Christian Möllmann, Michael A. St.John, Rabea Diekmann and Georgs Kornilovs

F

PS AC

Climate/Hydrography

C SPR

Trophic cascade

Indirect effects of climate- and

overfishing-induced zooplankton changes on ecosystem structure –regime shifts, trophic cascades and feedback

loops in a simple ecosystem

P2P-loop

University of Hamburg

The Baltic SeaCharacteristics•large semi-enclosed brackish water body• stratified water-column with a permanent halocline• low diversity• high productivity• eutrophication • pronounced climate influence through variability in temperature & salinity• high fishing pressure

Central Baltic Sea

”Ecosystem Assessment”– Data & Methods”

Time series from 1974-2005

65 variables (12 fish-related, 6zooplankton, 20 phytoplankton-related, 8nutrient, 19 physical datasets)

Principal Component Analysis

Traffic-light plot

Source: ICES/HELCOM Working Group on Integrated Assessments of the Baltic Sea [WGIAB]

Ecosystem Regime Shift – ”Traffic lights”

AcartiaSprat

CodPseudocalanus

Coding: blue – low; red - highVariable

Aims

Demonstrate the influence of climate and human forcing on the pelagic Baltic ecosystemCentral importance of zooplankton (2 dominant

copepod species) for the ecosystem

1)Climate effects on Pseudocalanus acuspes and Acartia spp.

2) Zooplankton effects on cod and sprat recruitment

3) A trophic cascade & feedback-loops

Data & Methods

Data:Zooplankton

- Latvian Fish Resources Agency- Spring- Judai-Net (160µm)

Phytoplankton- Biomass- ICES Database

Hydrography- Temperature, salinity, cod reproductive

volume (RV)- ICES Database

Fish- Biomass and recruitment (age 0)- Multispecies Virtual Population Analysis

(MSVPA)- ICES Study Group on Multispecies

Assessments of the Baltic Sea

Methods:Regime shift analysis

- Sequential Regime Shift Detection Method (Rodionov 2004)

Statistical modelling- Linear Models- Generalized Additive

Models (GAMs)

Model selection- Linear Models: AIC

(Akaike Information Criterion)

- GAMs : GCV (Generalized Cross

- Validation criterion)

”Regime-shift” between key-species

Climate/Hydrography

AC

+ Temperature

PS

- Salinity

1975 1980 1985 1990 1995 2000 2005

Aca

rtia

spp.

0

20

40

60

80

Tem

pera

ture

1

2

3

4

5

6

Regime Shifts detected by Sequential Regime Shift Analysis

(Rodionov 2004)

1975 1980 1985 1990 1995 2000 2005

P. a

cusp

es

0

20

40

60

80

100

120

140

Sal

inity

6.6

6.8

7.0

7.2

7.4

7.6

PS – Pseudocalanus acuspes, AC – Acartia spp.

Pseudocalanus acuspes & Hydro-Climate

Results of GA-ModellingPredictors GCV R2 (%)

Salinity*** 0.313 63.0

Salinity***, NAO 0.311 72.0

*** p < 0.001

GCV – measure of model quality (the lower the better)

Acartia spp. & Hydro-climate

** p < 0.01Results of Linear-Modelling

Predictors AIC R2 (%)

Temperature** 94.73 23.3

Temperature, NAO 93.35 31.0

AIC – the lower the better ... !

”Regime-shift” between key-species

Climate/Hydrography

AC

SPR

+ Temperature

PS

C

- Salinity

1975 1980 1985 1990 1995 2000 2005

Spr

at

0

500

1000

1500

2000

2500

1975 1980 1985 1990 1995 2000 2005

Cod

0

200

400

600

800

1000

PS – Pseudocalanus acuspes, AC – Acartia spp., C – Cod, SPR - Sprat

Results of GA-Modelling

Cod recruitment & Hydro-Climate & P. acuspes

Predictors GCV R2

(%)

SSB*** 0.255 76.2

SSB***, Reproductive Volume 0.242 79.4

SSB***, Reproductive Volume*, P. acuspes** 0.187 85.8

SSB***, Reproductive Volume*,P. acuspes**, NAO

0.205 85.8

Response – ln(R/SSB) -> Recruitment success

***p < 0.001** p < 0.01* P < 0.05

Sprat recruitment & Hydro-Climate & Acartia spp.

Response – ln(R/SSB) -> Recruitment success

Results of Linear-Modelling

Predictors AIC R2

(%)

SSB** 68.46 20.0

SSB**, Temperature* 64.34 40.2

SSB***, Temperature*, Acartiaspp.* 63.66 49.52

SSB***, NAO*, Acartia spp.*, 76.34 87.6

***p < 0.001** p < 0.01* P < 0.05

1975 1980 1985 1990 1995 20000

500

1000

1500

Din

ofla

gella

tes

2

4

6

8

10

12

14

16C

od 0

200

400

600

800

1000

1200

0

500

1000

1500

2000

2500

Spra

t

0

200

400

600

800

1000

Pseu

doca

lanu

s ac

uspe

s0

500

1000

1500

2000

2500

3000

2

4

6

8

10

12

14

Species-level Trophic Cascade

Climate/Hydrography

AC

+ Temperature

SPR

PS

C

- Salinity

FTrophic cascade

SPR

PS

C

Cod

Fi

shin

g M

orta

lity

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

PS – Pseudocalanus acuspes, AC – Acartia spp., C – Cod, SPR – Sprat, F – Fishery

”Limiting” Predator –2- Prey LOOP

Climate/Hydrography

AC

+ Temperature

SPR

PS

C

- Salinity

FTrophic cascade

SPRC SPR

PS

CP2P-loop

Sprat0 500 1000 1500 2000 2500 3000

Pseu

doca

lanu

s ac

uspe

s

0

5

10

15

20

25Cod

0 200 400 600 800 1000 1200

Spra

t

0

500

1000

1500

2000

2500

3000

P2P-loop after Bakun (2007)

Pseudocalanus acuspes & Hydro-Climate & Top-down control

Results of GA-Modelling

Predictors GCV R2

(%)

Salinity*** 0.313 63.0

Salinity***, NAO 0.311 72.0

Salinity**, NAO, Sprat biomass** 0.230 80.8

***p < 0.001** p < 0.01

”Promoting” Predator –2- Prey LOOP ?

Climate/Hydrography

AC

+ Temperature

SPR

PS

C

- Salinity

FTrophic cascade

SPRC

D

1975 1980 1985 1990 1995 20000

500

1000

1500

Din

ofla

gella

tes

2

4

6

8

10

12

14

16C

od 0

200

400

600

800

1000

1200

0

500

1000

1500

2000

2500

Spra

t

0

200

400

600

800

1000

Pseu

doca

lanu

s ac

uspe

s0

500

1000

1500

2000

2500

3000

2

4

6

8

10

12

14

Cod

Fi

shin

g M

orta

lity

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

P2P-loop

PS – Pseudocalanus acuspes, AC – Acartia spp., C – Cod, SPR – Sprat, F – Fishery, D - Dinoflagellates

Acartia spp. & Hydro-Climate & Bottom-up control

Results of GA-Modelling

Predictors GCV R2 (%)

Temperature*, Dinoflagellates** 0.592 80.4

Temperature, Dinoflagellates, NAO 0.614 83.9

** p < 0.01* P < 0.05

Summary

Climate-induced changes in salinity and temperature have caused a Regime-shift in the pelagic Baltic ecosystem on all trophic levels

Regime-shift from salinity-controlled (P. acuspes/cod) to temperature-controlled (Acartia spp./sprat) species

P. acuspes and Acartia spp. are key ecosystem components, mediating the climate effect to the important fish stocks

Overfishing of cod cascades down to P. acuspes (and potentially to dinoflagellates ?), stabilizing the new regime through P2P-loops