Copyright © NISC Pty Ltd

African Journal of Marine Science 2008, 30(3): 437–452 Printed in South Africa — All rights reserved

AFRICAN JOURNAL OF MARINE SCIENCE ISSN 1814–232X EISSN 1814–2338 doi: 10.2989/AJMS.2008.30.3.1.635

Does vertical migratory behaviour retain fish larvae onshore in upwelling ecosystems? A modelling study of anchovy in the southern Benguela C Parada1*, C Mullon2, C Roy3,4, P Fréon2,5,6, L Hutchings5 and CD van der Lingen5 School of Aquatic and Fisheries Sciences, 1122 Boat Street NE, Seattle, WA 98105, USA IRD, UR097 ECO-UP, IRD, CRHMT, BP 171, 34203 Sète Cedex, France 3 IRD, Laboratoire de Physique des Océans, UMR 6523, CNRS-Ifremer-IRD-UBO, BP 70, 29280 Plouzané, France 4 Department of Oceanography, University of Cape Town, Private Bag X3, Rondebosch 7701, South Africa 5 Marine and Coastal Management, Department of Environmental Affairs and Tourism, Private Bag X2, Rogge Bay 8012, South Africa 6 Zoology Department and Marine Research Institute, University of Cape Town, Private Bag X3, Rondebosch 7701, South Africa * Corresponding author, e-mail: [email protected]

1

2

Manuscript received April 2008; accepted August 2008

A spatially explicit individual-based model (IBM) forced by 3D temperature and current fields simulated by a hydrodynamic model of the southern Benguela upwelling region was used to test two hypotheses concerning the role of diel vertical migration (DVM) by Cape anchovy Engraulis encrasicolus larvae and pre-recruits. These hypotheses were that: (1) DVM enhances alongshore transport of anchovy eggs and larvae from the spawning grounds to the nursery area while avoiding the lethal effect of low water temperatures in the upwelling system, and/or (2) DVM enhances the transport of larvae and pre-recruits from the offshore to the onshore domain of the nursery area, and then counteracts offshore advection by favouring retention. We tracked the trajectories of virtual particles in the model and calculated a pre-recruitment index as a proxy for transport success to the nursery area (onshore and offshore) and found that the index

increased from 10% to 20% after the incorporation of larval vertical migration into the IBM, with virtual individuals held at depths of around 60 m showing maximal pre-recruitment index values. Hence, DVM does appear to enhance transport to the nursery area (offshore) for early and late larvae. Model outputs showed coarse-scale horizontal distribution patterns of larvae by age/size class that are similar to field observations for early, small larvae but not for large larvae and pre-recruits. Observations show that early/ small larvae are located offshore whereas older/ larger larvae and pre-recruits are found closer to the continental shelf and the inner nursery grounds. This disparity between model results and field observations does not support the hypothesis that DVM is one of the mechanisms involved in the onshore movement of early life-history stages, especially for large larvae.

Keywords: anchovy, individual-based model, larval vertical migration, recruitment, southern Benguela, transport, upwelling

Introduction The life history of Cape anchovy Engraulis encrasicolus in the southern Benguela upwelling system has been well studied. The species spawns serially over the Agulhas Bank during austral spring to summer (September–March) with a peak in November (Huggett et al. 1998, van der Lingen et al. 2001), and eggs spawned there are transported by a jet current to the nursery grounds on the West Coast between Cape Columbine and the Orange River (Figure 1). The jet current accelerates along the Cape Peninsula, bypasses the active upwelling centres off Cape Town and Cape Columbine (Hutchings et al. 2002) and develops three components north of Cape Columbine: offshore,

alongshore and onshore (Boyd et al. 1992). Efficient transport by the jet current of early stages of anchovy from the Agulhas Bank spawning ground to the West Coast nursery area, and retention within that area, are thought to be critical factors influencing their recruitment variability (Hutchings et al. 1998). Intense, wind-induced offshore Ekman drift and strong upwelling dominate the oceanography of the nursery ground during the anchovy spawning season, and mesoscale instabilities triggered by upwelling lead to the development of eddies and filaments capable of transporting shelf waters hundreds of kilometres offshore, resulting in most cases in the loss of biota to oligotrophic

438

Parada, Mullon, Roy, Fréon, Hutchings and van der Lingen

F AFRICA

SOUTH AFRICA

South Africa

Orange River

m

ore

0 10

ore sh On

fsh Of

0 20

Nursery grounds

SOUTH AFRICA

m 0m 20 m 0 50

0 10 m

33° S

32° S

Cape Town

E

EABOn EABOff

Cape Town

34° S

EABOn EABOff

WAB 36° S

D

−2

5 000

16° E

4 000

Eggs m 3 000 2 000 1 000

18° E

CABOff B

C

WAB 36° S

Cape Columbine

Spawning grounds

CABOff

20° E

A

22° E

16° E

18° E

20° E

Figure 1: Location of anchovy spawning and nursery grounds. Spawning grounds are separated by lines A–D: Western Agulhas Bank (WAB), Central Agulhas Bank offshore (CABOff), Eastern Agulhas Bank onshore and offshore (EABOn and EABOff respectively); and nursery areas are located between lines E–F. Some areas are subdivided into onshore and offshore in the individual-based model simulations, using the 100 m isobath to separate the onshore from offshore spawning grounds, and the 200 m isobath to separate the onshore from the offshore nursery grounds. The outermost cross-shelf boundary of all offshore areas is the 500 m isobath. The insert shows a composite distribution map of anchovy eggs collected using a CalVET net during annual pelagic spawner biomass surveys over the period 1984–2000 (from van der Lingen and Huggett 2003)

oceanic areas (Fowler and Boyd 1998). Passive offshore transport in the southern Benguela is considered to be a major cause of mortality of anchovy larvae and post-larvae aged 2–3 months (Shannon et al. 1996), because the surrounding waters are typically unfavourable for survival. Varying probabilities of larval survival are presumed for those spawned over the Agulhas Bank and dispersed along the productive West Coast, with larvae that drift far offshore being assumed to have less chance of recruiting successfully than those that remain close onshore (Hutchings et al. 1998). Offshore larval drift is also considered to be a major process responsible for reduced survival in other pelagic fish species such as northern anchovy E. mordax off northern California (Hewitt and Methot 1982) and Atlantic menhaden Brevoortia tyrannus off North Carolina (Joyeux 1998). Previous efforts to understand the causes of recruitment variability of pelagic fish in upwelling systems focused on processes such as first feeding, offshore transport and retention mechanisms (Sinclair 1988, Platt et al. 2003). Bakun (1996) pointed out that fish larvae inhabiting the surface layer

in coastal upwelling regions would tend to be swept away from their favoured coastal habitat by strong offshore Ekman transport. In upwelling areas, offshore transport of surface waters is balanced by a compensatory onshore flow at depth. Organisms employing appropriate vertical migratory behaviour could take advantage of this onshore circulation component to maintain themselves in the nearshore domain or to minimise offshore losses (Pillar and Stuart 1988, Hill 1991, Verheye et al. 1991, Batchelder et al. 2002). An individual-based model (IBM) for anchovy in the southern Benguela system by Huggett et al. (2003) represented individuals as Lagrangian particles and quantified the success rate of the transport of particles from the spawning to the nursery grounds. When the nursery ground was defined as the shelf and the adjacent slope regions (i.e. to the 500 m isobath) between Cape Columbine and the Orange River (Figure 1), the success rate of transport was relatively high at around 10% (Huggett et al. 2003). However, when the nursery area was restricted to the onshore shelf domain (i.e. onshore of the 200 m isobath),

439

African Journal of Marine Science 2008, 30(3): 437–452

the success rate of transport decreased to 2%. Historical records of the horizontal distribution of anchovy larvae along the west coast of South Africa shows a clear size gradient with small larvae (F)

Significance