An Early Life History Model for Pacific herring (Clupea pallasii) in Prince William Sound, Alaska

1
(No Transcript)
2
A Model for Early Life History Survival for
Pacific Herring in Prince William Sound

• Brenda Norcross, Seanbob Kelly,
• Peter-John Hulson, Terry Quinn
• School of Fisheries and Ocean Sciences
• University of Alaska Fairbanks

3
Herring An Important Species

• Forage fish

• Commercial fishery

4
Prince William Sound
http//marine.alaskapacific.edu/octopus/pws-map.ph
p
5
Recent History of Herring

• Fishery closed in March 1989 following the Exxon
  Valdez oil spill
• Stock collapsed 1993 due to (VHSV)
• Species has not recovered to pre-oil spill
  abundance

6
Spawning and Larvae

• Herring spawn onshore in April
• Larvae herring are advected counter-clockwise
  through open water

7
Years 1 and 2

• June-August metamorphosis
• Nursery habitats at heads of bays
• In nursery bays for 2 winters
• Leave bays and join adult schools

8
Early Life History to Age-1

• 4 stages - eggs, larvae, fall juveniles, winter
  juveniles
• Know mortality changes as life stage changes
• Determine which life stage is most influential

9
Early Life History Model

• Life-stage specific survival to age-1
• Builds on an earlier range-based study
• (Norcross Brown 2001)
• Statistical distributions of survival to account
  for uncertainty
• Data input from published estimates

10
Standard Year-Class Model (Quinn and Deriso 1999)

• Cumulative mortality - z for each stage
  multiplied by number of days per stage
• Total mortality - combines mortalities of
  sequential life stages to age-1

,
11
Delta Method (Seber 1982)

• Converts standard error of survival to that of
  mortality
• Assumed normal distribution
• Allows determination of 95 confidence intervals

.
12
Egg Stage - first 21 days

• (Haegele 1993)
• Subsurface oophagy - crabs, sea anemones, and
  snails
• (Rooper et al. 1999)
• Duration of air exposure - exposure abiotic
  forces, and avian predation

13
Larval Stage - next 92 days

• (McGurk et al. 1993)
• Larval mortality caused by advection, predation,
  and inability to feed
• Data from Auke Bay, Southeast Alaska
• Comparable to estimates from British Columbia

14
Fall Juvenile Stage - next 92 days

• (Stokesbury et al. 2002)
• Greatest mortality due to predation
• Averaged over four bays and two years

15
Winter Juvenile Stage - next 135 days

• (Patrick 2000)
• Energy reserves (WBEC) and water temperature
  affect survival
• Age-0 winter mortality due to starvation
• Averaged over 12 bays

16
Egg Stage
Larval Stage
gtAge-3 spawn
hatch
next 92 days mortality 0.07 d-1
first 21 days
mortality 0.07 d-1
drift
Winter Juvenile Stage
Fall Juvenile Stage
nursery bay
Age-1

• next 135 days
• mortality 0.004 d-1

next 92 days mortality 0.01 d-1
17
Results

• Total survival through age-1
• 118 herring out of 1 million eggs
• Compare to range-based 1-6,500
• Consistent with the results of age-structured
  assessment (ASA)

18
Distribution of total mortality
ASA
ELH
Total mortality

• Average ELH mortality is lower than ASA mortality
• ASA incorporates mortality ages-0 through -3
• Greater uncertainty in the distribution of ELH
  mortality

19
Single-Stage Sensitivity Analysis

• Altered mortality of each life stage by 10
• Total survival was most affected by the larval
  stage
• Length of stage (92 days) and mortality level (
  0.07 d-1) is cause
• Other life stages had an order of magnitude less
  effect

20
Conclusions

• Life stages did not contribute equally to
  mortality and survival
• Larval stage has the largest influence on total
  survival
• This model shows that there is high uncertainty
  in the early life history

21
Acknowledgments

• Exxon Valdez Oil Spill Trustee Council
• Jeep Rice
• Mark Carls

22
Daily Survival and Daily Mortality
.
23

• Results of single-stage sensitivity analysis.
  Both of the textured series are the results from
  increasing (left) or decreasing (right) daily
  mortality (zi) while the black series is the
  total survival estimated from the base early life
  history (ELH) model

24
Interaction Sensitivity Analysis

• Determined all possible paired combinations of
  life stages
• Altered each pair of mortality estimates by 10
• Larval stage combined with any other life stage
  contributed the most to total survival
• Total survival maximized by decreasing mortality
  for larval and egg stages