FISHERY BULLETIN: VOL. 84, NO. 3 



additional confounding factor is the lack of biological 

 information in the model; consistent larval behavior 

 patterns (e.g., diurnal vertical migrations) and 

 spatially heterogeneous mortality could produce 

 distributional patterns differing from those pre- 

 sented here. In spite of these caveats, the simula- 

 tions do demonstrate that variations in northern 

 anchovy spawning location and time, and changes 

 in the magnitude of offshore directed Ekman trans- 

 port, can have significant consequences for the 

 subsequent larval distribution. By inference, these 

 changes in distribution can result in increased or 

 reduced larval mortality, and ultimately affect adult 

 northern anchovy population size. 



Offshore transport was not significant in the sim- 

 ulations done with the unaugmented or "normal" 

 March (seasonal April geostrophic + March Ekman) 

 currents. A majority of the northern anchovy lar- 

 vae that began drift at the four starting locations 

 were inshore of their starting points after 30 d of 

 drift, and the cross-shore distributions indicated that 

 most larvae occupied a relatively narrow range of 

 distances close to shore. Mais (1974) and Methot 

 (1981) reported that most juvenile northern anchovy 

 occupy inshore areas in the fall, and the model in- 

 dicates that this inshore movement could be facil- 

 itated by passive drift. As mentioned earlier, the 

 consensus is that nearshore regions provide more 

 hospitable food conditions for the northern anchovy 

 larvae. Lasker (1978, 1981) summarized the results 

 of surveys of larval food distributions in the South- 

 ern California Bight. His figures indicate that suit- 

 able larval food concentrations decline rapidly as one 

 progresses offshore. O'Connell (1980) reported the 

 results of a survey for starving northern anchovy 

 larvae in the Southern California Bight, the degree 

 of starvation being defined by histological criteria. 

 He found apparently healthy larvae at locations as 

 far as about 250 km offshore (at lat. 32°30'N, long. 

 120°W), where model concentrations were <10~ 7 

 after 30 d in all March simulations except for lar- 

 vae begun at A, where they were <10~ 4 . Despite 

 the good condition of these offshore larvae, the 

 simulation results indicate a low likelihood of their 

 being recruited to the nearshore juvenile population. 

 The low offshore larval concentrations will also 

 hinder the development of schooling (Hewitt 1981a). 



The minimal offshore transport situation found in 

 the March current simulations was also generally 

 true when simulations were done using currents 

 from other seasons, except for northern anchovy lar- 

 vae begun at location A. This point is the most in- 

 terior starting location within the Southern Califor- 

 nia Bight proper and is primary northern anchovy 



spawning habitat (Hewitt 1980) and where seasonal 

 changes in the currents are especially important 

 (Tsuchiya 1980). Spring is a time when currents in 

 the Southern California Bight are not as well 

 organized as other times of the year, and the 

 Southern California Eddy is often absent (Hickey 

 1979; Owen 1980). It is interesting that currents 

 during March, the peak spawning period, produced 

 the least offshore transport of larvae begun at loca- 

 tion A when compared with other seasons, even 

 though March is the time of greatest overall Ekman 

 transport (Bakun and Nelson 1976). There is signifi- 

 cant spawning in January (Methot 1981), and the 

 January simulations also had reduced dispersal of 

 larvae. The model results support the hypothesis of 

 Parrish et al. (1981) that northern anchovy spawn- 

 ing in the Southern California Bight do so at a time 

 and place that minimizes offshore transport of eggs 

 and larvae. 



It is clear that the overall 30-d alongshore distribu- 

 tions of northern anchovy larvae produced by 

 normal March currents depended largely on the 

 spawning location's proximity to the well-defined 

 southeasterly current present near the coast in the 

 southern half of the modeled region. Larvae that 

 started drift near this current underwent extensive 

 downshore transport. Larvae begun farther into the 

 Southern California Bight (location A), and farther 

 offshore (location D), were also transported down- 

 shore, but to a much lesser extent. This again con- 

 firms the role of the Southern California Bight as 

 an area where minimal transport of spawning 

 products takes place. The southwesterly, offshore 

 transport that occurred in many of the simulations 

 at the southern margin of the modeled region 

 (between CalCOFI lines 110 and 120) is consistent 

 with the evidence that this region forms a faunal 

 boundary between species of the Southern Califor- 

 nia Bight and those of Baja California to the south, 

 and that this faunal boundary is created by current 

 patterns (Hewitt 1981b). This is also a region of in- 

 creased surface convergence (Parrish et al. 1981). 



The extent of alongshore transport was marked- 

 ly different for northern anchovy larvae begun at 

 the same starting location when currents from the 

 different seasons were used. Depending on start- 

 ing location, seasonal changes in currents could pro- 

 duce almost complete reversals between predom- 

 inantly upshore or downshore transport. March 

 currents consistently produced the greatest down- 

 shore transport. These effects were due to the pres- 

 ence or absence of the Southern California Eddy and 

 the Southern California Countercurrent. Because 

 the Southern California Countercurrent is present 



600 



