Comparing internal and external drivers in the southern Benguela

and southern Humboldt Upwelling ecosystems

Shannon, Neira, Moloney, Roy, Cury

• 1978-2003 --- 27 time series:• 10 effort/F driver series;• Model fitted to 7 catch series and 10 abundance series

TROPHIC MODELSOUTHERN BENGUELA

(SOUTH AFRICA):

Southern Humboldt

(Central Chile): 1970-2004

75° 70°

31°

40°

32°

33°

34°

35°

36°

37°

38°

39°

71°72°73°74°

Concepción

Valparaiso

Valdivia

Exploring drivers of ecosystem variability

• Drivers considered during model fitting: i) Internal forcing: trophic flow controlsii) External forcing: fishing and the environment.

• Fishing carefully managed in the southern Benguela (previous studies shown most resource variability attributed to internal forcing and environmental forcing)

• Fishing shown to have played relatively major role in driving ecosystem changes observed in the southern Humboldt

• 2 hypotheses tested:i) Upwelling ecosystems function as wasp-waist systemsii) Although completely different environmental drivers act in the two ecosystems, the processes whereby these effects are transferred through the ecosystems and manifest themselves as ecosystem changes and observed resource dynamics, are essentially similar

Model investigations

• Compare 25 interactions for which the model is most sensitive in terms of Vs

• Find “best” required forcing function acting on PP, which improves model fits

• Explore additional FFs applied to specific interaction(s)

• Compare the FF applied to PP and/or specific interactions with existing environmental time series

Birds andSardine in

South Africaand Namibia

(from Crawford 1999)

0

2

4

1983 85 87 89 91 93 950

0,4

0,8

0

0,3

0,6

1956 61 66 71 76 81 86 91 960

500

1000

1500

Table 1a: The top 25 most sensitive interactions in the model of the southern Benguela

Anc

hovy

Sar

dine

Adu

lt ho

rse

mac

kere

l

Mes

opel

agic

fish

Sno

ek

Oth

er la

rge

pela

gic

fish

Cep

halo

pods

Sm

all M

. cap

ensi

s

Larg

e M

. cap

ensi

s

Sm

all M

. par

adox

us

Larg

e M

. par

adox

us

Sea

ls

Bird

s

Phytoplankton T Microzooplankton T Mesozooplankton B T T Macrozooplankton T B T B Anchovy B B Sardine T B B T T B B Juvenile horse mackerel B T Adult horse mackerel T Small M. capensis T B Small M. paradoxus B T

Table 1b. Top 25 sensitive interactions in the southern Humboldt model and the resulting vulnerability value in terms of bottom-up (B if v<2) or top-down (T if v>2).

Predator

Prey phyt

opla

nkto

n

zoop

lank

ton

I

zoop

lank

ton

II

zoop

lanc

ton

III

jellie

s

mac

robe

ntho

s

anch

ovy

com

mon

sar

dine

mes

opel

agic

s

hors

e m

acke

rel

hake

(lar

ge)

hake

(sm

all)

pela

gics

I

dem

ersa

l fis

h I

dem

ersa

l fis

h II

chon

dric

thth

yan

pela

gics

II

ceph

alop

ods

sea

lion

sea

bird

s

ceta

cean

s

phytoplankton B Bzooplankton I T Tzooplankton II B B Bzooplancton III T T Bjelliesmacrobenthos T T Banchovy Tcommon sardine Tmesopelagics Bhorse mackerel Thake (large) T Thake (small) B T Tpelagics Idemersal fish Idemersal fish IIchondricththyanpelagics IIcephalopods Tsea lion Tsea birdscetaceansDetritus B

Model fits – Southern Humboldt

Scenario Total1 F series Top 25 vulnerabilities EwE Anomaly

Phytoplankton 28 21 13 62

2 F series EwE Anomaly Phytoplankton

Top 25 vulnerabilities

28 16 21 64

3 F series Top 25 vulnerabilities EwE Anomaly Phytoplankton

EwE Anomaly Phytoplankton and

small pelagics

EwE Anomaly Phytoplankton and all significant interactions

28 21 13 - -

4 F series Environment on small pelagics

Environment on small pelagics+ 25 top v's

(Environment on small pelagics+ 25 top v's) +

EwE Anomaly 28 3 29 8 68

5 F series 25 top v's + EwE Anomaly

28 42 70

6 F series Environment on small pelagics

Environment on small pelagics+( 25 top v's +

EwE Anomaly)

28 3 44 75

0.5

1

1.5

2

1970 1975 1980 1985 1990 1995 2000 2005-1.5

-1

-0.5

0

0.5

1

1.5EwE anomaly in PP

SST anomalya

0.5

1

1.5

2

1970 1975 1980 1985 1990 1995 2000 2005

SST anomaly

EwE anomaly in PPb

ENVIRONMENTAL FORCING

IN THE SOUTHERN HUMBOLDT

Squat lobster

0

1

2

3

4

5

1970 1975 1980 1985 1990 1995 2000

ObservedEwE sc 1EwE sc 2 (PP+vs)EwE sc4

Model fits – Southern BenguelaScenario 1

Top 25 Vs FF applied to PP A 2nd FF, applied to anchovy-meso/macrozooplankton interaction

Total % SS reduced

32 9 9 51 Scenario 2

Top 25 Vs FF applied to PP A 2nd FF (actual data) applied to anchovy-meso/macrozooplankton interaction

Total % SS reduced

32 9 1 43 Scenario 3

Top 25 Vs FF1 applied to anchovy-meso/macrozooplankton interaction

FF2 applied to sardine-meso/macrozooplankton interaction

FF3 applied to PP

Total % SS reduced

32 12 4 7 55 Scenario 4

Top 25 Vs FF1 applied to anchovy as a prey group

FF2 applied to sardine as a prey group

FF3 applied to PP

Total % SS reduced

32 22 4 3 61 Scenario 5

Vs and FF1 (combination from actual upwelling anomalies) applied to PP

FF2 (hypothetical) applied to anchovy-meso/macrozooplanton interaction

Total % SS reduced

6 20 26 Scenario 6

Top 25 Vs & FF applied to PP simultaneously

FF1 applied to anchovy-meso/macrozooplankton interaction

FF2 applied to sardine-meso/macrozooplankton interaction

Total % SS reduced

48 8 3 59

25Vs+FF1 anch as prey;FF2 sard as prey; FF3 on PP SCENARIO 4

02468

10121416

1 38 75 112 149 186 223 260 297

FF1 applied to anchovy as aprey speciesFF1 anchovy-meso/macrozooplankton

00.5

11.5

22.5

33.5

4

1 39 77 115 153 191 229 267 305

FF2 applied to sardine as aprey speciesFF2 sardine-meso/macrozooplankton

INVERSE Forcing functionsfound for Anchovy as prey, & for anchovy –meso/macroZooplankton Interactions

(1986, 91 and 99-2001 high anchovy abundance)

86 9199

INVERSE Forcing functionsfound for Sardine as prey,& for sardine – meso/macro zooplankton interactions

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

1978 1982 1986 1990 1994 1998 2002

sardine catch fishingfishing+25Vs fishing+25Vs+FF on PPfishing+25Vs+3 sequentail FFs FFs anch&sard as prey

0

0.5

1

1.5

2

2.5

3

1978 1982 1986 1990 1994 1998 2002

anchovy catch fishingfishing+25Vs fishing+25Vs+FF on PPfishing+25Vs+3 sequentail FFs FFs anch&sard as prey

Anchovy catch

Sardine catch

Comparing drivers

• Southern Benguela and southern Humboldt upwelling systems are wasp-waisted in terms of flow control

• Fishing: much more important driver of ecosystem changes in SH than in SB (10-13% of variability explained by fishing in Benguela versus 28% in Humboldt)

• Trophic parameters (Vs) explain more variability than the anomaly on PP, suggesting effects of environmental forcing or bottom-up changes will depend on the internal functioning of the food web in terms of vulnerability of groups to predation

Conclusions:Environmental forcing

• S. Humboldt:Environment forces anchovy and common sardine in same way, affecting their consumption of zooplankton and their consumption by predators; Enviro-driven variability in PP also important

• S. Benguela:Sardine and anchovy forced by opposite environmental drivers / react in opposite ways to the same environmental driver)Environment affects recruitment, effects then propagated up and down from the wasp-waist

Final comments• Models as exploratory tools to get at underlying

mechanisms• Anchovy-sardine interactions with zooplankton

and predators - critical in driving changes in both systems, although the mechanisms differ

• S. Benguela: environmental data series to be combined into forcing functions as index of mechanisms determining the way an ecosystem functions –

“environmental suitability indices”