384 



Fishery Bulletin 102(2) 



Mean residence time 

 (area under each curve) 



Post Office Creek (20.4 days) 

 Stacey Creek (15.2 days) 



100 



Mean residence time 



(area under each curve): 



i East Fork (25.5 days) 

 i West Fork (20.4 days) 



40 60 



Days at large 



100 



Figure 6 



Estimates of mean residence times of marked shrimp 

 in (A) tidal creeks sampled in 1998 and ( B ) creeks sam- 

 pled in 1999. Estimates are based on the area under the 

 curves describing the proportion of recaptured marked 

 shrimp in each creek system that still remained to be 

 captured after the indicated number of days-at-large. 



displacements m). In contrast, one shrimp marked at 

 PO demonstrated a net displacement of 3 km when it was 

 recaptured at SC after 61 days at large. Nine shrimp (at 

 large from 18 to 49 days) marked at WF were recaptured 

 at EF. and two (at large 19 and 45 days) tagged at EF were 

 recaptured at WF. It was not possible to determine pre- 

 cisely when these shrimp moved out of the creek in which 

 they were tagged or how long they were present in the 

 creek subsystem where they were ultimately recaptured. 

 For the growth rate analyses, it was assumed that most 

 growth occurred while the shrimp were in the creek and 

 where individuals were marked. The mean (±SD) final 

 size (mm, TL) of individuals that moved between creek 

 subsystems was significantly (separate variance estimate 

 <=2.62, df=16.9, P=0.018) larger (78.9 ±7.4) than that of 

 the group tagged and recaptured in the same subsystem 

 (71.3 ±13.1); the initial mean size of the two groups was 

 nearly identical (57.5 ±10.8 and 57.7 ±10.4, respectively). 

 Two shrimp (at large 7 and 17 days) tagged at EF were 

 recaptured at high tide in flume weirs located 25 m into 



1000 

 800 

 600 

 400 

 200 



F=7.10, df = 1.37 

 P = 0.01, r 2 = 0.16 



40 50 60 70 80 90 100 110 

 Final size (TL. mm) 



Figure 7 



The effect of shrimp size at recapture on mean dis- 

 tance between mark and recapture locations (dis- 

 placement!. Summary results from the linear regres- 

 sion ANOVA performed on the data are shown. Values 

 of mean displacement were based on data from 2-11 

 individuals within each size. 



the interior of the intertidal marsh drained by that tidal 

 creek subsystem. The flume weir samples were part of an 

 ongoing study (Kneib, unpubl. data) to determine nekton 

 use of the intertidal marsh surface (see Kneib, 1991, 1997; 

 Kneib and Wagner, 1994). 



Discussion 



Growth 



Mean growth rates of juvenile white shrimp measured 

 in this study (0.6-0.9 mm/d) were near the lower end of 

 the range of estimates previously reported for juvenile 

 white shrimp along the U.S. Atlantic and Gulf coasts 

 (Table 2). The principal difference between the present 

 and previous studies is that the values presented in this 

 study were based on direct measurements of free-rang- 

 ing individual juvenile shrimp rather than on extrapola- 

 tions from batch mark-recaptures of larger individuals 

 or changes in modal size frequencies. The open nature of 

 estuarine ecosystems, prolonged seasonal recruitment to 

 the nursery, and ontogenetic differences in mortality and 

 movement all may confound the interpretation of size-fre- 

 quency data (Haywood and Staples, 1993). Given that our 

 growth values were based on actual changes in the size 

 of individuals rather than estimated from the apparent 

 growth trajectories of cohorts, we are confident that the 

 mean growth rates reported here accurately reflect those 

 of free-ranging juvenile white shrimp (40-80 mm TL) in 

 the polyhaline portion of the tidal marsh nursery habitat 

 of coastal Georgia. 



Temporal differences in observed growth rates in this 

 study may have resulted from either variation in environ- 

 mental conditions or spatial variation in habitat quality. 



