ECOSYSTEM JELLO

1
ECOSYSTEM JELLO

• Kerim Aydin Bob Francis Pat Livingston

2
Ecosystem properties

• Stability Many Many terms
• Resilience
• Resistance
• Food web structure?
• Diversity
• High biomass or other adaptations in susceptible
  groups
• MEASURE VARIANCE

3
Need to understand control

• TOP DOWN?
• BOTTOM UP?
• MIDDLE OUT?
• Which (if any) is dominant in marine systems??
• Explanations may be easy to find and hard to
  confirm!

4
Kitchellization
Marine Mammals
Sharks

• The process of losing beer money by making poor
  guesses about ecosystem responses

Large Fish
???
Forage Fish
Zooplankton
Phytoplankton
5
Revenge??
Marine Mammals
Sharks
Large Fish
Forage Fish
Zooplankton
Phytoplankton
6
Top-down-bottom-up-ENSO-PDO-fishing?

• Figure the Frequency!!!

7
Where can you go from a marine food web?

• Process models
• Investigate specific environmental linkages
• Address species concerns (endangered spp.)

Food Webs

• Multi-species harvest models
• Predict returns from harvest strategies

• Biomass dynamics models
• Measure broad ecosystem properties
• Assess risks of regime shifts
• Determine the need for mechanisms

8
Why use the most general models?Regimes Happen!

• Biomass dynamics models
• Measure broad ecosystem properties
• Assess risks of regime shifts
• Determine the need for mechanisms

• There needs to be a way to abstract ecosystems,
  and look for risks.
• This is one example of using whole ecosystem
  properties to examine this risk.

9
Food web forcing?
Marine Mammals
Sharks
Large Fish
????
Forage Fish
Zooplankton
Phytoplankton
10
Correlation requires Variation
11
Variation comes with history

• This is maybe not so important in labs and lakes
  (white systems).
• But in large marine systems.

12
History vs. the correlation method - equilib. Vs.
time...

• TOP DOWN OR BOTTOM UP???

13
How can variability vary?

• Amplitude
• Frequency
• Cadence

14
What are the characteristics of a food web
component?
15
What are the characteristics of a food web
component?

• Biomass (tons)
• Production (tons/year)
• P/B (1/year)-
• inverse of replacement time
• related to generation time

B
P
P/B
16
What are the characteristics of a food web
component?

• Biomass (tons)
• Production (tons/year)
• P/B (1/year)
• Trophic level

B
P
P/B
17
What are the characteristics of a food web
component?

• Biomass (tons)
• Production (tons/year)
• P/B (1/year)
• Trophic level
• Network characteristics
• Dissipative characteristics

B
P
P/B
18
B vs. P/B
P

• R vs. K selected...

P
B
B
P/B
P/B
19
This experiment

• Start with two Actual Food Webs

The East. Bering Sea Shelf system with pollock
as the dominant fished species The East.
Tropical Pacific Tuna are dominant fished
species Both have shown interannual variation in
primary production which may be linked to climate
signals.
20
Need rules for species interactions
????

• One possibility of many
• P/B then B of predator respond to increases of
  prey biomass
• overly stable - only looking for chances of
  regimes
• Ratio-dependent predator/prey model with
  satiation (ECOSIM based)
• Mimics surplus production (Pella-Tomlinson) model
  when predator, prey fixed.
• Next step is to add better age-structure (single
  biggest weakness of the model).

21
Dynamics of overlap
Its cold down there!
V
Bj
B-V
aijVijBj
vij (Bi-Vij)
Bi - Vij
Vij
Assume fast equilibrium for Vij
vijVij
dVij /dt vij(Bi-Vij) - vijVij - aijVijBj
22
The appearance of Density Dependence

• dVij /dt vij(Bi-Vij) - vijVij - aijVijBj 0
• Vij vijBi/(2 vij aijBj)
• Cij (Bi,Bj) aijvijBiBj
• (2 vij aijBj)

Prey biomass
Cij (or Minstant)
Cij /Bj
Predator Biomass
23
P/B vs. Trophic Level - EBS

• (note missing microzooplankton)
• B vs. Trophic Level? No firm relationship

24
P/B vs. Trophic Level - ETP
25
Insert bottom up variation

• Use natural signal (e.g. ENSO)

26
Insert bottom up variation

• Use natural signal (e.g. ENSO)
• Vary amplitude, frequency, cadence

27
Insert bottom up variation

• Use natural signal (e.g. ENSO)
• Vary amplitude, frequency, cadence

28
Insert bottom up variation

• Use natural signal (e.g. ENSO)
• Vary amplitude, frequency, cadence

29
Results

• Heres a couple 100-year time tracks...
• We care about
• Amount of variability transmitted (CV over 100
  years)

30
Amplitude and cadence

• Increasing the amplitude of forcing increases the
  amplitude of response.
• Cadence is complex, and depends too heavily on
  (unknown) parameters.

31
Frequency and Variation
32
Frequency and Variation
33
Frequency and Variation
34
Frequency II - CV (tuna)
35
What about correlations and control??
36
One theoretical explanation...

• Simply hitting the resonant frequencies of each
  model component?
• Useful for model (and real life) analysis of
  important terms.
• Real life P/B values may imply natural resonant
  frequencies.

37
Small observations

• Missing seasonal/micronekton
• Frequencies, P/B are the same unit (1/time)
• Trophic Level less of a fit

38
Fish must follow history or be history

• If P/B of a species is in the range that it is
  excited by the balance of top-down/bottom up,
  does it need extra biomass to be stable (avoid
  regimes)?
• We dont know the frequency of primary production
  variation in many systems.
• Need to look at more ecosystems

39
What responds to each frequency range?

• Forage fish, micronekton response peaks near
  ENSO-scale forcing.

• This doesnt mean that they vary on an ENSO
  scale, but that they are most susceptible to
  crashes when the bottom-up forcing is at that
  scale.

40
What responds to each frequency range?

• Forage fish, micronekton response peaks near
  ENSO-scale forcing.

• Larger commercial fish response peaks at regime
  (10-50 year) forcing.

41
So what about fishing?

• General principle of surplus production As B
  goes down, P/B goes up, due to
• more food per fish
• smaller, faster growing fish

P
P
B
B
P/B
P/B
42
Are our fish becoming anchovies?

• Beyond multispecies MSY, have we changed the
  natural time scale of animals (P/B, replacement,
  generation time) without changing the natural
  time scale (frequency) of input variation?

43
Shift happens (with a little help from)

• Fishing may push a species in or out of the high
  CV range.
• Will regime changes occur MORE or LESS frequently
  with fishing?
• It can occur in both directions the fin whale
  control?
• What is the range of fishing change compared to
  natural variability in P/B?

44
Preliminary until...

• More ecosystems and parameters
• Devil in the details
• Model type, (L.V., ECOSIM, Spatial, parameters)
• Fishing changes (historical)
• Still, true for reasonable forms
• Middle-out Forcing

45
CONCLUSIONS

• The frequency of primary production variation
  may be strongly connected to P/B.
• The frequency associated with regime shifts
  (10-50 year period) is the frequency at which
  most currently fished species show the strongest
  response.
• Fishing may push P/B into or out of the range of
  greatest variation, depending on the frequency of
  natural forcing in the ecosystem.