FISHERY BULLETIN: VOL. 84, NO. 3 



used for all locations and all times in the model. The 

 numerical method incorporated diffusion as the 

 weighting factor coth [(uh)/(2K)] for the flux at each 

 grid cell interface, where u is the current velocity 

 at the interface and h is the 37 m grid spacing (see 

 Power 1984 for further details). Hence, diffusion 

 becomes important in regions of low current 

 velocity, and at higher velocities diffusion is less im- 

 portant and advection dominates the flux. For the 

 current velocities in most of the modeled region, the 

 above hyperbolic cotangent function is usually 

 evaluated to a magnitude near unity, making the 

 contribution of turbulent transport to larval drift 

 minimal relative to advective (current velocity) 

 transport. 



Simulations were carried out by starting an ini- 

 tial point source of northern anchovy eggs or larvae 

 at various locations historically known to be larval 

 anchovy habitat (Hewitt 1980). Examples of simula- 

 tions for four starting locations (Table 1; Fig. 1), 

 which are representative of the overall patterns pro- 

 duced by the simulations, are presented here. The 

 four locations will be referred to in the text by their 

 letter designations indicated in Figure 1 and Table 

 1. Northern anchovy larvae begin to school at about 

 27 d (Hunter and Coyne 1982); therefore larval 

 distributions after 30 d of drift are presented. 

 Thirty-d-old larvae are also rapidly increasing their 

 "patchiness" (Hewitt 1981a), indicating that they 

 could then exert significant control over their posi- 

 tion. The time step in the simulations was 1 d. 

 Results from a simulation using the actual northern 

 anchovy egg distribution found in 1982 as the ini- 

 tial condition can be found in MacCall (1983). 



Table 1.— Geographic and CalCOFI coordinates of start- 

 ing locations for simulations presented in this paper. Letter 

 designation corresponds to the same locations in Figure 1 . 



Northern anchovy larval concentrations in the 

 contour plots are relative to starting concentration; 

 the unitless contour value of 10~ 2 represents a lar- 

 val concentration two orders of magnitude below the 

 starting concentration, and only concentrations 

 down to 10 " 7 are illustrated. Larvae were per- 

 mitted to be advected out the borders of the modeled 

 area, except for the border along the coast. Grid cells 

 bordering the Santa Barbara Channel (at about lat. 



34°N, long. 120°W; Fig. 1) between the Channel 

 Islands and Point Conception were open, and lar- 

 vae advected into this region were considered to be 

 lost from the system. Larvae were not permitted 

 to be transported across any of the islands in the 

 modeled region. Because March is the peak spawn- 

 ing time of northern anchovy, the effects of different 

 starting locations on the 30-d larval distributions 

 during March conditions will be presented first. The 

 effects of spawning in different seasons and en- 

 hanced offshore Ekman transport during March will 

 then be presented for comparison. The simulation 

 results nominally represent larval northern anchovy 

 distributions, but the results also apply to any plank- 

 tonic species that begin drift at the same locations 

 and maintain themselves in the top 50 m of the 

 California Current. 



The overall extent of onshore-offshore and along- 

 shore transport was of major interest in this study. 

 A convenient way of summarizing the simulated 

 larval distributions relative to their cross-shore 

 distribution was to sum all larval concentrations in 

 the cells having the same CalCOFI station coordin- 

 ates. These sums were converted to percentages of 

 the total number of larvae at 30 d, and the cumula- 

 tive percentage of larvae present as one progressed 

 offshore was plotted versus CalCOFI station coor- 

 dinates. A similar procedures using CalCOFI line 

 coordinates was done to summarize alongshore 

 transport. 



RESULTS 



Effects of Starting Location, 

 Normal March Currents 



Northern anchovy larvae that began drift at loca- 

 tion B, near the coast, were transported downshore 

 by March currents (Fig. 3B). This was an effect of 

 the nearshore southeasterly current (Fig. 2), and 

 because of this flow only 15% of the larvae were at 

 or upshore of the starting location after 30 d of drift 



Figure 3.— Distribution of northern anchovy larvae after 30 d of 

 drift in March currents. Letter designation corresponds to a simula- 

 tion with northern anchovy begun at the corresponding lettered 

 location in Figure 1 and Table 1; starting location is marked in 

 this and subsequent contour plots with asterisks. Locations A and 

 C share the same CalCOFI station coordinate; points B and D have 

 the same CalCOFI line coordinate. Concentration contour inter- 

 vals are proportions of the starting concentration, decreasing in 

 order of magnitude steps. Tic marks around perimeter are at whole 

 degrees of latitude and longitude; dots are at intervals of 3.33 

 CalCOFI line units from lines 70 to 120 and intervals of 10 station 

 units offshore to station 120 (i.e., every 74 km). 



590 



