FISHERY BULLETIN: VOL. 79, NO. 2 



46i 



SALINITY 



''/oo 401 



8'^c 



Figure 5. — Seasonal trends in shrimp 

 8"C at bay sea grass station 30. All 

 samples are composite samples; tem- 

 peratures were taken between 0900 and 

 1200 h on each collection date. 



34 



16i *» • 



-14 



-12 



o o o 



a 



a 



& s 



a f ^ 



60 210 



JULIAN DAY. 1978 



C 



tSHRIMP 

 SPECIES: 



• BROWN 

  PINK 

 o GRASS 

 « ARROW 

 D SNAPPING 



360 



tively). These seasonal differences between means 

 were significant offshore (P<0.01, ^test) but not 

 at sea grass station 50 (P>0.05, ^test ). 



DISCUSSION 



To use 8*^C data to monitor shrimp migrations, 

 isotopically distinct feeding grounds must be iden- 

 tified in which shrimp acquire significantly differ- 

 ent, unique isotopic tags during feeding and 

 growth. When the relationship between shrimp 

 foods and shrimp tissue S^^C is more precisely 

 known, it should be possible to predict shrimp 

 tissue S^^C at any locality from the 8^^C value of 

 locally available shrimp foods (Fry in prep.). This 

 study relied on a more indirect method of identify- 

 ing these isotopically distinct feeding grounds, 

 i.e., monitoring the tissues of the mobile shrimp. 

 In most cases, mean shrimp 8^^C values are proba- 

 bly a reliable index of local feeding conditions, as 

 evidenced by the close 8^^C similarity of migratory 

 brown shrimp and resident nonmigratory shrimp 

 both offshore (Table 2) and in sea grass meadows 

 (Figure 5). 



Tests for significant differences among stations 

 showed that sea gi'ass stations were distinct from 

 open bay stations which were in turn divided into 

 two groups (Figure 3). The 8*^C distinction be- 

 tween sea grass and the more positive group of bay 

 stations has been previously observed (Parker and 

 Calder 1970; Fry et al. 1977; Fry and Parker 1979). 

 These open bay stations occurred throughout the 

 Texas bay system (Figure 3). The more negative 

 open bay stations were found in areas which are 

 more heavily influenced by terrestrial inputs such 



as sewage (stations 100 and 13) or river-borne de- 

 bris (stations in Galveston and San Antonio Bays). 

 Numerous sediment studies have documented 

 that river-borne detritus derived from terrestrial 

 sources averages -25 to -SOX (e.g.. Hunt 1970; 

 Shultz and Calder 1976). Brown shrimp consump- 

 tion of this detritus should result in the observed 

 more negative shrimp 8*^C values. 



Seasonality 



The seasonal cycle observed in shrimp S^^C is 

 roughly correlated with the warmer growing sea- 

 son when the temperature increases and the 

 species composition and abundance of food sources 

 also change. These effects are difficult to separate 

 with the present data. Marine plant species are 

 sometimes enriched in ^^C when grown at higher 

 temperatures (Sackett et al. 1965; Degens et al. 

 1968; Wong and Sackett 1978), although this effect 

 does not appear to consistently apply to sea 

 grasses (Thayer et al. 1978) nor to natural phyto- 

 plankton populations existing at temperatures 

 typical of Texas bays (Sackett et al. 1974). 



Seasonal changes in species composition and 

 abundance occurring in sea grass meadows could 

 also change the mean isotopic composition of 

 shrimp foods; probably, the larger seasonal varia- 

 tions observed in sea grass meadows are due in 

 part to the increasing dominance of sea grass car- 

 bon in food chains leading to shrimp. The lack of 

 correlation between shrimp S^^C values and tem- 

 perature at a time when shrimp S^^C values could 

 change rapidly (mean weight increased 20-50 x 

 over the spring sampling period) argues against a 



342 



