Ju et al: Demographic assessment of Callinectes sapidus 



313 



to assess the age structure of the blue crab population in 

 the Chesapeake Bay through measurement of lipofuscin 

 and to compare the demographics with those determined 

 by the more traditional size-based approach. 



Materials and methods 



Sample collections 



Winter bottom-dredge surveys (WDS) are conducted annu- 

 ally in the Chesapeake Bay system to monitor blue crab 

 recruitment and size. Surveys were conducted according 

 to a stratified random design, and sampling intensity was 

 apportioned according to depth and site area (for details, 

 see Volstad et al., 2000). Although collections target all 

 sizes of crabs, juvenile crabs (<15 mm carapace width 

 [CW] ) are not fully susceptible to the dredge gear and are 

 underestimated in abundance (Rothschild et al., 1992). 

 Further, lipofuscin (LF) index in small crabs (<40 mm CW) 

 cannot be determined because of analytical limitations (Ju 

 et al., 1999). Only juvenile crabs >40 mm CW were selected 

 for our study. The elimination of crabs smaller than 40 mm 

 CW in our lipofuscin sample should result in under-rep- 

 resentation by juvenile crabs spawned late during the 

 previous spawning season (i.e. August-September) and by 

 other slow growing juveniles. Subsamples were collected 

 from three different regions of Chesapeake Bay: the east- 

 ern shore (Fishing Bay and Honga River), western shore 

 (Potomac River), and lower bay (James River) from Decem- 

 ber through February in 1998-99 and 1999-2000. A portion 

 of juveniles <70 mm CW (n=58) from the 1998-99 survey 

 was excluded for use in a growth and calibration study as 

 recently reported in Ju et al. (2001). 



Analysis of lipofuscin index 



Crabs were anesthetized on ice prior to being sacrificed. 

 Carapace width (mm) was measured and eyestalk tissues 

 were carefully dissected. Each collected tissue was trans- 

 ferred to a 4-mL amber vial for extraction of lipofuscins 

 with a solvent. Measurement of fluorescence intensity was 

 modified slightly from Ju et al. ( 1999) to improve sensitiv- 

 ity and accommodate larger sample numbers by switching 

 from individual sample detection to a scanning fluores- 

 cence spectrophotometer (Waters 474) equipped with a 

 flow cell. Volumes of 10 pL from each extract were injected 

 by an auto-sampler with methanol (MeOH) as the carrier 

 solvent (1 mlVmin). Fluorescence intensity was measured 

 at a maximum emission wavelength of 405 nm by using a 

 maximum excitation at 340 nm at constant temperature 

 (10°C). 



To provide a quantitative measure of lipofuscin, fluores- 

 cence intensities of extracted lipofuscin were calibrated by 

 using quinine sulfate (in O.IN H2SO4) and normalized to 

 protein content of extracted tissue measured by the modi- 

 fied bicinchoninic acid assay ( Nguyen and Harvey, 1994 ). Al- 

 though fluorescence intensity can be accurately calibrated 

 against external standards, the lipofuscin amount per unit 

 tissue volume ( wet or dry weight ) has often been measured 



either before or after extraction (e.g. Ettershank and 

 George, 1984). Such measures are inherently variable and 

 subject to differences among tissue types and processing 

 methods. The use of cellular protein as a basis for measure- 

 ment of extractable lipofuscin concentrations eliminates 

 many previously encountered difficulties. Protein-normal- 

 ized lipofuscin content is expressed as the lipofuscin index 

 or normalized-lipofuscin (/ig-LF/mg-protein). 



Statistical analysis 



The lipofuscin index (^g-LF-content/mg-protein) of 

 samples were natural-log-transformed before statistical 

 analyses to satisfy assumptions of homogeneity of vari- 

 ances and normality of residuals. In order to determine 

 whether the lipofuscin index varied between sexes in each 

 sampling year, analysis of covariance ( ANCOVA) with CW 

 as a covariate was performed. Regression analysis also was 

 performed to compare the relationship between size (CW) 

 and lipofuscin index for each sex in each sampling year 

 Statistical analysis was done with SAS (SAS, Inc., 1996). 



The frequency distribution of CW and lipofuscin index 

 frequencies were analyzed for modal separation. We speci- 

 fied that modal means should be separated by more than 

 two standard deviations (SD) (Gulland and Rosenberg, 

 1992). Lipofuscin index frequency distributions, which 

 showed more than two modes, were analyzed by using 

 ENORMSEP (FiSAT; Gayanilo et al., 1996), which is a 

 maximum likelihood method for identifying modes. A 

 chi-squared analysis was used to test the assumption 

 that frequencies were normally distributed for each mode. 

 Class interval was specified at 10 mm and 0.04 ug lipo- 

 fuscin-content/mg-protein for CW and lipofuscin index, 

 respectively. Although this exceeded the precision of the 

 lipofuscin measurement, it provided class size ranges 

 between 5 and 25 individuals to facilitate the analysis of 

 distribution modes. The age class of each mode was then 

 assigned based on the lipofuscin index accumulation rate 

 determined through rearing experiments (Ju et al., 2001). 

 The formula for this age assignment is 



Agekyr) = 0.824 x exp 



(1,133 » LF index) 



(1) 



Age classes were defined as 0, 1, and 2 and consisted of 

 individuals to <1, 1 to <2, and >2 yr old, respectively. 

 The chronological "age" in this study is related to hatching 

 dates of animals, i.e. the release of the first stage zoea. The 

 age class of each mode identified from modal analysis was 

 assigned according to the lipofuscin index accumulation 

 rate (Eq. 1). Size (CW)-based age classification of 0, 1, 2+ 

 age individuals corresponded to crabs of CW < 60, 60 < 

 CW < 120, and CW > 120 mm, respectively, according to 

 a recent Chesapeake Bay blue crab demographic assess- 

 ments (Rugolo et al., 1998). In our study, age 2+ references 

 all crabs >2 years of age. All crabs of CW > 60 mm (?i=217 

 and 208 for 1998-99 and 1999-2000 WDS, respectively) 

 were included to estimate the relative abundance of each 

 age class determined by either lipofuscin or CW. Although 

 some juveniles (<80 mm CW) were not measured, they 

 were assumed to be <1 yr old (age-0 class). 



