64 



Fishery Bulletin 104(1) 



A simple Lagrangian trajectory model was devel- 

 oped to estimate the drift of planktonic stages. The 

 model used the observed currents at ADCP moorings 

 A and B to calculate trajectories. Because of the lack 

 of information on spatial current variations and the 

 difficulty of extrapolating vertical current profiles 

 from one depth and bottom relief to other conditions, a 

 simple two-dimensional (horizontal) simulation model 

 was used. For the computations we selected the high- 

 est bin from the ADCP that had good data throughout 

 the tidal cycle at each station. This bin typically was 1 

 m below the mean water level at that location and was 

 the highest bin not affected by instrument sidelobe 

 interference. Because instantaneous bottom currents 

 were closely aligned in direction with surface currents 

 and because magnitudes were only 25% lower, the 

 tidal currents were barotropic and we used, therefore, 

 the surface currents to estimate the largest possible 

 distance traveled. 



The larval transport was calculated with the equation 



dxjdt = w,, 



where .r, = position vector of the larvae; 

 t = time; and 

 «, = the local current velocity. 



An Euler (forward in time) integration rule was used 

 to numerically solve this equation. Because only two 

 ADCPs (A and B) provided the current data for this 

 large region, no attempt was made to extrapolate a 

 current field from them. Simulations were run by using 

 current meters A and B independently and by assum- 

 ing that the current field was spatially homogeneous. 

 The trajectories therefore were two-dimensional in 

 the horizontal plane and the result was identical to a 

 progressive vector diagram. Comparison between the 

 two sets of trajectories (A and B) provided an estimate 

 of the possible variability. The model was used to ex- 

 plore the potential transport of planktonic stages under 

 various assumptions of behavior controlled by environ- 

 mental cues: 1) a behavioral response to salinity and 

 light, 2) a diel behavior, 3) a diel and tidal behavior 

 throughout the planktonic phase, and 4) an ontogenic 

 change that began with diel behavior and added a 

 tidal behavior at the 15''^ day. All four hypotheses of 

 larval behavior in relation to transport were simulated 

 in order that their effects on distance traveled could 

 be compared and contrasted. In all simulations, we 

 assumed that larvae traveled only at night. We also 

 assumed that the source of the pink shrimp larvae 

 was located immediately northeast of the Dry Tortugas 

 about 150 km from western Florida Bay (Cummings, 

 1961; Jones at al., 1970; Roberts, 1986). The program 

 simulated distances traveled by particles for a period 

 of 30 days (e.g., days 1-30, 31-60, 61-90, etc.), a pe- 

 riod that corresponds to the estimated developmental 

 period for pink shrimp from the time of hatching to the 

 postlarval stage when larvae are ready for settlement 

 (Dobkin, 1961; Ewald, 1965). 



Results 



Patterns of postlarval supply, SST, and winds 



The monthly influx of postlarvae through the Middle 

 Florida Keys channels (WH and PH) exhibited a highly 

 variable temporal pattern from year to year. Postlarvae 

 were observed every month through the three years 

 (Fig. 2A). Peaks of postlarvae through the Middle Keys 

 channels occurred in May, July, August, and October 

 2000; in January, July, and October 2001; and in Janu- 

 ary, March, June, and September 2002. In contrast, the 

 monthly influx of postlarvae through the northwestern 

 stations (SK and MG) showed a strong seasonal pattern 

 with one distinct high peak centered in summer from 

 July through September for each year of the 3-year 

 period (Fig. 2B). The number of postlarvae entering 

 through northwestern Florida Bay was much higher 

 than through the Keys stations. The mean concentration 

 of postlarvae per station over the 3-year period indicated 

 that concentrations of postlarvae entering northwestern 

 Florida Bay through SK and MG channels were about 

 eight times greater than through the Florida Keys chan- 

 nels of WH and PH (Fig. 3). Results from a two-way 

 ANOVA indicated that there was a significant effect 

 of site (northwestern stations vs. Florida Key stations) 

 and month on the supply of postlarvae entering Florida 

 Bay (Table 2). 



Winds showed a seasonal pattern; the spring and 

 summer were dominated by weak southeasterly winds 

 and the fall and winter, by strong northerly winds 

 (Fig. 2, C-D). The monthly average alongshore showed 

 a weak northward constituent in spring-summer of 

 each year (Fig. 2C). The monthly average cross-shelf 

 winds were consistently negative (toward the west) 

 and showed no seasonality (Fig. 2D). The average tem- 

 perature over the three-year time series was 26.1°C, 

 and winter temperatures in 2000-01 were lower than 

 in 2001-02 (Fig. 2E). In summer, during the period of 

 peak postlarval immigration through the northwestern 

 stations (MG and SK), the alongshore wind constituent 

 was mainly northward, the cross-shelf wind was west- 

 ward, and the SST was above average during the three- 

 year period (Fig. 2, A-E). Postlarval concentrations at 

 the northwestern stations were correlated with SST and 

 alongshore winds (Fig. 2, B-E; Table 3). Postlarval con- 

 centrations at the Florida Keys stations (WH and PH) 

 were not correlated with either winds or SST. 



Subtidal and tidal currents at the SW Florida shelf 



Advective displacement derived from two ADCP velocity 

 records indicated that the net current is primarily in the 

 alongshore direction (Fig. 4, A and B). The alongshore 

 flow from onshore mooring A was northward, had a total 

 mean velocity of 0.0062 m/sec, and a total water dis- 

 placement of 589.3 x 10' m over the three years (Fig. 4A). 

 The cross-shelf flow was westward, had a mean veloc- 

 ity of -0.0005 m/sec, and a total water displacement of 

 -289.7 X 10^ m. The alongshore flow from offshore moor- 



