identification of the fish found
in the PWS juvenile pink salmon stomachs by
Sturdevant (1999) would be useful. The hypothesis
that juvenile pink salmon predation substantially
influences age-0 herring survival has supporting
evidence and no contrary evidence.
Our findings that juvenile pink salmon and age-1
herring probably compete for limiting food resources
is in contrast with some other published reports on
the interactions between these species. Cooney
(1993) studied the influence of the release of
hatchery-reared pink salmon fry on the carrying
capacity of PWS and concluded that the release of
hatchery-reared fry are unlikely to have much impact
on zooplankton populations. Hence, Cooney (1993)
did not expect that the fry releases would cause
competition with other forage-fish species. However,
his estimates were based on the total production (as
grams of carbon) of broad trophic classes of organ-
isms, such as zooplankton or planktivores. This
simple approach assumes that all organisms within
a given trophic class are potential prey for the next
trophic level, regardless of size, energy content, or
the ease of capture. This approach also does not
include the potential for differences in prey quality or
the selection of high-energy-content prey, which
studies indicate have a strong bearing on growth
and survival of PWS juvenile herring (Paul and Paul
1998; Cooney et al. 2001; Norcross et al. 2001).
If PWS adult herring condition has been declining
since the mid-1980s, how does the strong PWS
1988 year class reconcile with oceanic factors and
juvenile pink salmon releases as impediments to
recovery? Pink salmon interactions are a combination
of predation by pink salmon juveniles on age-0
herring and competition of pink salmon juveniles
with age-1 herring. The 1984 and 1988 year classes
were 1 year old in 1985 and 1989, respectively,
which were two years of the highest zooplankton
abundance (Fig. 4). The pink salmon fry releases first
reached 600 million in 1989. There are no years of
high zooplankton abundance in the years following
1989 (Fig. 4).
There is evidence for food competition between
juvenile herring and juvenile pink salmon in PWS.
Age-0 herring and pink salmon fry overlap in the
nearshore areas of PWS from July through the fall.
Age-1 herring and juvenile pink salmon overlap in
PWS nearshore areas from May through the fall. There
is sufficient overlap in the preferred food items among
the juvenile forage fishes in PWS to indicate food
competition, especially for prey of high-energy con-
tent. The field studies of Sturdevant (1999) found
statistically significant and substantial decreases in
four feeding indices of juvenile herring when they co-
occur with juvenile pink salmon. Poor nutrition in
juvenile herring, which could result from competition
Rev Fish Biol Fisheries (2012) 22:95–135 125
123
Page 32
with pink salmon, is related to growth, survival, and
recruitment (Foy and Norcross 2001; Norcross et al.
2001; Brown and Norcross 2001; Norcross and Brown
2001). Because juvenile herring need sufficient ener-
getic reserves to survive the winter, decreased nutri-
tion from competition in the spring and summer,
followed by a poor prey field in the fall, likely would
seriously compromise overwintering ability and, in
turn, PWS herring year-class strength.
Marty et al. (2010) posit that I. hoferi affects both
recruitment and biomass in PWS herring, but a
mechanism for the effect on recruitment was not
offered and is not clear. The potential for a biolog-
ically plausible mechanism for I. hoferi to produce
low biomass appears to be greater. Poor condition in
Atlantic herring was associated with severe I. hoferi
infections (Kramer-Schadt et al. 2010). Also, I. hoferi
affects older fish and a loss of older herring could
decrease the biomass. But as discussed previously,
poor nutrition appears to precede disease expression.
Marty et al. (2010) used ASA modeling with two
disease covariates to assess the relative roles of
VHSV and I. hoferi in the decline and poor recovery
and concluded that both VHSV and I. hoferi had a
substantial influence on the population dynamics of
PWS herringreducing the stock biomass to levels
about three times below those that otherwise might
have been. Deriso et al. (2008) also used ASA
modeling but with 19 covariates and found some
support for both VHSV and I. hoferi using the chi-
square test but not when using the more accurate
randomization test. Interactions with hatchery-release
juvenile pink salmon proved to be the most consistent
and strongest influence on PWS herring population
dynamics (Deriso et al. 2008). Marty et al. (2010)
does not include the hatchery releases in their ASA
model so that the assessment of the higher relative
influence of juvenile pink salmon from hatchery
releases versus disease remains that of Deriso et al.
(2008).
The impact of I. hoferi infection on herring
populations is well documented regarding Pacific
herring (Hershberger et al. 2006), and the ASA
modeling of Deriso et al. (2008) (Table 3) and
Hulson et al. (2008) indicate that I. hoferi could be
influencing the population dynamics of PWS her-
ring. The observed prevalence patterns of I. hoferi
in PWS herring through 2006 (Table 5) suggest
that the pathogen could be limiting survival of
herring, particularly in the older year classes
(Hershberger et al. 2002). Given the poorly under-
stood relationship between infection intensity and
prevalence on population mortality risk, we con-
clude that these factors need to be more fully
understood before accurate conclusions and predic-
tions relating I. hoferi infection to population effect
can be made.
The influence of I. hoferi on PWS herring
recruitment is not supported, but some secondary
role of I. hoferi in producing low biomass in PWS
herring cannot be dismissed. The increase in prev-
alence of I. hoferi after 2000 offers some evidence
of a role and the outcomes of the more recent ASA
models posit this hypothesis. The loss of older fish
to disease from I. hoferi offers a plausible mecha-
nism for lower biomass. Experimental examination
of the disease process to determine whether poor
condition is the precursor to disease related mor-
bidity and mortality is needed before we can more
definitively assess whether there is a causal role for
I. hoferi in the low post-decline biomass levels in
PWS herring.
Whale predation appears to be reducing adult
biomass rather than the sub-adult biomass. Summer
feeding by humpback whales appears to be primarily
on euphausiids. Whale predation during the summer
is on the 2-year-old herring because age 0 and age 1
herring are in the nursery areas and the adults are in
the GOA. Our estimates indicate that summering
whales may be taking less than 24% of the PWS
recent sub-adult herring population. Because there
are no estimates of the age-2 biomass or numbers, we
used the age-3 abundance in our estimate. Conse-
quently, a more realistic estimate from the available
information would be that summering whales remove
substantially less than 24% of the sub-adult herring
population.
Our estimates are sensitive to the daily consump-
tion rate, the proportion of herring in the diet and
the number of whales. Our estimates for the
overwintering period are especially sensitive to the
high number of whales (112) observed in October
and by the preliminary diet assumption that 80% of
the overwintering whale diet is herring. This preda-
tion falls on adult herring overwintering in PWS.
Because of the wide range of potential influence that
whale predation may have on adult herring mortal-
ity, we recommend assessments of overwintering
126 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 33
whale abundance and diet over several years and
some studies to confirm the summer diet.
From ongoing studies, overwintering whales now
appear to be in greater abundance than previously
supposed and appear to feed on herring in the fall and
winter. Our estimates indicate that overwintering
whales may be taking about 26% of the recent levels
for the adult population. Preliminary estimates from
ongoing bioenergetics modeling suggest that over-
wintering whales take 18% to 23% of the pre-winter
PWS herring biomass (Rice 2008). Whereas these
estimated levels of whale predation are approxi-
mately equivalent to a 20% commercial harvest and
worthy of concern, the wide range of estimates, the
variety of assumptions, and the preliminary nature of
recent observations make it premature to declare that
a classic ‘‘predator pit’’ from whale predation is the
cause of poor recovery in PWS herring.
Furthermore, whale predation cannot be the sole
factor impairing recovery in PWS herring. Condition
and weight at age has not rebounded in PWS.
Enhanced growth under lower population levels
would have occurred if whale predation was remov-
ing large portions of the sub-adult and adult popu-
lation and other factors remained the same. Perhaps,
overwintering whales take not only herring but also
consume large amounts of the euphausiid prey also
needed for growth and survival of overwintering
adult herring. Whereas whale predation appears to be
a credible hypothesis for contributing to low biomass,
how whale predation influences recruitment and lack
of strong year classes remains less clear. We need to
examine other factors in PWS that would affect
recruitment.
The recovery hypotheses remaining are not mutu-
ally exclusive and could be acting on different
aspects of herring population dynamics at the same
time. A multifactor hypothesis would include the
following elements. First, regional-scale oceanic
factors set the limits for the prey field and, in turn,
the growth and survival of PWS herring. Second,
interactions with pink salmon juveniles released from
the hatcheries reduce age 0 herring survival through
predation and reduce age 1 growth and survival
through food competition and feeding disruption.
Third, disease as a secondary response following poor
nutrition could reduce biomass levels. Fourth, over-
wintering whales could remove large portions of
adult herring biomass through predation.
Implications for management and restoration
To be most effective, any management or restoration
efforts should address the mechanism behind the
decline and lack of recovery. If the mechanisms are
still operational, then restoration efforts that enhance
the population before any bottleneck will be wasted
effort. For example, because spawning habitat is not
limiting in PWS, enhancing spawning habitat is
unlikely to improve herring recovery.
When oceanic factors are influencing recovery,
then little can be done other than to manage the
commercial herring harvest. It is important to
remember that several important herring stocks took
decades to recover even under severe fisheries
management efforts. Recovery of PWS herring could
well depend on a future change in the conditions of
the marine environment and, especially, the prey
field.
Similarly, how to influence recovery if disease is
either a primary or secondary casual factor is elusive.
Perhaps all the fisheries manager can do is to
incorporate the disease variables, along with all other
available and relevant variables, into an ASA model
to predict the spawning biomass more accurately. If
the disease variables are a proxy for a population
showing the effects of an environmental stressor
rather than a causal factor, their incorporation in the
ASA models may still be of practical use in fisheries
management. Marty et al. (2010) cites recent
instances of forecasts that prevented unnecessary
mobilization of the herring fishing fleet. The available
information indicates that closer attention to sam-
pling design and the use of the improved techniques
to measure I. hoferi need to be incorporated into any
field efforts to monitor infectious diseases in PWS
herring.
Whale predation may prove to be removing large
amounts of PWS herring, but protected status pre-
cludes manipulating whale populations or individuals.
The only viable remedy to any effects of whale
predation may still be to manage the commercial
herring harvest with improved forecasts of the influ-
ence of whale predation. Tjelmeland and Lindstrom
(2005) found that including minke whale predation in a
stock assessment model of Barents Sea herring was
both feasible and beneficial in improving stock
assessments to support management decisions. Adding
whale predation to PWS assessment models will
Rev Fish Biol Fisheries (2012) 22:95–135 127
123
Page 34
require ongoing studies of whale abundance and
predation to be completed over a sufficient number of
years to provide reliable long-term information.
Improving nutrition among overwintering juvenile
herring clearly would be beneficial to PWS herring
recruitment, but how to do so on a scale that would be
effective is not immediately obvious. One approach
may be to switch the timing of hatchery pink salmon
fry release from the peak of plankton production to
smaller releases over a longer time period. Reduction
in the total amount of releases may also be effective
in preventing juvenile pink salmon from depleting
any zooplankton stocks that also benefit juvenile
herring. If ocean survival of juvenile and adult
salmon is the main factor governing the magnitude
of returns, then an evenly timed release or even a
reduced total release may not have a great influence
on the overall returns of adult pink salmon to PWS.
In their review of the lessons learned from the
collapse and recovery of North Sea herring, Dickey-
Collas et al. (2010) call for not only more research on
the mechanisms for collapse and recovery but also for
a move away from single-species management
approaches to multispecies management and, in the
long term, to ecosystem management. When com-
mitments to the management of several species are in
conflict (which appears to be the case with PWS),
reconciling the conflicting commitments will be best
done at the ecosystem level (Dickey-Collas et al.
2010).
Conclusions
We have assessed the evidence for and against the
principal decline and poor recovery hypotheses and
find no evidence that oil exposure from the Exxon
Valdez oil spill, harvest effects, spawning habitat
loss, the spawn-on-kelp fishery, or disease have led to
either the decline or poor recovery of PWS herring.
Our re-examination of available information and
recent modeling outcomes supports earlier conclu-
sions (Pearson et al. 1999) that poor nutrition is the
probable cause of the 1993 decline. Nutritional status
of PWS herring clearly began to decline in the mid-
1980s and reached a low in 1993 and was associated
with low zooplankton abundance. Recent informa-
tion, especially the disease analysis of Elston and
Meyers (2009), decreases the role of disease in the
decline. The evidence supports the contention that
disease during the decline was a secondary response
after a portion of the PWS herring population was
stressed by poor nutrition.
Poor recovery probably results from several fac-
tors. Since 1993, no strong year classes have emerged
in the GOA herring populations, including that at
PWS. This lack of strong year classes appears to
derive from regional-scale ocean environmental fac-
tors. Beyond the regional-scale factors, two other
factors specific to PWS appear to be reducing herring
biomass and recruitment. First, predation by an
increasing number of overwintering humpback
whales may prove to be removing a substantial
proportion of the adult herring in PWS. Second,
interactions with juvenile pink salmon released from
PWS hatcheries may be influencing nutrition in
juvenile herring and their subsequent growth, sur-
vival, and recruitment. Continued research is recom-
mended on PWS humpback whale predation and on
herring recruitment processes. Also recommended is
a move away from single-species management to
multi-species management, if not to the management
of PWS as a whole ecosystem.
Acknowledgments We thank Michael Anderson, Michael
Cobb, Lee Miller, Ann Skillman, Kathryn Sobocinski, and
John Southard of Battelle Marine Sciences Laboratory in
Sequim, Washington, and Karen Humphrey of Aquatechnics,
for their dedicated efforts in compiling and presenting the
information in this review. We thank Mark N. Maunder of
Quantitative Resource Assessment LLC for his modeling aid in
this effort. We thank Lawrence L. Moulton of MJM Research
for sharing his raw data on the co-occurrence of pink salmon
and herring in his tow net samples. We thank Kenneth A. Rose
of Louisiana State University and John R. Skalski of the
University of Washington for valuable discussions and for
reviewing an earlier version of this paper. We also thank
Douglas E. Hay of Nearshore Consulting for reviewing earlier
manuscripts. Exxon Mobil Corporation supported this effort.
Open Access This article is distributed under the terms of the
Creative Commons Attribution Noncommercial License which
permits any noncommercial use, distribution, and reproduction
in any medium, provided the original author(s) and source are
credited.
Appendix: Sources for Unpublished Data
from Alaska Department of Fish and Game
See Table 13.
128 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 35
Table 13 Herring fisheries data: Alaska Department of Fish and Game, Division of Commercial Fisheries
Year ADF&G report and link
1996 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/96catch.php
1997 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/97catch.php
1998 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/98catch.php
1998 Alaska herring roe fishery opening dates [1980–1998]
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/fishdate.php
1999 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/99catch.php
2000 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/00catch.php
2001 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/01catch.php
2002 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/02catch.php
2002 Craig/Klawock herring spawn-on-kelp pound fishery summary
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=1594
2002 Southeast Alaska winter food and bait herring fisheries
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=1936
2003 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/03catch.php
2003 Craig/Klawock herring spawn-on-kelp pound fishery summary
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=2150
2003 Prince William Sound herring fishery updates
http://www.cf.adfg.state.ak.us/region2/finfish/herring/pws/pwsupd03.php
2003 Sitka Sound herring fishery update
http://www.cf.adfg.state.ak.us/region1/finfish/herring/03sitka.php
2003 Southeast Alaska bait pound herring fisheries
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=2545
2004 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/04catch.php
2004 Craig/Klawock herring spawn-on-kelp pound fishery summary
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=2865
2004 Prince William Sound herring fishery updates
http://www.cf.adfg.state.ak.us/region2/finfish/herring/pws/pwsupd04.php
2004 Sitka Sound herring update #15
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=2828
2004 Southeast Alaska bait pound herring fisheries
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=2973
2005 Alaska commercial herring sac roe harvests and ex-vessel values
http://www.cf.adfg.state.ak.us/geninfo/finfish/herring/catchval/05catch.php
2005 Alaska peninsula–Aleutian Islands herring food and bait fishery update
http://www.cf.adfg.state.ak.us/region4/finfish/herring/05akpen_f&b.php
Rev Fish Biol Fisheries (2012) 22:95–135 129
123
Page 36
References
Angliss RP, Outlaw RB (2005) Killer whale (Orcinus orca):
eastern north Pacific northern resident stock. In: Alaska
marine mammal stock assessments, National Oceanic and
Atmospheric Administration Technical Memorandum
NMFS-AFSC-161, Seattle, pp 95–99
Armstrong JL, Boldt JL, Cross AD et al (2005) Distribution,
size, and interannual, seasonal and diel food habits of
northern Gulf of Alaska juvenile pink salmon, On-corhynchus gorbuscha. Deep-Sea Res Pt II 52:247–265
Ashe D, Gray D, Lewis B et al. (2005) Prince William Sound
management area 2004 annual finfish management report.
Fishery Management Report 05–65, Alaska Department
of Fish and Game, Anchorage, http://www.sf.adfg.
state.ak.us/FedAidPDFs/fmr05-65.pdf
Avendano-Herrera R, Toranzo AE, Magarinos B (2006)
Tenacibaculosis infection in marine fish caused by
Tenacibaculum maritinum: a review. Dis Aquat Organ 71:
255–266
Biggs ED, Haley BE, Gilman JM (1992) Historic database for
Pacific herring in Prince William Sound, Alaska,
1973–1991. Regional Information Report 2C91-11, Alaska
Department of Fish and Game, Anchorage
Blaxter JHS (1985) The herring: a successful species? Can J
Fish Aquat Sci 42(Suppl 1):21–30
Boehm PD, Neff JM, Page DS (2007) Assessment of hydro-
carbon exposure in the waters of Prince William Sound
after the Exxon Valdez oil spill: 1989–2005. Mar Pollut
Bull 54:339–367
Boehm PD, Page DS, Brown JS et al (2004) Polycyclic aro-
matic hydrocarbon levels in mussels from Prince William
Sound, Alaska, USA, document the return to baseline
conditions. Environ Toxicol Chem 23(12):2916–2929
Boldt JL, Haldorson LJ (2003) Seasonal and geographic vari-
ation in juvenile pink salmon diets in the northern Gulf of
Alaska and Prince William Sound. T Am Fish Soc 132:
1035–1052
Brady J, Morstad S, Simpson E et al (1991a) Prince William
Sound management area annual finfish management
report 1989. Regional Information Report 2C90-07,
Alaska Department of Fish and Game, Cordova
Brady J, Morstad S, Simpson E et al (1991b) Prince William
Sound management area 1990 annual finfish manage-
ment report. Alaska Department of Fish and Game,
regional information report No.2C91-14, Anchorage,
Alaska
Brady JA, Schultz K, Simpson E et al (1990) Prince William
Sound area annual finfish management report 1988.
Regional Information Report 2C90-02, Alaska Depart-
ment of Fish and Game, Cordova
Brady J, Sharr S, Roberson K et al (1987) Prince William
Sound area annual finfish management report 1986.
Alaska Department of Fish and Game, Cordova
Brady JA, Sharr S, Roberson K et al (1988) Prince William
Sound area annual finfish management report 1987.
Alaska Department of Fish and Game, Cordova
Brown ED (2004) Stock structure and environmental effects on
year class formation and population trends of Pacific
herring, Clupea pallasi, in Prince William Sound, Alaska.
Dissertation, University of Alaska, Fairbanks
Brown ED, Baker TT, Funk F et al (1994) The Exxon Valdez oil
spill and Pacific herring in Prince William Sound, Alaska: a
summary of injury from 1989–1994. In: Brown ED, Baker
TT (eds) Injury to Prince William Sound following the
Exxon Valdez oil. Final Report Fish/Shellfish Study Num-
ber 11. Alaska Department of Fish and Game, Anchorage
Brown ED, Norcross BL (2001) Effect of herring egg distri-
bution and environmental factors on year-class strength
and adult distribution: preliminary results from Prince
William Sound, Alaska. In: Funk F, Blackburn J, Hay D
et al (eds) Proceedings of the 18th Lowell Wakefield
fisheries symposium, Herring 2000: expectations for a
new millennium, Anchorage, AK, February 23–26, 2000,
AK-SG-01–04. University of Alaska Sea Grant College
Program, Fairbanks, pp 335–345
Table 13 continued
Year ADF&G report and link
2005 Prince William Sound herring fishery updates
http://www.cf.adfg.state.ak.us/region2/finfish/herring/pws/pwsupd05.php
2005 Sitka Sound herring update #12
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=3615
2005 Southeast Alaska winter food and bait herring fisheries
http://www.documents.cf1.adfg.state.ak.us/AdfgDocument.po?DOCUMENT=3471
2006 Herring age, weight, and length data for Kodiak herring, 1976, 1980–2005
Obtained from the Division of Commercial Fisheries, Juneau, Alaska
2006 Herring age, weight, and length data for Prince William Sound (1993, 1994, 1996–2005) and Sitka (1976–2005)
Obtained from the Division of Commercial Fisheries, Juneau, Alaska
2006 Herring age, weight, and length (1971–1999);
biomass (1978–2005), and age-class composition (1978–2005) data for Lower Cook Inlet
Obtained from the Division of Commercial Fisheries, Homer, Alaska
130 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 37
Brown ED, Norcross BL, Short JW (1996) An introduction to
studies on the effects of the Exxon Valdez oil spill on early
life history stages of Pacific herring, Clupea pallasi, in
Prince William Sound, Alaska. Can J Fish Aquat Sci
53:2337–2342
Brown ED, Seitz J, Norcross BL et al (2002) Ecology of her-
ring and other forage fish as recorded by resource users of
Prince William Sound and the outer Kenai Peninsula,
Alaska. Alsk Fish Res Bull 9(2):75–101
Bruce G (2006) Herring data from LCI, PWS, and Sitka.
Alaska Department of Fish and Game, Division of Com-
mercial Fisheries, Juneau, Alaska
Calambokidis J, Steiger GH (2001) Movements and population
structure of humpback whales in the north Pacific. Mar
Mamm Sci 17(4):769–794
Cardinale M, Casini M, Arrhenius F et al (2003) Diel and
spatial distribution and feeding activity of herring (Clupeaharengus) and sprat (Sprattus sprattus) in the Baltic Sea.
Aquat Living Resour 16:283–292
Carls MG, Marty GD, Hose JE (2002) Synthesis of the toxi-
cological impacts of the Exxon Valdez oil spill on Pacific
herring (Clupea pallasi) in Prince William Sound, Alaska,
USA. Can J Fish Aquat Sci 59(1):153–172
Carls MG, Marty GD, Meyers TR et al (1998) Expression of
viral hemorrhagic septicemia virus in prespawning Pacific
herring (Clupea pallasi) exposed to weathered crude oil.
Can J Fish Aquat Sci 55(10):2300–2309
Carls MG, Rice SD, Hose JE (1999) Sensitivity of fish embryos
to weathered crude oil: part I: low-level exposure during
incubation causes malformations, genetic damage, and
mortality in larval Pacific herring (Clupea pallasi).Environ Toxicol Chem 18(3):491–493
Collie JS (1991) Herring population dynamics and manage-
ment in Sitka Sound, Alaska. In: Wespestad V, Collie J,
Collie E (eds) Lowell Wakefield fisheries symposium:
proceedings of the international herring symposium,
Anchorage, AK, October 23–25, 1990, AK-SG-91–01.
University of Alaska Sea Grant College Program, Fair-
banks, pp 463–476
Cooney RT (1993) A theoretical evaluation of the carrying
capacity of Prince William Sound, Alaska, for juvenile
Pacific salmon. Fish Res 18:77–87
Cooney RT, Allen JR, Bishop MA et al (2001) Ecosystem
controls of juvenile pink salmon (Oncorhynchus gorbus-cha) and Pacific herring (Clupea pallasi) populations in
Prince William Sound, Alaska, fisheries. Fish Oceanogr
10(Suppl 1):1–13
Cross AD, Beauchamp DA, Armstrong JL et al (2005) Con-
sumption demand of juvenile pink salmon in Prince
William Sound and the coastal Gulf of Alaska in relation
to prey biomass. Deep-Sea Res Pt II 52:347–370
Deriso RB, Maunder MN, Pearson WH (2008) Incorporating
covariates into fisheries stock assessment models: Salmon
hatcheries inhibit herring recovery. Ecol Applications
18(5):1270–1286
Deriso RB, Maunder MN, Skalski JR (2007) Variance estimation
in integrated assessment models and its importance for
hypothesis testing. Can J Fish Aquat Sci 64(2):187–197
Dickey-Collas M, Nash RDM, Brunel T et al (2010) Lessons
learned from stock collapse and recovery of North Sea
herring: a review. ICES J Mar Sci 67:1875–1886
Dillingham P, Ryding K, Skalski JR (2004) Assessment of
fine-scaled interactions between Steller sea lions and
trends of local fisheries. Final report to National Ocean-
ographic and Atmospheric Administration by School of
Fisheries, University of Washington, Seattle
Donaldson W, Morstad S, Sharp D et al (1995) Prince William
Sound management area 1993 annual finfish management
report. Regional Information Report 2A95-20, Alaska
Department of Fish and Game, Anchorage
Donaldson W, Morstad S, Simpson E et al (1992) Prince
William Sound management area 1991 annual finfish
management report. Regional Information Report 2A92-
09, Alaska Department of Fish and Game, Anchorage
Donaldson W, Morstad S, Simpson E et al (1993) Prince
William Sound management area 1992 annual finfish
management report. Regional Information Report 2A93-
12, Alaska Department of Fish and Game, Anchorage
Dressel S (2006) ADF&G age-structured assessment Excel
spreadsheet program for Sitka Sound herring. Alaska
Department of Fish and Game, Division of Commercial
Fisheries, Douglas
Dunn BP, Stich HE (1976) Monitoring procedures for chemical
carcinogens in coastal waters. J Fish Res Bd Canada
33(9):2040–2046
Dunn BP, Young DR (1976) Baseline levels of benzo(a)pyrene
in southern California mussels. Mar Pollut Bull 7(12):
231–234
Elston RA, Drum AS, Pearson WH et al (1997) Health and con-
dition of Pacific herring Clupea pallasi from Prince William
Sound, Alaska, 1994. Dis Aquat Organ 31:109–126
Elston RA, Meyers TR (2009) Effect of viral hemorrhagic sep-
ticemia virus on Pacific herring in Prince William Sound,
Alaska, from 1989 to 2005. Dis Aquat Organ 83:223–246
Everson I (2000) Krill: biology, ecology and fisheries. Black-
well Science, Oxford, p 372
Farrington C (2004) Alaska salmon enhancement program
2003 annual report. Regional Information Report 5J04-02,
Alaska Department of Fish and Game, Juneau
Farrington JW, Quinn JG (1973) Petroleum hydrocarbons in
Narragansett Bay. I. Survey of hydrocarbons in sediments
and clams (Mercenaria mercenaria). Estuar Coast Mar S
1:71–79
Foy RJ, Norcross BL (1999) Spatial and temporal variability
in the diet of juvenile Pacific herring (Clupea pallasi)in Prince William Sound, Alaska. Can J Zool 77:
697–706
Foy RJ, Norcross BL (2001) Temperature effects on zoo-
plankton assemblages and juvenile herring feeding in
Prince William Sound, Alaska. In: Funk F, Blackburn J,
Hay D et al (eds) Proceedings of the 18th Lowell
Wakefield fisheries symposium, herring 2000: expecta-
tions for a new millennium, Anchorage, Alaska, February
23–26, 2000, AK-SG-01–04. University of Alaska Sea
Grant College Program, Fairbanks, pp 21–35
Funk F (1993) Preliminary forecasts of catch and abundance
for the 1993 Alaska herring fisheries. Regional Info Rep
No 5J93-06, Alaska Department of Fish and Game,
Juneau
Funk F (1994) Forecast of the Pacific herring biomass in Prince
William Sound, Alaska. Regional Information Report
5J94-04, Alaska Department of Fish and Game, Juneau
Rev Fish Biol Fisheries (2012) 22:95–135 131
123
Page 38
Funk F, Sandone G (1990) Catch-age analysis of Prince Wil-
liam Sound, Alaska, herring, 1973–1988. Fish Res Bull
90-01, Alaska Department of Fish and Game, Juneau
Hansen PC, Johnson TB, Schindler DE et al (1997) Fish bio-
energetics 3.0 for windows. Center for Limnology, Uni-
versity of Wisconsin, Madison
Hay DE, Toresen R, Stephenson R et al (2001) Taking stock: an
inventory and review of world herring stocks in 2000. In:
Funk F, Blackburn J, Hay D et al (eds) Proceedings of the
18th Lowell Wakefield fisheries symposium, herring 2000:
expectations for a new millennium, Anchorage, Alaska,
February 23–26, 2000, AK-SG-01–04. University of
Alaska Sea Grant College Program, Fairbanks, pp 381–454
Hedrick RP, Batts WN, Yun S et al (2003) Host and geographic
range extensions of the North American strain of viral
hemorrhagic septicemia virus. Dis Aquat Organ 55:
211–220
Hershberger PK, Gregg J, Pacheco C et al (2007) Larval Pacific
herring, Clupea pallasi (Valenciennes), are highly sus-
ceptible to viral haemorrhagic septicaemia and survivors
are partially protected after their metamorphosis to juve-
niles. J Fish Dis 30:445–458
Hershberger P, Hart A, Gregg J et al (2006) Dynamics of viral
hemorrhagic septicemia, viral erythrocytic necrosis, and
ichthyophoniasis in confined juvenile Pacific herring
Clupea pallasi. Dis Aquat Org 70:201–208
Hershberger PK, Stick K, Bui B et al (2002) Incidence of
Ichthyophonus hoferi in Puget Sound fishes and its
increase with age of Pacific herring. J Aquat Anim Health
14:50–56
Hilborn R (1997) The frequency and severity of fish stock
declines and increases. In: Hancock DA, Smith DA, Grant
DC, Beumer A et al (eds) Developing and sustaining
world fisheries resources, the state of the science and
management, proceedings of the 2nd world fisheries
congress, Brisbane, Australia, July 28–August 2, 1996.
CSIRO Publishing, Collingwood, pp 36–38
Hjort J (1914) Fluctuations in the great fisheries of the
Northern Europe viewed in the light of the biological
research. Cons Perm Int Explor Mer Rapp et Proc-Verb
Reun 20:1–228
Hjort J, Ruud JT (1929) Whaling and fishing in the North
Atlantic. Cons Perm Int Explor Mer Rapp et Proc Verb
Reun 56(Paper 1):1–123
Holst JC (1996) Estimating the prevalence of Ichthyophonushoferi (Plehn and Mulsow) in a herring stock (Clupeaharengus L.): observed effects of sampling gear, target
school density and migration. Fish Res 28:85–97
Hulson P-JF, Miller SE, Quinn TJ II et al (2008) Data conflicts
in fishery models: incorporating hydroacoustic data into
the Prince William Sound Pacific herring assessment
model. ICES J Mar Sci 65:25–43
Johnson SW, Carls MG, Stone RP et al (1997) Reproductive
success of Pacific herring, Clupea pallasi, in Prince
William Sound, Alaska, six years after the Exxon Valdezoil spill. Fish Bull 95(4):748–761
Johnston DW, Wildish DJ (1981) Avoidance of dredge spoil by
herring (Clupea harengus harengus). Bull Environ Con-
tam Toxicol 26:307–314
Kennedy CJ, Farrell AP, Sanders S (1999) Survival, perfor-
mance and reproduction in the pacific herring, Clupea
harengus pallasi: effects of environmental contamination,
viral hemorrhagic septicemia virus and Ichthyophonushoferi. In: Kocan RM, Marty GD, Kennedy CJ (eds)
Investigations of disease factors affecting declines of
Pacific herring populations in Prince William Sound,
Final Report, Restoration Project 97162, Exxon Valdez Oil
Spill Trustee Council, Anchorage, pp III-1–III-40, http://
www.evostc.state.ak.us/Store/FinalReports/1998-98162-
Final.pdf
Kocan RM, Hershberger PK, Elder N et al (2001) Epidemiol-
ogy of viral hemorrhagic septicemia among juvenile
Pacific herring and Pacific sand lances in Puget, Sound,
Washington. J Aquat Anim Health 13:77–85
Kocan RM, Hershberger PK, Mehl P et al (1999a) Pathoge-
nicity of Ichthyophonus hoferi for laboratory-reared
Pacific herring Clupea pallasi and its early appearance in
wild Puget Sound herring. Dis Aquat Organ 35:23–29
Kocan RM, Hershberger PK, Winton JR et al (1999b) Con-
trolled field and laboratory studies on VHS and Ichthy-ophonus in Pacific herring. In: Kocan RM, Marty GD,
Kennedy CJ (eds) Investigations of disease factors
affecting declines of Pacific herring populations in Prince
William Sound, Final Report, Restoration Project 97162,
Exxon Valdez Oil Spill Trustee Council, Anchorage, pp II-
i–II-73, http://www.evostc.state.ak.us/Store/FinalReports/
1998-98162-Final.pdf
Kocan RM, Dolan H, Hershberger P (2011) Diagnostic meth-
odology is critical for accurately determining the preva-
lence of Ichthyophonus infections in wild fish populations.
J Parasit (in press)
Kramer-Schadt S, Holst JC, Skaen D (2010) Analysis of
variables associated with the Ichthyophonus hoferi epi-
zootics in Norwegian spring-spawning herring, 1992–2008.
Can J Fish Aquat Sci 67:1862–1973
Marty GD, Freiberg EF, Meyers TR et al (1998) Viral hem-
orrhagic septicemia virus, Ichthyophonus hoferi, and other
causes of morbidity in Pacific herring, Clupea pallasi,spawning in Prince William Sound, Alaska, USA. Dis
Aquat Organ 32:15–40
Marty GD, Quinn TJ II, Carpenter G et al (2003) Role of
disease in abundance of a Pacific herring (Clupea pallasi)population. Can J Fish Aquat Sci 60:1258–1268
Marty GD, Quinn TJ, Miller SE et al (2004) Effect of disease
on recovery of Pacific herring in Prince William Sound,
Alaska. Restoration Project 030462, Final Report, ExxonValdez Oil Spill Trustee Council, Anchorage, http://www.
evostc.state.ak.us/Store/FinalReports/2003-030462-Final.
pdf
Marty GD (2007) Evidence that the Exxon Valdez oil spill did
not cause the 1993 disease epidemic in the Pacific herring
population of Prince William Sound Alaska. Chapter 2.
In: Rice SD, Carls MG Prince William Sound Herring: An
updated synthesis of population declines and lack of
recovery. Exxon Valdez Trustee Council Restoration
Project 050794, Final Report
Marty GD, Miller SE, Hulson P-JF, Quinn TJII, Moffitt SD,
Merion RA, Meyers TM. (2007) Role of Ichthyophonushoferi, viral hemorrhagic septicemia virus, and cutaneous
ulcers in preventing recover of a Pacific herring (Clupeapallasi) population. Chapter 5, in: Rice SD and Carls MG
Prince William Sound Herring: An updated synthesis of
132 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 39
population declines and lack of recovery. Exxon Valdez
Trustee Council Restoration Project 050794, Final Report
Marty GD, Hulson P, Miller SE et al (2010) Failure of popu-
lation recovery in relation to disease in Pacific herring.
Dis Aquat Org 90:1–40
Matkin C, Hobbs R (1999) Mammals: baleen whales. In: Okey
TA, Pauly D (eds) Trophic mass-balance model of Alas-
ka’s Prince William Sound ecosystem, for the post-spill
period 1994–1996. Final Report, Restoration Project
99330-1, Exxon Valdez Oil Spill Trustee Council,
Anchorage, p 56, http://www.evostc.state.ak.us/Store/
FinalReports/1999-99330BAA-Final.pdf
McQuinn IH (1997) Metapopulations and the Atlantic herring.
Rev Fish Biol Fish 7:297–329
Meyers TR, Short S, Lipson K et al (1994) Association of viral
hemorrhagic septicemia virus with epizootic hemorrhages
in the skin of Pacific herring Clupea harengus pallasifrom Prince William Sound and Kodiak, Alaska, USA.
Dis Aquat Organ 19:27–37
Meyers TR, Short S, Lipson K (1999) Isolation of the North
American strain of viral hemorrhagic septicemia virus
(VHSV) associated with epizootic mortality in two new
host species of Alaska marine fish. Dis Aquat Organ 38:
81–86
Moffitt S (2006) ADF&G age-structured assessment Excel
spreadsheet program and data for Prince William Sound
herring. Alaska Department of Fish and Game, Division
of Commercial Fisheries, Cordova
Moulton LL (1997) Early marine residence, growth, and
feeding by juvenile salmon in northern Cook Inlet,
Alaska. Alaska Fish Res Bull 4(2):154–177
Neff JM, Bence AE, Parker KR et al (2006) Bioavailability of
polycyclic aromatic hydrocarbons from buried shoreline
oil residues thirteen years after the Exxon Valdez oil spill:
a multispecies assessment. Environ Toxicol Chem 25(4):
947–961
Neff JM, Burns WA (1996) Estimation of polycyclic aromatic
hydrocarbon concentrations in the water column based on
tissue residues in mussels and salmon: an equilibrium
partitioning approach. Environ Toxicol Chem 15(12):
2240–2253
Norcross BL, Brown ED (2001) Estimation of first-year sur-
vival of Pacific herring from a review of recent stage-
specific studies. In: Funk F, Blackburn J, Hay D et al (eds)
Proceedings of the 18th Lowell Wakefield fisheries sym-
posium, herring 2000: expectations for a new millennium,
Anchorage, Alaska, February 23–26, 2000, AK-SG-01-04.
University of Alaska Sea Grant College Program, Fair-
banks, pp 535–558
Norcross BL, Brown ED, Foy RJ et al (2001) A synthesis of the
life history and ecology of juvenile Pacific herring in
Prince William Sound, Alaska. Fish Oceanogr 10(Suppl
1):42–57
Okey TA, Pauly D (1999) Trophic mass-balance model of Alas-
ka’s Prince William Sound ecosystem, for the post-spill
period 1994–1996. Final Report, Restoration Project
99330-1, Exxon Valdez Oil Spill Trustee Council, Anchor-
age, http://www.evostc.state.ak.us/Store/FinalReports/1999-
99330BAA-Final.pdf
Overholtz WJ, Link JS (2007) Consumption impacts by marine
mammals, fish, and seabirds on the Gulf of Maine—
Georges Bank herring (Clupea harengus) complex during
the years 1977–2002. ICES J Mar Sci 64:83–96
Patterson KR (1996) Modeling the impact of disease-induced
mortality in an exploited population: the outbreak of the
fungal parasite Ichthyophonus hoferi in the North Sea
herring (Clupea harengus). Can J Fish Aquat Sci 53:
2870–2887
Paul AJ, Paul JM (1998) Comparisons of whole body energy
content of captive fasting age zero Alaskan Pacific herring
(Clupea pallasi Valenciennes) and cohorts over-wintering
in nature. J Exp Mar Biol Ecol 226(1):75–86
Payne MR, Hatfield EMC, Dickey-Collas M et al (2009)
Recruitment in a changing environment: the 2000s North
Sea herring recruitment failure. ICEAS J Mar Sci 66:
272–277
Pearson WH, Elston RA, Bienert RW et al (1999) Why did the
Prince William Sound, Alaska Pacific herring (Clupeapallasi) fisheries collapse in 1993 and 1994? Review of
hypotheses. Can J Fish Aquat Sci 56:711–737
Pearson WH, Moksness E, Skalski JR (1995) A field and
laboratory assessment of oil spill effects on survival and
reproduction of Pacific herring following the Exxon
Valdez spill. In: Wells PG, Butler JN, Hughes JS (eds)
Exxon Valdez oil spill: fate and effects in Alaskan waters.
STP 1219, Publication 04-012190-16, American Society
for Testing and Materials, Philadelphia, pp 626–661
Platt JR (1964) Strong inference. Science 146:347–353
Quinn TJ II, Marty GD, Wilcock J et al (2001) Disease and
population assessment of Pacific herring in Prince Wil-
liam Sound, Alaska. In: Funk F, Blackburn J, Hay D et al
(eds) Proceedings of the 18th Lowell Wakefield fisheries
symposium, Herring 2000: expectations for a new
millennium, Anchorage, Alaska, February 23–26, 2000,
AK-SG-01-04. University of Alaska Sea Grant College
Program, Fairbanks, pp 363–379
Quinn TJ II, Deriso RB (1999) Quantitative fish dynamics.
Oxford University Press, New York
Rice S, Carls MG (2007) Executive summary. In: Rice SD, Carls
MG (eds) Prince William Sound herring: an updated syn-
thesis of population declines and lack of recovery. Exxon
Valdez Trustee Council Restoration Project 050794, Final
Report
Rice S (2008) Significance of whale predation on natural mor-
tality rate of Pacific herring in Prince William Sound.
Exxon Valdez Trustee Council Restoration Project 080804,
Progress Report
Rice S (2009) Significance of whale predation on natural mor-
tality rate of Pacific herring in Prince William Sound.
Exxon Valdez Trustee Council Restoration Project
PJ090804, Progress Report
Rose KA, Werner F, Megrey BA et al (2007) Simulated herring
growth responses in the northeastern Pacific to historic
temperature and zooplankton conditions generated by the
3-dimensional NEMURO nutrient-phytoplankton-zoo-
plankton model. Ecol Model 202:184–195
Rothschild BJ (1986) Dynamics of marine fish populations.
Harvard University Press, Cambridge
Ruggerone GT, Nielsen JL (2004) Evidence for competitive
dominance of pink salmon (Oncorhynchus gorbuscha)
over other salmonids in North Pacific Ocean. Rev Fish
Biol Fish 14:371–390
Rev Fish Biol Fisheries (2012) 22:95–135 133
123
Page 40
Sanders SM (2005) Immunotoxicology of Pacific herring: deter-
mination of reference ranges and their application to assessing
exposure to the water-soluble fraction of crude oil. Disserta-
tion, Simon Fraser University, Burnaby, British Columbia
Sandone GJ (1988) Age, sex, and size composition of Pacific
herring sampled from the Prince William Sound Manage-
ment Area, 1984–1987. Regional Information Report
2A88-08, Alaska Department of Fish and Game, Anchorage
Short JW, Harris PM (1996) Petroleum hydrocarbons in caged
mussels deployed in Prince William Sound after the
Exxon Valdez oil spill. In: Rice SD, Spiess, Rb, Wolfe
DA et al (eds) Proceedings of the Exxon Valdez oil spill
symposium. American Fisheries Society Symposium 18
Short JW, Maselko JM, Lindberg MR et al (2006) Vertical
distribution and probability of encountering intertidal
Exxon Valdez oil on shorelines of three embayments
within Prince William Sound, Alaska. Environ Sci
Technol 40(12):3723–3729
Sindermann CJ (1958) An epizootic in Gulf of Saint Lawrence
fishes. In: Trefethen JB (ed) Transactions of the twenty-
third north American wildlife conference March 3–5,
1958. St. Louis, Missouri. Wildlife Management Institute,
Washington, DC, pp 349–360
Smith TD (1988) Stock assessment methods: the first fifty
years. In: Gulland JA (ed) Fish population dynamics, the
implications for management, 2nd edn, John Wiley and
Sons, New York, ch 1
Southward AJ, Boalch GT, Maddock L (1988) Fluctuations in
the herring and pilchard fisheries of Devon and Cornwall
linked to change in climate since the 16th century. J Mar
Biol Assoc UK 68:423–445
Spies RB (2007) Long-term ecological changes in the Northern
Gulf of Alaska. Elsevier, Amsterdam, p 589
Stephenson RL (1997) Successes and failures in the manage-
ment of Atlantic herring fisheries: do we know why some
have collapsed and others survived. In: Hancock DA,
Smith DA, Grant DC, Beumer A et al (eds) Developing
and sustaining world fisheries resources, the state of the
science and management, proceedings of the 2nd world
fisheries congress, Brisbane, Australia, July 28–August 2,
1996. CSIRO Publishing, Collingwood, pp 49–54
Stephenson RL (2001) The role of herring investigations in
shaping fisheries science. In: Funk F, Blackburn J, Hay D
et al (eds) Proceedings of the 18th Lowell Wakefield
fisheries symposium, herring 2000: expectations for a new
millennium, Anchorage, Alaska, February 23–26, 2000,
AK-SG-01-04. University of Alaska Sea Grant College
Program, Fairbanks, pp 1–20
Stokesbury KDE, Foy RJ, Norcross BL (1999) Spatial and
temporal variability in juvenile Pacific herring, Clupeapallasi, growth in Prince William Sound, Alaska. Environ
Biol Fish 56:409–418
Stokesbury KDE, Kirsh J, Brown ED et al (2000) Spatial
distributions of Pacific herring, Clupea pallasi, and
walleye pollock, Theragra chalcogramma, in Prince
William Sound, Alaska. Fish Bull 98:400–409
Sturdevant MV (1999) Forage fish diet overlap, 1994–1996.
Final Report, Restoration Project 97163C, Exxon ValdezOil Spill Trustee Council, Anchorage, http://www.evostc.
state.ak.us/Store/FinalReports/1997-97163C-Final.pdf
Sturdevant MV, Brase ALJ, Hulbert LB (2001) Feeding habits,
prey fields, and potential competition of young-of-the-
year walleye pollock (Theragra chalcogramma) and
Pacific herring (Clupea pallasi) in Prince William Sound,
Alaska, 1994–1995. Fish Bull 99:482–501
Sverdrup HU, Johnson MW, Fleming RH (1942) The oceans:
their physics, chemistry, and general biology. Prentice-
Hall, Englewood Cliffs
Taft DG, Egging DE, Kuhn HA (1995) Sheen surveillance: an
environmental monitoring program subsequent to the
1989 Exxon Valdez shoreline cleanup. In: Wells PG,
Butler JN, Hughes JS (eds) Exxon Valdez oil spill: fate
and effects in Alaskan waters. STP 1219, Publication
04-012190-16, American Society for Testing and Mate-
rials, Philadelphia, pp 626–661
Thomas GL, Thorne RE (2003) Acoustical-optical assessment
of Pacific herring and the predator assemblage in Prince
William Sound, Alaska. Aquat Living Resour 16:247–253
Thorne RE, Thomas GL (2008) Herring and the ‘‘Exxon Val-
dez’’ oil spill: an investigation into historical data con-
flicts. ICES J Mar Sci 65:44–50
Tibbo SN, Graham TR (1963) Biological changes in herring
stocks following an epizootic. J Fish Res Bd Canada
20(2):435–449
Tjelmeland S, Lindstrom U (2005) An ecosystem element
added to the assessment of Norwegian spring-spawning
herring: implementing predation by minke whales. ICES J
Mar Sci 62:284–294
Tynan CT (2004) Cetacean populations on the SE Bering Sea
shelf during the late 1990s: implications for decadal
changes in ecosystem structure and carbon flow. Mar Ecol
Prog Ser 272:281–300
Von Ziegesar O, Miller E, Dahlheim ME (1994) Impact on
humpback whales in Prince William Sound. In: Loughlin
TR (ed) Marine mammals and the Exxon Valdez. Aca-
demic Press, San Diego, pp 173–191
Wertheimer AC, Celewycz AG (1996) Abundance and growth
of juvenile pink salmon in oiled and non-oiled locations of
western Prince William Sound after the Exxon Valdez oil
spill. In: Rice SD, Spies RB, Wolfe DA, et al (eds) Pro-
ceedings of the Exxon Valdez oil spill symposium,
Anchorage, Alaska, February 2–5, 1993, American Fish-
eries Society Symposium 18, Bethesda, pp 518–532
White B (2005) Alaska salmon enhancement program 2004
annual report. Fishery Management Report 05-09. Alaska
Department of Fish and Game, Anchorage, http://www.sf.
adfg.state.ak.us/FedAidPDFs/fmr05-29.pdf
Willette TM (2001) Foraging behaviour of juvenile pink sal-
mon (Oncorhynchus gorbuscha) and size-dependent pre-
dation risk. Fish Oceanogr 10(Suppl 1):110–131
Willette, TM, Cooney RT, Hyer K (1999a) Some processes
affecting mortality of juvenile fishes during the spring
bloom in Prince William Sound, Alaska. In: Ecosystem
approaches to fisheries management: proceedings of the
symposium on ecosystem considerations in fisheries
management, September 30–October 3, 1998, Anchorage,
Alaska. Alaska Sea Grant Program AK-SG-99-01, Uni-
versity of Alaska Sea Grant Program, Fairbanks
Willette TM, Cooney RT, Hyer K (1999b) Predator foraging
mode shifts affecting mortality of juvenile fishes during
134 Rev Fish Biol Fisheries (2012) 22:95–135
123
Page 41
the subarctic spring bloom. Can J Fish Aquat Sci 56(3):
364–376
Willette TM, Cooney RT, Patrick V et al (2001) Ecological
processes influencing mortality of juvenile pink salmon
(Oncorhynchus gorbuscha) in Prince William Sound,
Alaska. Fish Oceanogr 10(suppl 1):14–41
Willette TM, Sturdevant MV, Jewett S (1997) Prey resource
partitioning among several species of forage fishes in
Prince William Sound, Alaska. In: Rice SD, Spies RB,
Wolfe DA et al (eds) International fisheries symposium,
role of forage fishes in marine ecosystems, Anchorage,
Alaska, 1996, AK-SG-97-01. University of Alaska Sea
Grant College Program, Fairbanks, pp 11–29
Williams EH, Quinn TJ II (2000a) Pacific herring, Clupeapallasi, recruitment in the Bering Sea and north-east
Pacific Ocean. I: relationships among different popula-
tions. Fish Oceanogr 9(4):285–299
Williams EH, Quinn TJ II (2000b) Pacific herring, Clupeapallasi, recruitment in the Bering Sea and north-east Pacific
Ocean. II: relationships to environmental variables and
implications for forecasting. Fish Oceanogr 9(4):300–315
Winters GH, Wheeler JP (1994) Length-specific weight as a
measure of growth success of adult Atlantic herring
(Clupea harengus). Can J Fish Aquat Sci 52:1169–1179
Winters GH, Wheeler JP, Dalley EL (1986) Survival of a
herring stock subjected to a catastrophic event and fluc-
tuating environmental conditions. J Cons Int Explor Mer
43:26–42
Witteveen BH, Foy RJ, Wynne KM (2006) The effect of pre-
dation (current and historical) by humpback whales
(Megaptera noveangliae) on fish abundance near Kodiak
Island, Alaska. Fish Bull 104:10–20
Wolfe DA, Hammedi MJ, Galt JA et al (1994) The fate of the
oil spilled from the Exxon Valdez. Environ Sci Technol
28(13):561–568
Zebdi A, Collie JS (1995) Effect of climate on herring (Clupeapallasi) population dynamics in the Northeast Pacific
Ocean. In: Beamish RJ (ed) Climate change and northern
fish populations conference, Victoria, British Columbia,
October 19–24, 1992, Can Spec Publ Fish Aquat Sci
121:277–290
Rev Fish Biol Fisheries (2012) 22:95–135 135
123
