628 



Fishery Bulletin 97(3), 1999 



were probably colder in May of both years at the 

 sample sites (Fig. 2). Temperature, salinity, and dis- 

 solved oxygen were similar among stations (Table 

 1). Visual observations of debris in trawl samples 

 indicated shell deposits at station 1 that were not 

 apparent at other stations. Grain size analysis indi- 

 cated similar sediments between stations 1 and 5, 

 and between stations 2 and 4, whereas station 3 had 

 a significantly higher silt-clay fraction compared with 

 all other stations (ANOVA, P<0.05; Table 1). 



From these nursery habitats, we collected 7507 

 age-0 red snapper from 536 10-min trawl tows in 1994 

 and 1995. Significantly more fish were collected in 

 1995 than in 1994 (ANOVA, P<0.05; Fig. 3). Age-0 

 red snapper first settled to these habitats in late 

 June, showed highest abundance July through Sep- 

 tember, then steadily declined in the fall of both years 

 (ANOVA, P<0.05; F'ig. 3). Most fish, 80-81%, were 

 collected at station 1 (ANOVA, P<0.05: Fig. 4). At 

 station 1, significantly higher peaks in CPUE ±SE 

 for 1994, were 63.6 ±8.9 in July, 60.8 ±26.2 in Au- 

 gust, and 40.3 ±13.4 in September (ANOVA, P<0.05). 

 In late August 1995 at station 1 , we observed a CPUE 

 (712 ±243) that far exceeded all previous estimates 

 from this study or any previous study. Other signifi- 

 cant peaks in CPUE for 1995 at station 1 were 76.2 

 ±21.4 in mid-August, and 81.4 5.0 in mid-September 

 (ANOVA, P<0.05). 



Age-0 red snapper ranged from 17.8 to 124.4 mm 

 SL. Fish first settled from the plankton after they 

 reached 17.8 mm SL. The smallest (<20 mm SL) were 

 present both years into mid-September, after which 

 no new settlers were detected (Fig. 5). Significantly 

 higher mean SLs were detected by late August in 

 both years compared with earlier sample periods; size 

 significantly increased with season (ANOVA, P<0.05; 

 Table 2; Fig. 5). Fish were significantly larger earlier 

 in 1995, but fish from 1994 caught up in size by Sep- 

 tember, afl:er which 1994 fish were significantly larger 

 than 1995 fish (ANOVA, P<0.05; Table 2; Fig. 5). 



In 1994, age-0 red snapper were first abundant 

 (CPUE=63.6) at station 1 and had limited settlement 

 (CPUE=2.0) at station 3 (Fig. 6). Also, fish were sig- 

 nificantly larger (ANOVA, P<0.05), less abundant, 

 and showed up later at all other stations compared 

 with fish at station 1 in 1994. We did not detect pat- 

 terns in inshore-offshore movement from fish size, 

 location, and seasonality; we did determine that af- 

 ter first settlement at station 1, fish showed expan- 

 sion in all directions (Fig. 6). In 1995, new recruits 

 were most abundant in July at station 3 (CPUE=78). 

 After July, patterns were similar to 1994, and most 

 fish were observed at station 1 and fewer larger fish 

 were observed at other stations later in the season 

 (Fig. 7). 



\ ] \ 1 1 i 1 1 1 



MJJASONDJ 



Month 



Figure 2 



Bottom water temperature ( C). salinity (ppt), and dis- 

 solved oxygen (ppml pooled by stations, in the northeast 

 Gulf of Mexico, from June 1994 to January 199.5, and 

 June 1995 to January 1996. Lines without symbols are 

 mean daily surface temperatures based on 10-d moving 

 averages, obtained from a NOAA moored buoy. 



The required precision (<10% ), was shown for 57% 

 of all otolith counts (Table 3). Age of red snapper was 

 26 to 144 days; thus fish may spend up to four months 

 in these habitats. We detected separate late May-June 

 and July cohorts in 1994, and May and June cohorts in 

 1995, ft-om hatching-date frequencies (Fig. 8). 



Growth rates were significantly different among 

 cohorts ( ANCOVA, P<0.05). The fastest growth rates 

 were observed for June (0.77 mm/d) and July (0.71 

 mm/d( cohorts in 1994. Growth rates for May (0.51 

 mm/d) and June (0.67 mm/d) cohorts in 1995 were 

 significantly less than the previous year and signifi- 

 cantly different from each other (ANCOVA, P<0.05; 

 Fig. 9). Growth rate estimates from SL on varying 

 dates (0.52 mm/d in 1994; 0.62 mm/d in 1995; Fig. 5) 

 were significantly lower than most cohort growth 

 estimates, with the exception of the slow growing 

 May 1995 cohort (ANCOVA, P<0.05). In station com- 

 parisons for 1995, fish from station 2 had a signifi- 

 cantly faster growth rate (0.86 mm/d), followed by 

 fish from station 1 (0.71 mm/d), whereas slower 

 growth rates were observed for station 3 (0.54 mm/ 

 d) and station 5 (0.60 mm/d; ANCOVA, P<0.05; Table 



