SMYTH: CALUNECTES LARVAE IN MIDDLE ATLANTIC BIGHT 



diel change similar to diel patterns of total 

 abundance — except during fall 1976. As distance 

 from shore increased and abundance decreased, 

 however, Callinectes larvae (late zoeae and 

 megalopae) were generally collected at the surface 

 only at night. Ten of 15, and 10 of 12 occurrences 

 (megalopae) during winter and spring 1977 were 

 at night. 



Larvae were taken at salinities ranging from 

 30.5 to 35.8%o and temperatures from 1 1 .0° to 25.7° 

 C (surface temperature and salinity); peak abun- 

 dance occurred in the ranges 31.6-34.9%o and 

 20.6°-25.7° C. Mean temperature, salinity, and 

 distance from shore, weighted for abundance, for 

 all neuston collections of Callinectes larvae were 

 22.9° C, 31.9%o, and 55.9 km. Plots of temperature 

 and salinity vs. abundance indicated no clear rela- 

 tionships among these variables. 



For the independent variables — temperature, 

 salinity, distance from shore, and depth — simple 

 (bivariate) correlation analysis indicated 

 strongest correlation between mean neuston 

 abundance per station and salinity and weakest 

 correlation of abundance with bottom depth 

 (Table 4). 



Table 4. — Simple correlation matrix for surface abundance of 

 Callinectes larvae and selected environmental variables. 



Table 5. — Partial correlation coefficients for surface abundance 

 log,(,[X-i-l]) oi Callinectes larvae with selected environmental 

 variables. 



-•p< o.oi,---p<o.ooi. 



When considered together, the variables form a 

 multivariate population. Partial correlation 

 analysis (Table 5) indicated a very weak relation- 

 ship between bottom depth and larval abundance 

 for all second and third, and most first order corre- 

 lations. Depth was, therefore, deleted from further 

 analysis. Second order correlations among tem- 

 perature, salinity, and distance from shore re- 

 vealed strongest correlation of abundance with 

 temperature, followed by distance from shore and 

 salinity. 



Based on partial correlation analysis, indepen- 

 dent variables were entered into a multiple 

 regression equation in the order temperature, 

 distance from shore, salinity, and depth. These 

 variables explained 66.09^^ of the variation in 

 abundance (Table 6), the maximum possible for 



'c indicates variable which is controlled (effects removed). 



any linear combination of these variables. Depth 

 contributed negligibly to explained variance, and 

 salinity very little (Table 6). The regression equa- 

 tion containing only the variables temperature 

 and distance from shore, explaining 62.4% of the 

 variation in abundance, is 



A = 0.1393 + 0.1124r - 0.0115Z) 



where A = abundance (logio[Z + 1]), 



T = temperature in degrees Celsius, 

 D = distance from shore in kilometers. 



The regression of abundance on temperature and 

 distance from shore was highly significant (Table 

 7). 



The temperature-distance from shore- 

 abundance relationship for all cruises is sum- 

 marized in Figure 5. Summer collections formed a 

 unique group, distributed across the shelf. Abun- 

 dance appeared relatively uniform at least to a 

 distance of 100 km from shore, with a slight in- 

 crease at the outermost stations. Temperature did 

 not appear to be a limiting factor for these summer 

 collections. Relationships are less clear for other 

 seasons. Fall collections generally decreased in 

 abundance with decreasing temperature and in- 

 creasing distance from shore. Winter and spring 

 collections formed groups which were limited to 

 the outer shelf. 



Cooccurring Decapods 



Collections made during periods of peak abun- 

 dance of Callinectes (in the summer) included 



259 



