usual community member of this non- 

 vegetated bottom. All other species 

 have occurred often enough to support 

 their inclusion as typical members. 

 Reasons for species having large 

 changes in abundances from one year 

 to the next are not understood, for 

 example, E. gula appears to alternate 

 years of high then low abundance. 



Extreme cold in January and 

 February 1977 (Gilmore et al. 1978) 

 may be responsible, in part, for the 

 two lowest abundances recorded in the 

 dry season. Arius felis , Bagre 

 marinus , Eucinostomus gula , Lagodon 

 rhomboides Bairdiella chrysura , 

 Leiostomus xanthurus and Menticirrhus 

 spp . have been reported as being 

 killed by cold snaps in central and 

 southern Florida (see Gilmore et al. 

 1978) at one time or another. Among 

 these species, only three specimens 

 of Lagodon rhomboides were caught 

 during this period. No dead or dying 

 specimens were observed in the sample 

 area. Ambient water temperature dur- 

 ing January and Febraury 1977 was 

 similar in 1978 (Figure 7) but 

 slightly warmer than in 1976. Among 

 cold-sensitive species, only Men- 

 ticirrhus americanus and Eucinostomus 

 gula were caught in 1976 and again in 

 1978 along with Leiostomus xanthurus , 

 Arius felis, Lagodon rhomboides and 

 Biardiella chrysura. Gilmore et al. 

 (1978) suggest that the rate of cool- 

 ing and duration of cold influences 

 mortality. Even though temperatures 

 may appear similar, based on sampling 

 once or twice a month, daily observa- 

 tions are needed to account for 

 abundance changes that may be related 

 to winter minima. 



Relative abundances generally 

 are lower near the end of the wet 

 season (Table 7). Among the environ- 

 mental changes resulting from high 

 flows, both salinity and dissolved 

 oxygen decrease (Figure 8) . These 

 two factors may be responsible, in 

 part, for low abundances. Each Sep- 

 tember sample produced low abundances 



along with lowest dissolved oxygen, 

 high temperature, and low salinities 

 (Figure 7), as well as low pH and 

 high color (unpublished obs . ) . 



The upper third of the harbor 

 probably experiences severe flow- 

 related changes usually during Sep- 

 tember each year (Table 1 and Figure 

 7). Abundances may be reduced to 

 near zero in higher flow years since 

 the long-term average in September is 

 33 percent greater than the average 

 for 5 years of study. Extreme abun- 

 dance reductions may, in part, be re- 

 lated to previous high flow months, 

 for example, June to September 1976 

 and June to August 1979 (Table 7). 

 Thus, if the flows for a given wet 

 season were similar to the period of 

 record average for each month, we 

 would expect to observe not only a 

 low average abundance for the wet 

 season but also extremely reduced 

 abundances during September. 



Although simple rankings with 

 abundance data suggest no obvious 

 trends (Table 8), grouping on the 

 basis of flow patterns (Figures 1 and 

 2) does suggest a trend. The wet 

 seasons of 1975, 1977 and 1979 had 

 higher abundances than 1976 or 1978. 

 Likewise, the dry seasons of 1977- 

 1978, 1978-1979 and 1979-1980 had 

 higher abundances than 1975-1976 

 or 1976-1977. 



Figures 8 and 9 show these flow- 

 abundance relationships for particu- 

 lar flows and seasonal abundance. 

 When the mean river flow for two 

 short periods of time is plotted with 

 the mean seasonal abundance of the 12 

 taxa, a nearly straight line rela- 

 tionship is formed (R = .95 for both 

 figures). Data in Tables 4 and 5 

 show that no single species is re- 

 sponsible for the trend of the points, 

 A sum of the four most abundant 

 species will produce a similar line 

 in the wet season and the seven most 

 abundant species will produce a simi- 

 lar line in the dry season from 



313 



