ABSTRACT Title of Document: FIBER DIGESTION IN THE JUVENILE BLUE CRAB, CALINECTES SAPIDUS RATHBUN Andrea Lauren Alman, Master of Science, 2006 Directed By: Profesor Alen R. Place Center of Marine Biotechnology Five experiments were performed to determine the importance of chitin and celulose in the diet of juvenile C. sapidus. A compartmentalized recirculating system was established to provide optimal conditions, maintaining the animals with litle mortality. The appropriate ration, compartment size, and an adequate baseline diet were established. We replaced 20% of a commercial diet with varying amounts of chitin and celulose. Crabs fed the celulose-containing diet had higher growth rates, conversion eficiencies, molt increments and frequencies than crabs fed the chitin- containing diet, but were equal to the control diet. We then asayed for chitinase and celulase in gut tisues. Chitinase had lower specific activity (0.072 + 0.159 mU mg - 1 min -1 ) than celulase (3.52 + 0.16 mU mg -1 min -1 ) in the foregut and hepatopancreas. There was no efect of diet on specific activity. The results show juvenile C. sapidus is capable of utilizing celulose, but not chitin, when delivered as 20% of a diet. FIBER DIGESTION IN THE JUVENILE BLUE CRAB, CALINECTES SAPIDUS RATHBUN By Andrea Lauren Alman Thesis submited to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfilment of the requirements for the degre of Master of Science 2006 Advisory Commite: Profesor Alen R. Place, Chair Profesor David Secor Anson Hines, Ph. D. ? Copyright by Andrea Lauren Alman 2006 i Acknowledgements NOA and Philips Seafood provided financial support for this project. We would like to thank Odi Zmora and Eric Evans for technical asistance and support. We are indebted to Bridget Bystry, Steven Snopkowski, and Aaron Stoler for asistance with feding crabs. We would also like to thank Jesica Hil for genetic screning of the brood and interpretation of the genetic data. This work was funded by a grant from NOA Award #NA17FU2841 Blue Crab Advanced Research Consortium Project to A.R.P. ii Table of Contents Acknowledgements....................................................i List of Tables.........................................................v List of Figures........................................................vi Chapter 1: Introduction..................................................1 Chapter 2: The efect of cage size and ration on growth of the juvenile blue crab, Callinectes sapidus (Rathbun), in a recirculating system........................12 Summary..........................................................12 Introduction........................................................13 Materials and Methods...............................................15 Experimental Design...............................................17 Growth.........................................................22 Conversion Eficiency..............................................24 Consumption.....................................................25 Genetic analysis...................................................25 Results............................................................26 Growth.........................................................30 Conversion Eficiency..............................................36 Consumption.....................................................38 Genetic analysis...................................................39 Discussion.........................................................40 Chapter 3. The efect of dietary fiber on growth of the juvenile blue crab, Callinectes sapidus (Rathbun), in a recirculating system: comparing four combinations of chitin and celulose............................................................47 Summary..........................................................47 Introduction........................................................48 Methods..........................................................48 Diet Formulations.................................................50 Growth.........................................................52 Consumption.....................................................54 Results............................................................54 Growth: prior to switching diets.......................................56 Growth: after switching diets.........................................59 Consumption.....................................................61 Discussion.........................................................61 Chapter 4. The efect of dietary fiber on growth of the blue crab, Callinectes sapidus (Rathbun), in a recirculating system: using alginate to improve water stability of a peletized diet........................................................67 Summary..........................................................67 Introduction........................................................68 Methods..........................................................68 Diet Formulations.................................................70 Growth.........................................................72 Conversion Eficiency..............................................73 Consumption.....................................................74 iv Results............................................................74 Growth.........................................................76 Conversion Eficiency..............................................80 Consumption.....................................................81 Discussion.........................................................82 Chapter 5. Activity of chitinase and celulase in the gut and hepatopancreas of the juvenile blue crab, Callinectes sapidus (Rathbun), fed diferent diets..............86 Summary..........................................................86 Introduction........................................................86 Methods..........................................................88 Results............................................................92 Discussion.........................................................94 Chapter 6: Discussion.................................................100 The experimental system.............................................100 Experimental diets..................................................100 Rations..........................................................102 Measures of Growth................................................104 Dietary Celulose...................................................106 Dietary Chitin.....................................................107 Endogenous vs. Exogenous Origins.....................................108 Other Discoveries..................................................109 References..........................................................111 v List of Tables Table 1. Esential ion composition mgL -1 of saltwater from our hatchery, compared to literature values of seawater from Bowen (1979)..........................16 Table 2. Ingredients in Ziegler 1/8" peletized shrimp diet, expresed as percent of the total diet, fed to juvenile blue crabs, Callinectes sapidus (Rathbun)............17 Table 3. Water parameters for the course of Experiment 1(21 wek duration) and Experiment 2 (6 wek duration).......................................18 Table 4. Mean disolved oxygen for each cage size and ration combination..........19 Table 5. ean growth values for al juvenile C. sapidus, + standard eror...........27 Table 6. Least-squares mean wet weights (g) over the course of 6 weks, + SEM.....32 Table 7. Growth parameters during molt 1 and 2 for juvenile C. sapidus in experimental system, + SEM. Within each molt, top half shows Experiment 1 and bottom half shows Experiment 2 data............................................34 Table 8. Growth parameters for molts 3-5 for juvenile C. sapidus in experimental system, + SEM..........................................................35 Table 9. ean gross conversion eficiency for juvenile C. sapidus acording to MIH-SR locus in crab genome...............................................40 Table 10. Formulations for juvenile C. sapidus diets, produced by Melick Aquafed (Catawisa, PA). Ingredients expresed in percent by weight. Highlighted portions were varied betwen diets............................................51 Table 11 a, b. Proximate analyses of diets with and without a nutritional oil coating, expresed in mean percentages + SEM. N=3 for al except control, where N=1. ADF performed at a later date than proximates............................52 Table 12. Mean growth values for al juvenile C. sapidus, + SEM.................55 Table 13. ANCOVA performed on final ln(wet weight) using initial ln(wet weight) as a covariate for the first time period (prior to switching diets)...................56 Table 14. Water quality parameters during the course of the experiment............69 Table 15. Proximate analyses of alginate-coated diets, expresed in percent composition + standard eror. N = 3 for each diet...................................71 Table 16. Mean growth values for juvenile C. sapidus, + standard eror.............75 Table 17. ANCOVA for final wet weight, using initial wet weight as a covariate......76 Table 18. Specific activity (mU mg -1 min -1 ) of celulase and chitinase in the gut of C. sapidus fed 20% celulose vs. chitin diets, + SEM. N=5 for hepato., foregut samples; N=1 for other tisues........................................92 Table 19. Statistical results from factorial ANOVA performed on log-transformed specific activity (mU mg -1 min -1 ) of foregut and hepatopancreas samples........93 Table 20. Statistical results from factorial ANOVA performed on log-transformed specific activity (mU mg -1 min -1 ) of pooled midgut, hindgut, and gastric ceca samples.........................................................93 Table 21. Growth parameters for juvenile C. sapidus in large cages of experimental system, + standard eror. From Chapter 2..............................103 Table 22. Intermolt periods and percent increase in carapace width (CW) for juvenile C. sapidus and C. similis: Comparative data from previous and current studies....105 vi List of Figures Figure 1. Rations fed to juvenile blue crabs, Callinectes sapidus (Rathbun), based on the recommended rations for Penaeus shrimp (Guilaume, Kaushik, Bergot, Metailer, 1999). DW= dry weight; W= wet weight..............................20 Figure 2. Layout of recirculating, compartmentalized tank system (not including sump). Numbers indicate ration delivered to each tank half........................21 Figure 3. Wet weight versus carapace width regresion for Experiment 1 and 2 values of intermolt juvenile C. sapidus; N = 207 and 36, respectively. No repeated measures. Standard eror of the estimate = 0.094. Mean values from (Milikin et al. 1980) juxtaposed for comparison...........................................28 Figure 4. Dry weight versus carapace width regresions for Experiment 1 and 2 values of intermolt juvenile C. sapidus; N = 176 and 32, respectively..................29 Figure 5. Wet weights for al juvenile C. sapidus over the course of Experiment 1. Linear regresions are plotted for each ration and cage size tested..............30 Figure 6. Wet weights for al juvenile C. sapidus during Experiment 2, measured every two weks. Linear regresions are plotted for each ration treatment............31 Figure 7. Weight of juvenile C. sapidus for each ration and cage size combination at wek 21. Legend indicates ration fed. Eror bars= LSD. Statistical tests were performed on LSM. Bars with diferent leters = sig. dif. at p < 0.05 (Tukey's HSD)...........................................................32 Figure 8. Mean specific growth rate (log (final weight) ? log (initial weight) time -1 (days) for al rations tested in Experiment 2. Eror bars represent standard eror..36 Figure 9. Mean conversion eficiency for each ration x cage size for juvenile C. sapidus. Test performed on LSM wet weight gained (wet weight fed) -1 ; reported as dry weight (DW gained (DW fed) -1 ). Eror bars = LSD; leters = sig. dif. at p< 0.05 (Tukey's HSD)....................................................37 Figure 10. Mean food conversion eficiency (dry weight gained (dry weight fed) -1 ) for each ration tested. Eror bars represent standard eror......................38 Figure 11. Mean proportion of food consumed in 4 hours for each ration x cage size for juvenile C. sapidus. Legend indicates ration fed. Eror bars = LSD. Statistical tests were performed on LSM. Bars with diferent leters = sig. dif. at p<0.05 (Tukey's HSD)...........................................................39 Figure 12. Wet weight of al juvenile C. sapidus during the experiment, measured 24 hours post-molt. Vertical division in the plot represents switching of diets at day 88. Linear regresions are plotted for each dietary treatment.....................55 Figure 13. Specific growth rate (log(final wt.) ? log (initial wt.) days -1 ) of juvenile C. sapidus in fed diferent diets for the first half of the experiment. Eror bars = SEM. Bars with diferent leters = sig. dif. at p < 0.05 (Tukey's HSD)...............57 Figure 14. Mean time to molt for juvenile C. sapidus fed diferent diets for the first half of the experiment. Eror bars represent standard eror of the mean. For molt 2, bars with diferent leters = sig. dif. at p < 0.05 (Tukey's HSD)...................58 Figure 15. Mean increment (increase in carapace width)(carapace width prior to molt) -1 ) for juvenile C. sapidus fed diferent diets for the first half of the experiment. Eror bars = SEM. For molt 2, bars with diferent leters = sig. dif. at p < 0.05.......59 vii Figure 16. Specific growth rate (log(final wt.) ? log (initial wt.) time -1 (days) and total molts crab -1 of juvenile C. sapidus in diferent diet treatments for the second half of the experiment. Eror bars = SEM. Bars with diferent leters = sig. dif. at p < 0.05 (Tukey?s HSD)....................................................60 Figure 17. Wet weight of al juvenile C. sapidus throughout the experiment, measured 24 hours post-molt. Linear regresions are plotted for each dietary treatment.......75 Figure 18. Mean specific growth rate (log (final wt.) ? log (initial wt.) time -1 (days) for experimental diets. Eror bars = SEM. Bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD)............................77 Figure 19. Mean molt increment (increase in carapace width)(carapace width prior to molt) -1 ) for experimental diets, measured molt-by-molt. Eror bars = SEM. Within each molt: bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD)....................................................78 Figure 20. Mean time to molt for experimental diets, measured molt-by-molt. Eror bars = SE. Within each molt: bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD)..................................79 Figure 21. Mean total molts per crab for experimental diets. Eror bars = SEM. Bars with diferent leters denote significant diferences betwen treatments at p < 0.05 (Tukey?s HSD)....................................................80 Figure 22. Mean food conversion eficiency (dry weight gained (dry weight fed) -1 ) for experimental diets. Eror bars represent standard eror of the mean. Bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD)...81 Figure 23. Mean proportion of diet consumed in 4 hours for diferent diets, taking leaching into acount. Eror bars = SEM. Bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD)............................82 1 Chapter 1: Introduction Stock abundance of the blue crab, Callinectes sapidus (Rathbun), in the Chesapeake Bay (USA) is declining. The baywide catch per unit efort (kg crabs caught per crab pot) declined by nearly 75% betwen 1945-95 (Rugolo, Knotts, Lange, 1998). From 1968-2000, the percent of legal males and the average size of male crabs have declined significantly near Calvert Clifs, Maryland (Abbe, 2002). In the lower Chesapeake Bay, female spawning stock abundance has declined by 81% from 1988- 2000 (Lipcius, Stockhausen, 2002). After the extremely low stock abundance in 2001, the crab population in Chesapeake Bay increased slightly, but remains at historicaly low levels and is stil above the overfished threshold (Commision, 2005). While overfishing has had a large part to play in the decline in stock abundance, it is unclear to what extent other factors are also involved. Many parameters afecting stock abundance are unknown. The efect of toxicants, such as pesticides, fertilizers, heavy metals, and countles other waste products has barely been investigated for blue crabs (Le, Noone, 1995). To understand the influence of such toxicants, we must have a sufficient understanding of the basic biology of the blue crab. Litle is known about the biochemical pathways influencing reproduction (McCarthy, Skinner, 1979), growth, and molting (Le, Watson, 2002; Le, Watson, Roer, 1998; Watson, Le, Borders, Dircksen, Lily, 2001) in C. sapidus, and until these pathways are elucidated there is no way to discern betwen specific environmental triggers and their efects on blue crabs. In addition to the unknown efects of toxicants, there is a lack of understanding about the trophic niche of the blue crab. C. sapidus has been described as a detritivore, omnivore, scavenger, cannibal, and a predator (Clark, Wolcott, Wolcott, Hines, 1999; 2 Laughlin, 1982; Stoner, Buchanan, 1990). Although these studies describe the wide variety of items consumed in the natural habitat, to what extent diferent portions of the crabs? diet contribute to growth is uncertain. Variations in abundance of nutritionaly valuable food items have a direct efect on the stock abundance of a population. In order to more acurately predict fluctuations in stock abundance, the animals? ecological niche must be taken into consideration. Fisheries managers are moving towards an ecosystem approach, taking trophic level, the efects of polutants, and many other factors into acount when determining stock abundances (Panel, 2004). As managers begin to take a multispecies, ecosystem approach for the Chesapeake Bay, providing more basic information about the factors afecting blue crabs would prove extremely useful. To learn more about the blue crab?s basic biology including endocrinology and digestive capabilities, it is necesary to maintain healthy animals in captivity. At the Center of Marine Biotechnology (Baltimore, MD), blue crab hatchery technology is being developed to produce thousands of juvenile crabs of approximately 25mm carapace width (Zmora, Findiesen, Stubblefield, Frenkel, Zohar, 2005). These crabs are involved in field experiments at the Smithsonian Environmental Research Center (Edgewater, MD) to ases the feasibility of stock enhancement (Davis, Young-Wiliams, Hines, Zohar, 2005). In addition to the hatchery, the Center of Marine Biotechnology has nursery and grow-out capabilities in which we can investigate basic biological questions for animals greater than 25 m carapace width. The facilities available are an established compartmentalized recirculating aquaculture system, making full control of the environmental parameters involved in a captive rearing study atainable. In this way, the long-term efects of diet and environmental toxicants can be observed. 3 The maintenance and survival of juvenile crabs in a recirculating system for several months may be dificult, as stresed animals in captivity have a higher frequency of disease (Mesick, Sindermann, 1992). However, much succes has been achieved in maintaining juvenile crabs in recirculating systems, and this approach appears to be a useful avenue for future research (Cadman, Weinstein, 1988; Milikin, Biddle, Siewicki, Fortner, Fair, 1980; Winget, Maurer, Anderson, 1973; Winget, Epifanio, Runnels, Austin, 1976). A number of parameters need to be investigated to provide optimal conditions for the health of animals in a recirculating system. Of the environmental parameters previously studied, temperature, salinity and disolved oxygen have been main focal points. Temperature is crucial in determining growth rates of juvenile crabs. Within a 13- 34 o C range, as temperature increases, molt frequency increases while molt increment (growth per molt) declines (Cadman, Weinstein, 1988; Lefler, 1972; Winget, Epifanio, Runnels, Austin, 1976). Since molt frequency contributes more to the overal growth rate than increment in these studies, growth increases with higher temperature. This needs to be applied with caution however, since increasing temperature also results in a higher mortality rate (Cadman, Weinstein, 1988; Lefler, 1972). Metabolic rate does not increase much from 20-27 o C, indicating that blue crabs have a rather broad thermoneutral range (Lefler, 1972). Salinity may also be important in determining growth of crabs, although it appears to be of les importance than temperature based on the literature. Cadman and Weinstein (1988) found higher growth at a salinity of 30 gL -1 than at 15 gL -1 in juvenile blue crabs. However, Guerin and Stickle (1997) found no diferences in growth rates of juvenile crabs at 2.5, 10, or 30 gL -1 . Blue crabs appear to be relatively sensitive to 4 hypoxic conditions. When exposed to oxygen concentrations of 0.5 mgL -1 , blue crabs have an LT 50 of 2.56 days, lower than the majority of epifaunal species tested from the York River, Virginia (Sagasti, Schafner, Duffy, 2001). When ofered a concentration gradient, blue crabs prefer, on average, 5.2 mgL -1 of oxygen saturation (Das, Stickle, 1994). While temperature, salinity, and disolved oxygen have been extensively studied for C. sapidus, appropriate dietary rations have never been determined. When initialy studying nutritional requirements of a species, the tendency is to overfed or deliver diets ad libitum (Biddle, Milikin, Fair, Fortner, 1978; Milikin, Biddle, Siewicki, Fortner, Fair, 1980; Winget, Epifanio, Runnels, Austin, 1976). This method of feding provides more food than is needed for survival and maintenance, so it may confound growth data for treatments where nutrients are limiting (Tenore, Hanson, McClain, Macubbin, Hodson, 1984). Overfeding is also wasteful and expensive. In various fish species, as rations increase, growth rate increases rapidly and then approaches an asymptote at some maximal value (Wootton, 1998). The maximum growth rate corresponds to a maximal ration, C max . Gross conversion eficiency (total weight gained (total weight fed) -1 ) does not follow the same trend, but reaches a maximum value at some ration les than C max and then decreases as rations approach C max . The ration resulting in the highest conversion eficiency is designated C opt , and provides the most growth per unit of diet delivered to the animals. This level of ration is of particular importance in an aquaculture production system, and should be determined for juvenile C. sapidus. In addition to overfeding, aquaculturists tend to provide a diet with more caloric and protein content than is needed for the species reared. This practice is not only 5 expensive, but it is also damaging to world fish supplies since 1/3 of the world fisheries are harvested for fishmeal for aquaculture and farm feds, and 2-5 times the protein that comes out of intensive aquaculture goes into the fed (Naylor, Goldburg, Primavera, Kautsky, Beveridge, Clay, Folke, Lubchenco, Mooney, Troel, 2000). To determine the composition of a potential diet, basic nutritional requirements need to be determined. In general, crustacean diets need to be more water-stable than fish diets due to their slower feding behavior. Crustaceans tend to pick apart the diet and consume it gradualy over several hours. Water stability must be balanced, however, with the release of smal amounts of substances atractive to the animals, so that the diet remains palatable (New, 1976). A few major aspects of diet have been investigated for blue crabs. Milikin, Biddle, Siewicki, Fortner, and Fair (1980) determined no significant diference in growth betwen crabs fed 37% and 49% protein, however growth was depresed when fed 23% protein diets. Conversion eficiency also tends to decrease in les nutritionaly valuable diets (Wootton, 1998), which has been shown for C. sapidus fed 27% fiber, compared to 9% fiber (Biddle, Milikin, Fair, Fortner, 1978). Since specific nutritional requirements of juvenile crabs are stil largely unknown, feding experiments are an important area for future research. From a generalistic ?control? diet, the specific requirements for C. sapidus can then be fine-tuned and the digestive capabilities defined. In order to define trophic level, the relative importance of fiber in the diet of C. sapidus needs to be determined. Juvenile blue crabs use seagras beds as nursery habitat in the Chesapeake Bay (Heck, Thoman, 1984). Examination of gut contents shows large portions of detrital material, including plant fragments, in young crabs. In a study in the Apalachicola Bay, Florida, detritus and plant mater composed 21.84% (by weight) of the 6 diets of juvenile crabs smaler than 31m carapace width, decreasing with increasing carapace width to 4.74% in animals >60mm carapace width (Laughlin, 1982). Growth rates of juvenile crabs in field enclosures were faster in vegetated than unvegetated substrata in the lower York River, Virginia (Perkins-Viser, Wolcott, Wolcott, 1996). This may have been partly due to the ability of blue crabs to digest maltose, a component of one of the primary complex storage sugars in plant material, starch (?-1,4-glucan) (McClintock, Klinger, Marion, Hsueh, 1991). The fact that blue crabs are able to digest starch suggests that they are not strictly carnivores, but are rather omnivores with capabilities to derive nutritional value from an extensive range of food items. If crabs are able to digest starch, then they may also be able to digest more complex polysacharides. Celulose, followed by chitin, is the most abundant carbon-based biomas worldwide and is the primary source of fiber (McDonald, Edwards, Grenhalgh, Morgan, 2002). In a feding experiment with varying proportions of dietary fiber (celulose), Biddle, Milikin, Fair and Fortner (1978) found similar growth betwen blue crabs fed 3 and 9% celulose, but low survival and depresed growth in crabs fed 27% celulose. The high inclusion rate (9-27%) suggests that crabs may have some ability to derive nutritional value from celulose. Celulose is a polymerized form of glucose with ?-l,4-linkages with a complex structure of crystalized microfibrils among amorphous matrixes. This structure limits aces to most hydrolyzing enzymes. However, some organisms have developed celulases that actively hydrolyze celulose fibrils (Watanabe, Tokuda, 2001). Usualy thre types of enzymes are needed to convert celulose to glucose: endoglucanases that cleave internal ?-1,4-glucosidic bonds, exoglucanases that cleave celobiosyl units at the 7 ends of the celulose molecules, and ?-glucosidase that releases glucose from celobiose and celodextrins (Byrne, Lehnert, Johnson, Moore, 1999). Celulases have been known to be produced by symbiotic protozoans living within the hindgut of termites and catle, but it was not believed until recently that higher animals could produce celulases endogenously (Watanabe, Tokuda, 2001). Celulase has been discovered in a variety of crustaceans and mollusks, including the freshwater prawn Macrobrachium rosenbergii (Gonzalez-Pena, Anderson, Smith, Moreira, 2002), six species of Tasmanian crabs (Johnston, Freman, 2005), and the spiny lobster Jasus edwardsi (Johnston, 2003). The hepatopancreas is the site of endogenous carbohydrase synthesis in crustaceans, which is then released into the foregut in the form of gastric juice (Johnston, Freman, 2005; Johnston, Yelowles, 1998). The cDNA encoding endogenous endoglucanases have recently been discovered and characterized in the red claw crayfish Cherax quadricarinatus (Byrne, Lehnert, Johnson, Moore, 1999; Xue, Anderson, Richardson, Xue, Mather, 1999), the abalone Haliotis discus (Suzuki, Ojima, Nishita, 2003), and in the blue mussel, Mytilus edulis (Xu, Janson, Selos, 2001; Xu, Helman, Erson, Janson, 2000). Al except the mussel endoglucanase belong to the same glycosyl hydrolase family (9) (Henrisat, 1991), and exhibit similar characteristics of being relatively smal proteins (181-594 a.a. residues) lacking celulose-binding domains (with the exception of abalone) and having high thermostability (30-50 o C temperature optima)(Byrne, Lehnert, Johnson, Moore, 1999; Suzuki, Ojima, Nishita, 2003; Xu, Helman, Erson, Janson, 2000; Xue, Anderson, Richardson, Xue, Mather, 1999). Chitin forms the exoskeleton of arthropods in combination with proteins and salts such as calcium carbonate. It consists of monomers of N-acetyl-D-glucosamine, 8 connected by ?(1-4) bonds. In its natural form, it is composed of crystaline fibers embedded in a matrix (much like celulose), which reduces the acesibility of chitinolytic enzymes to the molecule (Esaiasen, Myrnes, Olsen, 1996). At 10 days postmolt, the cuticle of C. sapidus consists of 11.4% chitin (Vigh, Dendinger, 1982) and approximately 73% mineral salts (Cameron, 1985). During molting, chitinase is necesary in order to break down the old cuticle to alow resorption of organic content. Approximately half of the organic content of the shel is absorbed prior to molting (Cameron, Wood, 1985). Because of this, al crustaceans are able to synthesize some level of chitinase (Warner, 1977). In the shrimp Palaemon seratus, chitinase activity in the gut fluctuates with molt cycle (Spindler-Barth, Van Wormhoudt, Spindler, 1990). While we can asume that C. sapidus has some chitinase for resorption of chitin prior to molting, it remains to be determined whether blue crabs can derive nutritional value from the chitin in their diet. Blue crabs have been observed consuming their own exoskeletons after molting, from the larval stages (Sandoz, Rogers, 1944) to juveniles and adults (A. Alman, pers. obs.). Chitin is also consumed when eating other crustaceans and when cannibalizing each other (Laughlin, 1982; Stoner, Buchanan, 1990). Since salts such as calcium carbonate are absorbed directly from the surrounding water via the gils (Cameron, Wood, 1985), the potential value of consuming shed exuviae exists in the resorption of chitin and the various proteins asociated with the molecule. Generaly, two types of chitinolytic enzymes are expresed in crustaceans: chitinase, which degrades chitin to soluble oligosacharides; and ?-N- acetylhexosaminidase, which hydrolyzes the oligosacharides to N-acetylglucosamines (Watanabe, Kono, Aida, Nagasawa, 1998). However, a system of several chitinases is 9 often required to react with chitins of difering substrate specificities (Esaiasen, Myrnes, Olsen, 1996). Endogenous chitinases have ben purified from the hepatopancreas of the prawns Penaeus japonicus and Pandalus borealis (Esaiasen, Myrnes, Olsen, 1996; Kono, Matsui, Shimizu, Koga, 1990; Watanabe, Kono, Aida, Nagasawa, 1998), and from the American lobster Homarus americanus (Lynn, 1990). These chitinases have molecular weights from 37000-150000, with widely varying pH and temperature optima. Depending on the species, the addition of dietary fiber may or may not influence the production of carbohydrases. In the prawn Macrobrachium rosenbergii, celulase activity positively correlated to the level of dietary celulose (Gonzalez-Pena, Anderson, Smith, Moreira, 2002). In the mud crab, Scyla serata, crabs fed 47% starch or celulose had a fourfold higher specific activity than when fed 29% starch (Pavasovic, Richardson, Anderson, Mann, Mather, 2004). However, starvation had litle efect on specific activity of proteases in Penaeus vannamei (Le, Smith, Lawrence, 1984). Species capable of altering their specific activities in response to diet exhibit high phenotypic plasticity. It is important to make the connection betwen growth rates and enzyme activity in order to determine at what level specific activity is no longer important to digestion. There are also many other factors involved in asimilation besides the presence of enzyme, such as food particle size, structural components of the diet inhibiting the availability of substrate to the enzyme, and potential antinutrients asociated with the diet (Francis, Makkar, Becker, 2001). As a result, there is not always a direct link betwen enzyme activity and growth when the substrate is present in the diet. Very few studies cover both growth rates and enzymatic profiles, but linking these two subjects is the only way to determine whether an enzyme is metabolicaly valuable to the animal. 10 A series of experiments were performed in order to define the importance of chitin and celulose in the diet of juvenile C. sapidus, beginning with the determination of the appropriate ration and compartment size. A compartmentalized recirculating system was established to provide optimal conditions acording to previous literature values, and maintained the animals with litle mortality for the time period studied. The first experiment compared thre rations and two compartment size treatments. We predicted that as rations increased, growth would also increase while conversion eficiency peaked, and then dropped. We also predicted that crabs would have higher growth rates and conversion eficiencies when maintained in larger compartments. The appropriate compartment size and an adequate baseline diet were established. To beter model the efect of ration on growth rates, a second experiment was performed. We delivered fractions of the lowest ration from the previous experiment to juvenile C. sapidus. We predicted a continuation of the trends sen in the previous experiment, with decreasing rations resulting in decreasing growth rates and a peak in conversion eficiencies at the C opt . The ?optimal? ration determined from this experiment was not necesarily practical, since rapid growth rates are more cost-efective in our grow-out system. The third and fourth experiments involved formulating diets supplemented with chitin and celulose. We set out to determine if the addition of fiber in the diet afects growth and survival of juvenile C. sapidus. If digestible, the addition of dietary chitin and celulose should have no deleterious efects on growth of juvenile crabs in the laboratory. Since there is evidence that other crustaceans produce endogenous carbohydrases, we predicted no suppresion of growth or decrease in conversion eficiency with the addition of dietary celulose or chitin. 11 In the fifth experiment, gut and hepatopancreas samples from the fourth experiment were asayed for chitinase and celulase activity. This was done to ascertain the presence of enzyme, their location in the gut, and whether dietary composition influenced the concentration of enzyme. We predicted that we would find both chitinase and celulase, since al crustaceans synthesize chitinase for molting, and crabs had similar growth rates when fed celulose-containing diets as when fed the control. We expected to find more celulase activity than chitinase, since crabs had higher growth rates when fed 20% celulose than when fed 20% chitin diets. Since carbohydrases are produced in the hepatopancreas and then secreted into the foregut via gastric juice (Johnston, Freman, 2005; Johnston, Yelowles, 1998), we expected to se the highest specific activities in the hepatopancreas and foregut, with decreased specific activity in other portions of the gut. We expected enzyme activity to vary with diet, with presence of a particular substrate in the diet resulting in higher coresponding enzyme activity. 12 Chapter 2: The efect of cage size and ration on growth of the juvenile blue crab, Callinectes sapidus (Rathbun), in a recirculating system Sumary Blue crab (Callinectes sapidus (Rathbun)) hatchery technology is being developed at the Center of Marine Biotechnology (Baltimore, MD, USA). To more eficiently produce juveniles and adults for research purposes, the appropriate cage size and ration need to be determined. We performed a 21-wek experiment, using a commercial peletized shrimp diet and thre rations representing a range of recommended rations for penaeid shrimp. We developed a compartmentalized recirculating system with two cage sizes (10x10x10cm and 15x15x10cm) to determine if smaler cages limit growth. Final wet weights (using initial weight as a covariate) and gross conversion eficiencies were significantly afected by cage size and ration. Percent of food consumed was significantly diferent among treatments, due to a diference within the smal cages where the intermediate ration was significantly higher than the other two rations. From this data we concluded that smal cages inhibit growth. In addition, al the rations were sufficient to maintain basic metabolic functions, but a sufficient range of rations was not tested to determine the optimum ration (C opt ). Genotyping of the brood revealed evidence of multiple paternity, and the food conversion eficiency was significantly diferent betwen crabs with one of two diferent fathers. Since C opt was not determined from the first experiment, we performed another experiment in which we fed crabs 100%, 75%, 50%, 25% and 0% of the lowest ration from the previous experiment. Crabs in the 25% and 0% ration treatments had consistently lower wet weights than the other treatments throughout the experiment. The 13 specific growth rate (ln(w 1 ) ? ln(w 0 ) time -1 ) and molt increment exhibited the same patern, both increasing as rations increased up to the 50% ration, and then leveling off. The average time required to reach the first molt decreased linearly betwen the 0% and 100% ration treatments. Food conversion eficiency (weight gained (weight fed) -1 ) peaked at the 50% ration, and then decreased with further increase in rations. From this, we concluded the 50% ration to be the C opt for juvenile C. sapidus in our system. Overal, growth rates were lower than in the first experiment. In our system, faster growth is more cost-efective and shows treatment efects more quickly, so a ration higher than C opt is more appropriate. Introduction Nutrition experiments with blue crabs have involved feding ad libitum only (Biddle, Milikin, Fair, Fortner, 1978; Milikin, Biddle, Siewicki, Fortner, Fair, 1980; Winget, Epifanio, Runnels, Austin, 1976). This method of feding provides more food than is neded for survival and maintenance, so it may confound growth data for treatments where nutrients are limiting (Tenore, Hanson, McClain, Macubbin, Hodson, 1984). Proper rations are necesary to evaluate the sufficiency of dietary components in future feding experiments. From an energetic viewpoint, the optimal ration is the ration resulting in the highest gross conversion eficiency (C opt ), providing the most growth per unit of diet fed. This level of ration is normaly les than the ration resulting in the highest growth rate (C max ). Both of these rations should be determined for farmed species, and the ideal ration chosen acording to the needs and goals of the establishment. To date, no studies have investigated the parameter of compartment size and its efects on juvenile crabs. It has been noted that crabs grown in a compartmentalized 14 system have lower growth rates than crabs grown in ponds (Ju, Secor, Harvey, 1999), however, this is not necesarily a result of the size of the container. Other factors such as diet and cannibalism may have a role in enhancing growth in communal ponds. Smal compartment size is known to impinge upon growth in the redclaw crayfish, Cherax quadricarinatus (Manor, Segev, Leibovitz, Aflalo, Sagi, 2002). Putting juvenile blue crabs into separate cages is important since they are extremely cannibalistic starting from the megalopa stage (Moksnes, Lipcius, Pihl, Van Montfrans, 1997), and competition and predation make controlled feding experiments impossible. However, if cage size is limiting in the blue crab, this may obscure treatment efects on growth. While ration and compartment size have not been studied, environmental factors such as temperature and salinity have been extensively investigated for C. sapidus. Both molt frequency and mortality increase with increasing temperature betwen 13 and 34 o C, and metabolic rate changes litle betwen 20-27 o C (Lefler, 1972). The highest molt increment for juvenile crabs occurs at 23 o C, and then decreases as temperature increases (Cadman, Weinstein, 1988). From salinity studies investigating cation concentrations in C. sapidus serum, the approximate isoosmotic point is 27 gL -1 (Colvocoreses, Lynch, Webb, 1974). These values give us a reference from which we can investigate our topic of interest. We set out to determine adequacy of a peletized commercial diet, as wel as to determine an appropriate ration and cage size for optimal growth and survival of these animals in a recirculating system. We started with a basic commercial shrimp diet to evaluate diet adequacy. Diet adequacy was defined by high survival and the ability to maintain normal metabolic functions such as molting. We atempted to determine what 15 rations corresponded to the highest conversion eficiency (C opt ) and highest growth (C max ). We predicted that as rations increased, growth would also increase while conversion eficiency peaked, and then dropped. The system we established had individual compartments to avoid cannibalism during molting. Using two cage sizes, we examined whether growth was more limited in the smaler versus the larger compartments. We predicted that crabs would have higher growth rates and conversion eficiencies when maintained in larger compartments. Since growth and conversion eficiency exhibited linear relationships with ration, we determined that the range of rations tested was not broad enough to show the overal nature of these relationships. We performed a second experiment, lasting for 6 weks, delivering fractions of the lowest ration from the previous experiment to juvenile C. sapidus. We predicted decreased rations to result in decreasing growth rates and a peak in conversion eficiencies at the C opt . Materials and Methods For both experiments, juvenile C. sapidus were maintained in a recirculating system of seven tanks. Each tank held a water volume of 297 L with dimensions 60x90x55 cm; one of these tanks was used as a sump containing mechanical and biological filtration. The biological filtration was KMT (Water Management Technologies, Baton Rouge, LA, USA) media filing approximately 100L of the sump with an established community of denitrifying bacteria. The flow rate from the sump to each tank was approximately 1.3 Lmin -1 , and a 20% water change was performed every thre weks. Approximately 38 tank volumes per day pased through the biological filter. Thre difusers in each tank provided aeration. Water depth in each cage was 16 approximately 5 cm, and the cage bottom was covered with nylon screning material (1 m mesh) to alow flow through the tank while retaining al but particulate food in the cages. We added brine and trace minerals to dechlorinated city water to make the saltwater. Saltwater was formulated as a compromise betwen miicking natural seawater as closely as possible while stil alowing for cost-eficiency (Table 2). The light source was fluorescent lights in the ceiling on a controlled photoperiod, which mirored the daily photoperiod in Baltimore, Maryland. We considered this aceptable since no significant diference was found in growth of crabs when 8-hour versus 16-hour daylengths were tested (Winget, Maurer, Anderson, 1973). Light intensity at the water surface during the day was 300 Lux. Table 1. Esential ion composition mgL -1 of saltwater from our hatchery, compared to literature values of seawater from Bowen (1979). We used a commercial peletized 1/8 inch (3.13 m) diameter shrimp diet (Ziegler Bros, Inc., Gardners, PA, USA). Milikin (1980) found no diference in growth rates of juvenile crabs when fed 37 or 49% protein diets but depresed growth at 23% 17 protein. From this study, we selected 30% protein to be an aceptable level for optimal growth and lowest protein content. The stated protein: fat: fiber ratio, measured in dry weight as fed, was 30: 7: 4. Proximate analyses were performed (New Jersey Fed Lab, Inc., Trenton, NJ, USA) to confirm the stated values, and al were within an aceptable range for our study (Table 1). We chose a shrimp diet because shrimp and blue crabs exhibit similar feding behaviors, both feding intermitently and leaving large portions of their food for later consumption. Therefore, water stability is an important factor in diet choice. However, more water-stable diets may not leach enough nutrients into the surrounding water to make them atractive to the animals, so a balance betwen water stability and palatability is required (New, 1976). The commercial shrimp diet chosen is water stable for several hours, and is a diameter that is easily consumed by the crabs. The crabs consumed it readily, so it appeared to have a good compromise betwen palatability and water stability to met the needs of the animals. Table 2. Ingredients in Ziegler 1/8" peletized shrimp diet, expresed as percent of the total diet, fed to juvenile blue crabs, Callinectes sapidus (Rathbun). Experimental Design 18 Experiment 1: Ration / cage size Two hundred and ten hatchery-reared crabs from a single brood were obtained from the Center of Marine Biotechnology. These crabs were on average 24 m (18-39 m range; standard eror = 0.15 m) carapace width and 1 gram wet weight (0.5-4.5 g range; mean = 1.24 g). Eight weks after commencing the experiment, cages were covered with a plastic grid and atached to the sides of the tank with a crossbar to prevent crabs from escaping. Water temperature was maintained at 21 + 1 o C. After wek 15, temperature was raised to 26 o C to facilitate molting. Salinity was maintained at 28 + 2gL -1 . Amonia, nitrite, and pH levels were measured daily, while nitrate was measured monthly (Table 3). Disolved oxygen was measured in each individual cage on August 28, 2003, September 8, 2003, and January 2, 2004. Mean disolved oxygen values were significantly diferent in the large versus smal cages (Table 4). However, the average values were only 0.2 mgL -1 diferent from each other (6.41 versus 6.65 mgL -1 ), so this is not likely to have any significant physiological efect on the crabs. Table 3. Water parameters for the course of Experiment 1(21 wek duration) and Experiment 2 (6 wek duration). 19 Table 4. Mean disolved oxygen for each cage size and ration combination. Rations were based on the recommended rations for Penaeid shrimp, expresed as a range of percent dry food per gram biomas per day (Guilaume, Kaushik, Bergot, Metailer, 1999). The minimum and maximum recommended values were designated ?low? and ?high? rations respectively, and the median was designated the intermediate ration (Figure 1). Crabs were fed once a day, six days a wek. When exces food acumulated in the cages from previous days, it was removed. 20 Figure 1. Rations fed to juvenile blue crabs, Calinectes sapidus (Rathbun), based on the recommended rations for Penaeus shrimp (Guilaume, Kaushik, Bergot, Metailer, 1999). DW= dry weight; WW= wet weight. The crabs were randomly asigned to treatments of a combination of cage size and ration. Cage sizes were 10x10x10 cm, designated ?smal?, and 15x15x10 cm, designated ?large?. The tank system was aranged as a nested experimental design. Six of the tanks were divided in half, with half of each tank containing large cages and the other half containing smal cages. Due to size constraints, there were 24 smal cages and 12 large cages per tank. Each tank half was randomly designated a ration, so that each combination of ration and cage size was asigned to two tanks (Figure 2). The experiment commenced on August 13, 2003. Wet weights (g) and carapace widths (m) were measured for al crabs every two weks until after wek 12, when they were measured every thre weks. To measure wet weights, crabs were blotted with a paper towel and placed on a scale. Carapace widths were measured across on the dorsal side of the 21 carapace with a ruler, from spine to spine. When a crab molted, the date was recorded. Molts were left in the cages to be consumed by the crabs. If a crab molted on the day of measuring weights and widths, it was left to harden for 24 hours and then measured. Figure 2. Layout of recirculating, compartmentalized tank system (not including sump). Numbers indicate ration delivered to each tank half. On August 28, thirten crabs (including ten crabs that were being held in the sump as ?extras?) were sacrificed and freze-dried to get initial dry weights. The experiment terminated on wek 21, January 9, 2004, and al crabs were removed, sexed, sacrificed, and freze-dried to get dry weights (g). Crabs were sacrificed by submersion in ice water for 5 minutes. Al dry weight values were pooled to make linear regresions determining the relationship betwen dry weight (DW) and wet weight (W), and dry weight and carapace width (CW). Experiment 2: Ration cuting For the second experiment, 36 hatchery-reared crabs from a single brood were obtained from the Center of Marine Biotechnology. These crabs were on average 26 + 1(standard eror) m carapace width (range = 19 - 34 m) and 1.5 + 0.11 grams wet weight (range = 0.54 - 3.13 g). Cages were covered with a plastic grid and atached to 22 the sides of the tank with a crossbar from the onset to prevent crabs from escaping. Prior to both experiments, crabs were aclimated to the experimental salinity and temperature for two weks in a communal tank, and then were moved to the experimental system and alowed to aclimate to tank conditions for two days. Water temperature was maintained at 23 + 0.2 o C. Salinity was maintained at 29 + 0.4 gL -1 . Amonia, nitrite, and pH levels were as before, but disolved oxygen was measured once, on March 11, 2004 ( Table 3). Mean disolved oxygen was 5.9 + 0.3 mgL -1 , an aceptable level acording to literature values (Das, Stickle, 1994), and there was no significant diference in disolved oxygen among rations. The ?low? ration from Experiment 1 was the highest (100%) ration delivered in Experiment 2. The other treatments delivered were fractions (75%, 50%, 25% and 0%) of this ration (Figure 1). Rations were fed to the crabs acording to the crab?s wet weight, in increments of 1 gram. The actual weight of the pelets fed was corrected to acount for 11.71% water weight determined experimentaly. Crabs were placed in 15x15x10 cm ?large? cages only in two adjacent tanks, and randomly asigned one of the five ration treatments. Rations were delivered and animals measured the same way as in Experiment 1. Al of the crabs were weighed and measured at 0, 2, 4, and 6 weks. The experiment commenced on February 10, 2004. On March 24, 2004, crabs were weighed, measured, sexed, and sacrificed by submersion in icewater for 5 minutes. Growth Al data analyses were performed using the SAS system V.8 (SAS 2000). When significant diferences were found in the main efects, pairwise comparisons within cage 23 sizes using Tukey?s HSD were performed. For both experiments, ANOVAs on initial wet weight and carapace width showed no significant diference betwen treatment groups (data not shown). However, the initial size of the animals contributed the majority of the variance in size. For this reason, analysis of covariance (ANCOVA) was performed for the final weights and carapace widths, using initial values as a covariate. Tests for normality and homogeneity of variance were applied to data prior to analysis, and data was transformed when appropriate. Statistical tests were performed on both wet weight and carapace width when both could be applied, but for the sake of brevity, only the methods for using wet weight wil be discussed. Means are reported + the standard eror. If significant diferences were found in growth betwen the treatments by ANCOVA, these diferences were further investigated. Diferences in growth could arise as a result of diferences in molt frequency or molt increment (increase in carapace width (carapace width prior to molt) -1 ). Therefore, time to each molt and percent increase in carapace width (increment x 100%) for each molt was analyzed by ANCOVA. If molting is so infrequent that the total number of molts is les for a treatment, then molt frequency becomes a poor measure of treatment efects. Therefore, the total number of molts per crab was analyzed by ANOVA. Since animals were maintained in a communal tank prior to the start of the experiment, the instar molt is unknown, and may be highly variable from animal to animal. Therefore, al measures of molting are relative to the start of the experiment. Linear measures of growth rates using the specific growth rate (log W 1 ? log W 0 ) time -1 ) was atempted for Experiment 1. However, the frequency distribution of 24 data for virtualy al time increments was bimodal, so the data could not be analyzed statisticaly. Growth was analyzed the same way in the second experiment as in the first, with the following exceptions. Linear measures of growth rates using the specific growth rate (log W 1 ? log W 0 ) time -1 ) had a normal distribution, so it was also used as a parameter for data analysis. Since specific growth rate inherently includes the initial size in the values, it was analyzed by ANOVA. The total number of molts per crab was analyzed by a Chi-squared test. Conversion Eficiency Gross conversion eficiencies were calculated by dividing the total change in wet weight by the total amount of food given per animal over the course of the experiment. An ANOVA was performed on the raw (wet weight) values, followed by pairwise comparisons within each cage size. The conversion eficiencies are also reported as dry weight gained and dry weight fed (DW) values (in grams) since these values are more relevant from an energetic viewpoint. Dry weight gained was calculated from the regresion relationship found betwen wet weight and dry weight, using the following equations (+ standard eror). For Experiment 1: DW = 0.3757(+0.0219)(W) 0.961(+0.0241) ; r 2 = 0.95. For Experiment 2: DW = 0.368 (+ 0.028) (W) 0.826 (+ 0.060) ; r 2 = 0.861. 25 Final calculated dry weight was subtracted from initial calculated dry weight to estimate the dry weight gained. Dry weight fed was calculated from the 11.71% water weight value found in the proximate analysis. Consumption For Experiment 1 only, a 4-hour consumption experiment was performed to determine what portion of the diet the crabs were eating. Rations were weighed and delivered to the crabs in the morning, including thre ?control? rations delivered to empty cages. Al rations were removed four hours later, rinsed with fresh water, and placed in a drying oven at 60 o C for thre days until a constant weight was reached. Then al diets were weighed, and the control diets were subtracted from and then divided by the initial calculated dry weight values. This gave an average value of 4.75(+0.57)% of the dry weight that was lost due to leaching. The proportion of dry weight consumed for the rest of the crabs was determined by the following equation: Proportion consumed = (0.883(+0.002) x ration W) ? (DW x 0.0475(+0.57)) ? leftover DW) (0.883(+0.002) x ration W) ? (DW x 0.0475(+0.57)) -1 A nested ANOVA was performed on the proportion of diet consumed by the crabs. The proportion of diet consumed was not relevant in Experiment 2, since the animals appeared to consume their entire diet within 4 hours. Genetic analysis After terminating Experiment 1, a random sample of sixty-eight of the freze- dried crabs were genotyped for a routine screning of hatchery-reared crabs. One claw from each crab was removed, and DNA was extracted using the FastDNA kit (QBIOgene, Carlsbad, CA, USA). To determine the genotype of a crab or ejaculate, 26 approximately 20 ng of total genomic DNA was used as template in a PCR with one of five microsatelite locus-specific primers (Steven, Hil, Masters, Place, 2005). PCR protocol was followed acording to Hil, Masters, and Place (In Review). The resulting data was analyzed with GeneScan 3.1 software and scored with GenoTyper v2.1 software (Applied Biosystems, Foster City, CA). Results Survival was high for both experiments. Out of 210 crabs in Experiment 1, only 12 crabs died. Thre of these were cannibalized by neighbors that escaped, two were sacrificed for initial dry weights, and thre were taken for use in another experiment, leaving only four that died from unknown causes. Twenty crabs escaped, most of these prior to placing lids on the cages. Five crabs were outside an initial carapace width range of 20-30 m, so they were not included in the data analysis. Out of 36 crabs in Experiment 2, only 4 crabs died during the experiment (1 in the 0% ration, 1 in the 50% ration, and 2 in the 75% ration). Crabs that died from unknown causes or were used in another experiment were not included in the data analysis. When their data was available, crabs that escaped, were sacrificed, or were cannibalized were included in al of the following data analyses because they were considered healthy before their date of death or disappearance. Of the 173 crabs that were present throughout Experiment 1, al of the crabs molted at least twice, 96.5% of the crabs reached the third molt, 63.0% reached the fourth molt, and 7.5% reached the fifth molt. Of the 32 crabs that were present throughout Experiment 2, al of the crabs molted once, and 10 crabs molted twice. For both experiments, the molt frequency decreased with each succesive molt (Table 5). The 27 percent increase in carapace width did not show an obvious trend, but was maintained around 20% for each molt (Table 5). Table 5. Mean growth values for all juvenile C. sapidus, + standard eror. The relationship betwen wet weight and carapace width was plotted for both experiments, and a regresion curve was found an r 2 = 0.99 (Figure 3). The regresion also included the initial (wek 0) values for crabs that escaped, were sacrificed, or were cannibalized by escaped neighbors, and therefore did not have final values. 28 Figure 3. Wet weight versus carapace width regresion for Experiment 1 and 2 values of intermolt juvenile C. sapidus; N = 207 and 36, respectively. No repeated measures. Standard eror of the estimate = 0.094. Mean values from (Milikin et al. 1980) juxtaposed for comparison. Although the relationship betwen wet weight and carapace width is the same betwen the first and second experiments, dry weights tend to fal slightly lower in the second experiment for the same carapace widths (DW = 2.5 X 10 -4 (+ 1.9 x 10 -4 )(CW 2.4(+ 0.21) , r 2 = 0.81) than in the first experiment (DW = 9 X 10 -5 (+ 2 X 10 -5 )(CW 2.74(+0.074) , r 2 = 0.95). There is no significant diference betwen the regresions when analyzed by ANCOVA, but when plotted, a diference in slopes can be visualized (Figure 4). 29 Figure 4. Dry weight versus carapace width regresions for Experiment 1 and 2 values of intermolt juvenile C. sapidus; N = 176 and 32, respectively. Since the regresions had high r 2 values, wet weight, dry weight, and carapace width al appeared to be equal measures in determining statistical diferences betwen treatment groups. Therefore, when al thre could potentialy be the dependent variable, only the results for wet weight values wil be reported. There was no significant diference betwen growth of male and female crabs in any of the statistical tests performed. 30 Growth Size showed increasingly large variance over time in both experiments (Figure 5, Figure 6). Within a brood, animals that started at a range of 0.72-1.52 g (Experiment 1) and 0.54-3.10 g (Experiment 2) ended the experiments betwen 4.68-32.07 g and 1.24- 11.34 g respectively. Figure 5. Wet weights for all juvenile C. sapidus over the course of Experiment 1. Linear regresions are plotted for each ration and cage size tested. 31 Figure 6. Wet weights for all juvenile C. sapidus during Experiment 2, measured every two weks. Linear regresions are plotted for each ration treatment. There was a significant diference betwen ration (Nested ANCOVA, F (4,167) = 3.44, p = 0.01) and cage size (Nested ANCOVA, F (1,167) = 7.59, p = 0.01) treatments at the end of Experiment 1 (Figure 7). In the smal cages, crabs in the low ration were significantly smaler (10.38 + 0.77 g) than crabs in the intermediate ration (13.12 + 1.01 g). In the large cages, crabs in the low ration were significantly smaler (12.83 + 1.22 g) than crabs in the high ration (15.01 + 1.54 g). 32 Figure 7. Weight of juvenile C. sapidus for each ration and cage size combination at wek 21. Legend indicates ration fed. Eror bars= LSD. Statistical tests were performed on LSM. Bars with diferent leters = sig. dif. at p < 0.05 (Tukey's HSD). There was no significant ration efect on size of the animals at the end of Experiment 2 (F (4,26) = 1.67, p = 0.19). But as early as wek 2, crabs fed the 0% and 25% rations had reduced wet weights compared to crabs fed higher rations (Table 6). Table 6. Least-squares mean wet weights (g) over the course of 6 weks, + SEM. 33 Increments and frequencies are similar betwen experiments for the first two molts for al treatments with the exception of the 25% and 0% rations in the second experiment (Table 7). Similar values for crabs in the same treatment (the ?low? ration, ?large? cages in Experiment 1 vs. the 100% ration in Experiment 2) show that growth was consistent betwen the two experiments. In the second experiment, only 10 crabs molted twice (2 in the 25% ration, 5 in the 50% ration, 2 in the 75% ration, and 1 in the 100% ration) (Table 7), so statistics were only performed on data regarding molt 1. The average time required to reach the first molt decreased with higher ration, from 21 (+ 15) days at 0% ration to 7 (+ 4) days at the 100% ration. Despite the lack of significant statistical diference (F (5,27) = 1.22, p = 0.33), the trend is obvious and consistent among al treatments. A trend of larger molt increment with higher ration was also apparent, although not significant (F (5,26) = 1.63, p = 0.19). For the first molt, juveniles increased from 0.12 (+ 0.05) at the 0% ration to 0.21 (+ 0.05) at the 100% ration (Table 7). 34 Table 7. Growth parameters during molt 1 and 2 for juvenile C. sapidus in experimental system, + SEM. Within each molt, top half shows Experiment 1 and bottom half shows Experiment 2 data. Neither molt frequency nor molt increment was significantly diferent among treatments when looking at the overal average values or any of the individual molts except molt 3 in Experiment 1 (Table 8). There were significant ration efects in the ANCOVA model for both time to molt 3 (F (6,169) = 3.78, p = 0.02), and the percent increase in carapace width for molt 3 (F (6,161) = 3.05, p = 0.01). In both the frequency and increment values for molt 3, significant diferences appear betwen the low and intermediate rations within the smal cages. There was also a significant diference in the total number of molts per crab in Experiment 1 (F (5,167) = 2.30, p = 0.047). Tukey?s post- hoc test indicated that this diference was in the large cages, betwen the high ration (4.1 + 0.2) versus the intermediate (3.7 + 0.2) and low ration (3.6 + 0.1). 35 Table 8. Growth parameters for molts 3-5 for juvenile C. sapidus in experimental system, + SEM. Specific growth rate (log (final weight) ? log (initial weight) time -1 ) was examined for Experiment 2. The specific growth rate, despite having no significant ration efect (F (4,27) = 1.76, p = 0.17), had a general trend of increasing as rations increase up to the 50% ration. At the two higher rations, the growth rate appeared to level off or decrease (Figure 8). This may be atributed to the fact that more crabs reached the second molt in the 50% ration than in any other ration. 36 Figure 8. Mean specific growth rate (log (final weight) ? log (initial weight) time -1 (days) for all rations tested in Experiment 2. Eror bars represent standard eror. Conversion Eficiency In Experiment 1, there was a significant treatment efect on food conversion eficiency (F (5,168) = 7.79, p < 0.0001), since eficiency tends to decrease as rations increase (Figure 9). The mean food conversion eficiency was 40% (0.40 + 0.01; wet weight gained (wet weight fed) -1 ). The conversion ratio in the large cages was significantly higher than in the smal cages (F (1,168) = 5.01, p = 0.03). Within the large cages, crabs fed the low ration had a significantly higher eficiency (0.48 + 0.03) than the intermediate (0.40 + 0.02) and high rations (0.40 + 0.03). In the smal cages, crabs fed both the low (0.44 + 0.02) and intermediate rations (0.42 + 0.02) have significantly higher eficiencies than the high ration-fed crabs (0.31 + 0.01). 37 Figure 9. Mean conversion eficiency for each ration x cage size for juvenile C. sapidus. Test performed on LSM wet weight gained (wet weight fed) -1 ; reported as dry weight (DW gained (DW fed) -1 ). Eror bars = LSD; leters = sig. dif. at p< 0.05 (Tukey's HSD). Food conversion eficiency appeared to peak at the 50% ration, and then decrease with additional increase in rations. There was no significant diference in conversion eficiencies (dry weight gained (dry weight fed) -1 ) when comparing al rations in Experiment 2 by ANOVA (F (3,26) = 2.75, p = 0.06). But when the 0% and 100% ration treatments were eliminated (due to the low number of replicates in either treatment), a significant ration efect was found (F (2,24) = 3.64, p = 0.042). The pairwise comparisons revealed this diference to exist betwen the 25% (0.275 + 0.054) and 75% (0.184 + 0.041) ration treatments (Figure 10). The overal average food conversion eficiency in wet weight: wet weight values was 0.86 + 0.12. 38 Figure 10. Mean food conversion eficiency (dry weight gained (dry weight fed) -1 ) for each ration tested. Eror bars represent standard eror. Consumption There was a significant ration efect on the proportion of food consumed (F (4,168) = 3.36, p = 0.01) (Figure 11). This was only apparent in the post-hoc tests for the smal cages, in which crabs fed the intermediate ration consumed a significantly higher proportion of food (0.61 + 0.06) than both the low (0.38 + 0.06) and the high (0.36 + 0.05) rations. In fact, this treatment group had the highest proportion of food consumed than any of the other treatment combinations. Aside from this anomaly, the lack of a significant diference among the other treatments indicates that the animals were not compensating for the quantity of the food given to them by consuming more or les of it. 39 Figure 11. Mean proportion of food consumed in 4 hours for each ration x cage size for juvenile C. sapidus. Legend indicates ration fed. Eror bars = LSD. Statistical tests were performed on LSM. Bars with diferent leters = sig. dif. at p<0.05 (Tukey's HSD). Genetic analysis While it is known that blue crab females can undergo multiple inseminations (Jivoff, 1997), the possibility of multiple fathers in the efective brood has not been investigated. However, a routine DNA screning of the crabs used in Experiment 1 revealed more than two parents. Of the 68 crabs that were genotyped, 53 were succesfully discriminated by determining the aleles present at the MIH-SR locus (Steven, Hil, Masters, Place, 2005). With two aleles contributed from each parent, we would expect four aleles total and four posible combinations of these aleles. However, we found five aleles total and eight possible combinations present at this locus. Analysis of the mitochondrial genome (data not shown), which is pased from mother to offspring, established that the crabs shared the same mother so the extra aleles must come from a 40 second father. This finding indicates that it is possible for a single brood to have multiple paternity. When comparing growth parameters betwen crabs with diferent fathers, a significant diference was found in the food conversion eficiency (oneway ANOVA, F (1,51) = 4.59, p = 0.04). The conversion ratio for 22 crabs with one father was 0.17 + 0.01, while the 31 crabs with the other father was 0.14 + 0.01 (Table 7). Table 9. Mean gross conversion eficiency for juvenile C. sapidus according to MIH- SR locus in crab genome. Discusion Rearing juvenile blue crabs in a compartmentalized recirculating system was succesfully undertaken for a period of 21 weks. This can be dificult to achieve with blue crabs, and high mortality is common. Winget et al. (1976) reported mortalities of up to 50%, and the presence of a herpes-like virus in sick crabs reared in a recirculating system. It is important, however, to maintain animals for several months in feding experiments, since treatment efects may not manifest for months (Milikin, Biddle, Siewicki, Fortner, Fair, 1980). The extremely low mortality in our studies and absence of 41 any indication of disease demonstrated that this is an excelent system for future long- term feding experiments. Growth parameters were similar to literature values for blue crabs reared in captivity. Carapace width and wet weight values were similar to Milikin et al. (1980), with mean values faling roughly along the same regresion curve (Figure 3). The relationship betwen dry weight and carapace width for intermolt crabs had a high r 2 value, which was sen as unlikely by Cadman and Weinstein (1985) due to the large proportion of weight that is carapace. But since our weight to carapace width correlation was so high, we saw no problem in measuring the crabs during intermolt. Like Cadman and Weinstein (1988), we saw no efect of sex on growth of juveniles. Molt frequency was comparable to other studies with similar-sized crabs, both in actual values (about once a month) and in the trend of frequency decreasing with each succesive molt (Cadman, Weinstein, 1988; Guerin, Stickle, 1997; Lefler, 1972; Milikin, Biddle, Siewicki, Fortner, Fair, 1980). Percent increase in carapace width is comparable to published values at 20-27 o C, which fal betwen 12-20% (Guerin, Stickle, 1997; Lefler, 1972; Winget, Epifanio, Runnels, Austin, 1976). The comparison to other captive studies shows that our crabs exhibited normal growth and molting paterns. Growth rates are substantialy lower in laboratory experiments than they are in the field. Crabs in field enclosures at Point Lookout, Maryland starting at approximately 4.0 g wet weight and 35 m carapace width (determined using our alometric relationship betwen wet weight and carapace width) molted twice in 35 days (Brylawski, Miler, 2003). This is a substantialy higher molt frequency than we found for animals of the same size in our study (once every 20-37 days). There are various reasons why animals 42 grow faster in the field. Among them is temperature, which reaches several degres (celcius) higher in the Chesapeake Bay in summer than in our laboratory experiments (Cowan, Boynton, 1996). Preventing activity in the compartmentalized system also decreases metabolic rates, which may depres growth rates in the laboratory compared to the field (Brylawski, Miler, 2003). A more widely varied diet in the field may supply additional esential nutrients to C. sapidus that are largely unknown, which limits our ability to properly formulate diets for laboratory experiments. These various obstacles may be dificult to overcome in rearing studies, and should be further investigated to improve growth in the laboratory. Feding a commercial peletized shrimp diet proved a practical option for rearing C. sapidus juveniles. High survival, combined with growth and molting comparable to those found in other laboratory experiments showed the diet to be adequate for the animals? needs. It is easier to achieve a balanced diet including al esential amino acids, lipids, sterols, vitamins and minerals in a multi-ingredient diet containing diferent protein sources, such as the commercial diet we used (New, 1976). The combination of water stability and palatability of the diet proved efective in providing for the animals? neds. Water stability proved desirable not only because of the crabs? feding behavior of gradualy consuming it over several hours, but it also helped to preserve water quality in the recirculating system (New, 1976). Since the diet was a dry pelet, it may have been les eficient than a moist diet, however, the practicality of having an easily produced and storable diet proved important for a long-term feding experiment. The production cost of the diet was not high, because the company formulated it in bulk. If the intermediate ration was fed to crabs growing at our rates, we could supply food for juvenile crabs 43 starting at 25mm carapace width for five months at the cost of $0.04 per crab. In the future, various aspects of the diet composition should be investigated so that the diet is not providing more energetic and protein content than is neded, as this is wasteful and fish meal is expensive (Naylor, Goldburg, Primavera, Kautsky, Beveridge, Clay, Folke, Lubchenco, Mooney, Troel, 2000). However, using a commercial shrimp diet from the beginning gives us a jump-start for achieving the goal of economic feasibility. Al parameters of growth measured were significantly influenced by ration. This was not a residual efect of diferent percentages of the diet being consumed, most likely because we used low enough rations that the animals were food-limited for al of them. There was a trend towards les and les growth with lower ration, with the 25% and 0% rations showing the most dramatic decreases in wet weight, specific growth rate, increment, and molt frequency when compared to the other rations (Table 6, Table 7, Figure 7, Figure 8). Molt frequency and increment are both influenced by ration when juvenile crabs are seriously deprived of food. Some studies have determined only frequency and not increment to vary in response to diferent levels of dietary protein (Milikin, Biddle, Siewicki, Fortner, Fair, 1980), dietary fiber (Biddle, Milikin, Fair, Fortner, 1978), and salinity (Guerin, Stickle, 1997). Although diferences in increment were apparent in our experiments, they were stil subtle (range: 12-29%) and may be dificult to detect in a lower-resolution study. Depending on the priorities of the grow-out facility, optimal rations can be determined in various ways. From an energetic viewpoint, the ration resulting in the highest conversion eficiency is the best ration to deliver. The food conversion eficiency (dry weight gained (dry weight fed) -1 ) increased rapidly as ration increased up to the 50% 44 ration, and then declined. This conversion eficiency continued to decline from the low to the high rations in Experiment 1, showing a continuum of conversion eficiency betwen the two experiments. We can therefore conclude that the optimal ration to use (C opt ) is the 50% ration in Experiment 2. This results in the most economical use of diets, when diet is the highest cost involved in grow-out. However, we have determined diet to be a fairly inexpensive commodity, while electrical costs including water heating, aeration and water pumps for running the recirculating systems are relatively high. To reduce these costs, the priority is to grow animals as quickly as possible to maturity, and to perform experiments with equal speed. Therefore, the 50% ration is not the ideal ration for our purposes. Instead, we should use the intermediate or high ration from Experiment 1 for future diet studies, as they resulted in the highest growth rates. Even though the diferences in growth were not as obvious for the rations delivered in Experiment 1, they could stil be statisticaly distinguished from one another, so any treatment efects should be apparent in future dietary studies. Juvenile crabs were able to survive and molt, even in starvation conditions. Although the relationship betwen wet weight and carapace width are virtualy the same betwen Experiments 1 and 2, dry weights in Experiment 2 are lower for the same carapace widths. This is a result of les body tisue in animals when fed lower rations, but the water weight in the body cavities masks the loss of tisue when considering wet weight. Therefore, even when animals were surviving and molting, their live tisues may have been degenerating. In C. sapidus, the RNA: DNA ratio (an indicator of protein production and therefore growth) drops by 31% within 3 days of starvation, and by 61% after 30 days of starvation (Wang, Stickle, 1986). In our starvation treatment, one crab 45 molted at the beginning of the experiment, on day 6, while the other did not molt until day 36. This suggests that it is posible that cues for molting proced regardles of nutritional condition. Cage size had a significant efect on growth (Figure 7, Figure 9) and produced apparent diferences in molt frequency. The total number of molts in the smal cages was 3.62, versus 3.75 molts in the large cages. The mean time to molt 3 in the smal cages was 48 days, versus 45 days in the large cages. Within the smal cages, the intermediate ration had both a higher frequency and increment than the low ration. The intermediate ration also had a higher increment than the high ration, which is dificult to explain. Water quality did not sem to be the isue, since disolved oxygen was not significantly diferent betwen any of the rations in the smal cages. It may be that the crabs are inhibited in some way by the combination of plentiful food and limited space. This is very important in future experiments, and needs to be further investigated to determine at what point cage size is no longer a limiting factor in the growth of juvenile crabs. The variable k, representing the ratio of compartment area to carapace width squared (cm 2 cm - 2 ), is a common factor used to indicate when cel size inhibits growth. For the lobster Homarus americanus, growth was inhibited at k < 33 (Van Olst, Carlberg, 1978), while for the crayfish Cherax quadricarinatus, growth was inhibited at k < 50 (Manor, Segev, Leibovitz, Aflalo, Sagi, 2002). In this study, the initial k values for the smal and large cages were 17 and 38, respectively, while the final k values for the smal and large cages were 4 and 8, respectively. If the k that is limiting to growth in blue crabs is similar to that of other crustaceans, this may help explain why a cage efect was significant 46 throughout the experiment. In the future, a larger range of cage sizes should be studied to determine the true limiting k. Other factors besides the ratio of cage size to animal size also come into play when determining what compartment size to ultimately use. Water depth may significantly afect growth in C. quadricarinatus (Manor, Segev, Leibovitz, Aflalo, Sagi, 2002). If this is true for blue crabs, then it may be possible to use a smaler size cage in combination with deeper water to achieve the same growth. We plan to investigate the efect of water depth on the growth of blue crabs in future studies. Practicality may also come into play when determining what cage size to use. When analyzing the overal yield in C. quadricarinatus, smaler compartments produced more mas per unit area than larger compartments even though the individual animals were smaler (Manor, Segev, Leibovitz, Aflalo, Sagi, 2002). Depending on the goals of grow-out production, an aquaculturist may prefer to rear a larger biomas per unit area rather than larger individuals. Evidence of multiple mating was discovered by the presence of two ejaculates in 12.4% of wild females (Jivoff, 1997). Physiologicaly, both ejaculates have aces to the eggs, since the seminal fluid disappears over time, alowing spermatophores to mix in the spermathecae (Jivoff, 1997). However, subsequent fertilization of the eggs, sponge production, hatching, and survival of offspring have never been monitored to determine if the two fathers both played a role in brood production. This study proves that two males can both have offspring from a dual-inseminated female, and in a single brood. 47 Chapter 3. The efect of dietary fiber on growth of the juvenile blue crab, Callinectes sapidus (Rathbun), in a recirculating system: comparing four combinations of chitin and celulose Sumary Recent work indicates that many species of crustaceans have the capability to digest various forms of fiber. Celulose, followed by chitin, composes the largest biomas worldwide, posing an energetic advantage for animals that can digest them. To more eficiently produce juveniles and adults for research purposes, the appropriate diet needs to be formulated. A six-month experiment was performed where we replaced 20% of a commercial diet with varying amounts of chitin and celulose. The diet was coated with nutritional oil to improve water stability. During the first thre months of the experiment, crabs fed 20% chitin had lower wet weights and specific growth rates (ln(w 1 ) ? ln(w 0 ) time -1 ) than crabs fed 20% celulose after the second and third molts, while the 14% celulose / 6% chitin and 14% chitin / 6% celulose diets had intermediate growth rates. Time to molt was longer and molt increment was lower for the 20% chitin diet than for the other diets for thre succesive molts. No crabs fed 20% chitin molted a fourth time. After thre months, we switched the 20% chitin and 20% celulose dietary treatments. During the second half of the experiment, crabs fed the 20% celulose diet had significantly higher specific growth rates (0.00668 + 0.00054) than either crabs fed 20% chitin (0.00273 + 0.00065) or the control diet (0.00318 + 0.00120), indicative of compensatory growth. Only the crabs fed 20% celulose molted a second time. These findings suggest that celulose is digestible in blue crabs, while chitin is either indigestible or is inhibitory if digested. 48 Introduction Biddle, Milikin, Fair and Fortner (1978) performed an ad libitum feding experiment with varying proportions of celulose in the diet of juvenile C. sapidus. Similar growth was found betwen crabs fed 3 and 9% celulose, but there was low survival and depresed growth in crabs fed 27% celulose. In the wild, approximately 20% of the diet of juvenile crabs (< 80 m carapace width) consists of detritus and plant material (Laughlin, 1982; Stoner, Buchanan, 1990). We performed a feding experiment where we replaced 20% of a commercial shrimp diet with chitin and celulose, and delivered the diet acording to appropriate rations (se Chapter 2). We set out to determine if the addition of fiber in the diet afects growth and survival of juvenile C. sapidus. If digestible, the addition of dietary chitin and celulose should have no deleterious efects on growth of juvenile crabs in the laboratory. Methods One-hundred and forty-four juvenile crabs from a single hatchery-reared brood were obtained from the Center of Marine Biotechnology. The crabs averaged 1.04 (+ 0.02) grams wet weight and 23 (+ 0.2) m carapace width. We used a compartmentalized recirculating system consisting of seven tanks (described in Chapter 2). The crabs were placed in the experimental system in 15x15x10 cm ?large? compartments with wals constructed of PVC board. Treatments were asigned randomly to the compartments within each tank, acording to the random complete block design (with the tanks as blocks). Wet weights and carapace widths were measured 24 hours after each molt, and the date of each molt was recorded. 49 Water temperature was maintained at 25 + 0.4 o C as a compromise betwen rapid growth and high survival (Cadman, Weinstein, 1988; Lefler, 1972). Saltwater was formulated as in the ration and cage size experiment (Chapter 2). Salinity was maintained at 28 + 1 gL -1 acording to the approximate isoosmotic point (Colvocoreses, Lynch, Webb, 1974). Fluorescent lights in the ceiling were on a controlled photoperiod equivalent to the photoperiod in Baltimore, Maryland from July 8, 2004 to December 5, 2004. Amonia (0.250 + 0.029 mgL -1 ), nitrite (0.524 + 0.060 mgL -1 ), and pH (8.08 + 0.02) levels were measured daily, while salinity and temperature were measured wekly. Disolved oxygen for this system has ben previously measured as 5.9 + 0.3 mgL -1 at approximately the same temperature and salinity. The experiment was started on July 8, 2004. The crabs were aclimated to the same temperature and salinity (25 o C and 28 gL -1 ) in a communal tank for 1 wek before they were aclimated to the system for 2 days prior to the start date. Five experimental diets were delivered to the crabs, containing the following variations: 20% celulose (by wet weight), 20% chitin, 14% celulose / 6% chitin, 14% chitin / 6% celulose, and a control diet (no chitin or celulose). On September 29 and 30, al crabs in the two diet treatments that contained both chitin and celulose were weighed, measured, and sacrificed. Five crabs randomly sampled from each of the other thre treatments were also measured and sacrificed. Crabs were sacrificed by submersion in ice water for 5 minutes. Half of the sacrificed crabs were freze-dried to obtain dry weights. After removing the two treatments containing a combination of chitin and celulose, the 20% celulose and 20% chitin treatments were switched to determine the efect of changing diets on the growth of the crabs. The experiment was terminated on December 5, 2004, 50 and remaining crabs were weighed, sexed, measured and sacrificed. The animals were then freze-dried for obtaining dry weights. Al dry weight values were pooled to make linear regresions determining the relationship betwen dry weight (DW) and wet weight (W), and dry weight and carapace width (CW). A regresion relationship was also found betwen wet weight and carapace width. Values for alometric relationships were collected at the time of sacrifice, therefore no repeated measures were used. Diet Formulations Experimental diets were formulated by replacing 20% by weight of the ground wheat and fiber portions of a commercial shrimp diet with varying amounts of chitin and celulose (Table 10). The wheat portion of the diet is considered ?filer?, but blue crabs have the ability to digest maltose, a component of starch (?-1,4-glucan) (McClintock, Klinger, Marion, Hsueh, 1991). Therefore, we are replacing a portion of the diet that supplies some nutritional value to the animals, without removing crucial components such as protein. The diets were based on Melick Aquafed?s (Catawisa, PA) 1/8? shrimp pelets, which have a similar formulation to the shrimp diet that was proven succesful in maintaining juvenile blue crabs (Chapter 2). Juvenile C. sapidus readily consumed the Melick shrimp diet. The celulose used was Solka-floc celulose (International Fiber Corporation, North Tonawanda, NY), standard in many dietary formulations. The chitin was made from deproteinated, demineralized blue crab shels (ChitinWorks, Cambridge, MD). Fiber sources were sent to the Melick Aquafed facilities, and the company produced the diets in bulk acording to the formulations. Proximate analyses were performed on the diets by New Jersey Fed Lab, Inc. (Trenton, NJ) (Table 11a), and show the fiber content to be a lot les than was expected, so ADF 51 (acid-detergent fiber) analysis was then performed, resulting in more acurate numbers. This is due to the smal celulose particles (35 um diameter) pasing through the filters used when performing proximate analyses, but they were retained in ADF analysis. Table 10. Formulations for juvenile C. sapidus diets, produced by Melick Aquafed (Catawisa, PA). Ingredients expresed in percent by weight. Highlighted portions were varied betwen diets. The experimental diet, however, was not water-stable and fel apart within 5 minutes of contact with tank water. After testing diferent concentrations, volumes, and proportions of potential stabilizing additives, we developed a protocol for coating diets with nutritional oil to maintain their water stability for a minimum of 2 hours. Pelets were coated with a 1:2 (by volume) combination of pure lecithin ?Yelkinol TS? (Archer Daniels Midland Corp., Decatur, IL) and menhaden oil (Omega Protein, Redvile, VA). Pelets were weighed, aranged in a single layer on a flat baking tray, sprayed using compresed nitrogen with a paint sprayer containing the oil mixture on both sides for a specific amount of time and reweighed, and then placed in a drying oven at 50 o C for 8 hours. The lecithin and menhaden oil coating comprised 13.15% of the wet weight of the initial coated diet, and then was decreased to 7.38% after 6 weks (on August 25, 2004) 52 by cutting the spraying time in half. The final proximate analyses and ADF values after coating with the later amount of oil are shown in Table 11b. When the oil coating was decreased, pelets were water-stable for approximately 30 minutes, after which they visibly fel apart. Crabs readily consumed the oil-coated diet. The majority of the animals, however, left large portions of their diet to be consumed over a period longer than the water-stability of the diet, indicating that the pelets should be made more water- stable in the future. Table 11 a, b. Proximate analyses of diets with and without a nutritional oil coating, expresed in mean percentages + SEM. N=3 for all except control, where N=1. ADF performed at a later date than proximates. The animals were fed the median ration from a range of recommended rations for shrimp of the genus Penaeus (Se Chapter 2) during the wek, but were fed ad libitum on the wekend. Rations are expresed in grams dry weight per 100 grams body weight per day, and vary acording to the crabs? wet (body) weight. The actual weight of the pelets was corrected to acount for moisture content. Growth The SAS system software V.8 was used for data analyses (SAS 2000). When asumptions of normality and homogeneity of variance were not met, data was transformed as appropriate. Data was divided into two time periods and those periods 53 analyzed separately: prior to switching diets (from the start to September 30), and after switching diets (from September 30 to December 5). The size of the animals at September 30 was the ?initial? size for the second half of the experiment, and the number of molts started over at 1 for the second half. This is to prevent confounding of the data due to switching treatments. When a significant diet efect was found, pairwise comparisons were performed betwen diets using Tukey?s HSD at an experimentwise eror rate of p < 0.05. There was no significant diference betwen diets for wet weight (F (4,139) = 0.26, p = 0.91) or carapace width (F (4,139) = 0.22, p = 0.93)at the beginning of the experiment when analyzed by ANOVA. There was a significant diference betwen diets at the beginning of the second half of the experiment, however, for both wet weight (F (2,40) = 17.88, p < 0.0001) and carapace width (F (2,40) = 17.24, p < 0.0001). The majority of the variance in wet weight and carapace width at the end of each time period was related to the size at the beginning of the period. Because of this, analysis of covariance was performed using the size at the beginning of the period as a covariate for the size at the end of the period. Statistical results for wet weight and carapace width were redundant, so only the results for wet weight wil be reported. A linear measure of growth rate, specific growth rate (log W 1 ? log W 0 ) time -1 ) was also used to discern diferences betwen treatments. Since specific growth rate inherently includes the initial size in the values, it was analyzed by ANOVA. If significant diferences in growth betwen diets were found, these diferences were further investigated. Growth can difer as a result of molt increment (increase in carapace width) previous carapace width -1 ) or molt frequency (time to each molt in 54 days). Both of these parameters were found to covary with size prior to molt, so they were analyzed by ANCOVA. The total number of molts per crab was analyzed by ANOVA. Molt numbers are relative to the start of the experiment. Consumption An experiment was conducted to determine if al of the crabs were consuming the same proportion of their diet. On August 31, rations were delivered as normal in the morning, and then removed four hours later with a turkey baster. The food was placed in aluminum cups, rinsed with fresh water, and placed in a 50 o C drying oven until they reached a constant weight (72 hours). The moisture of the initial diets (as measured by the drying oven) was calculated by placing preweighed pelets directly into the drying oven and removing them 72 hours later. Control rations were placed in empty cages to acount for leaching and other los of the diet during imersion in the tanks. The percent of diet consumed could then be calculated by subtracting the final oven-dried diet from the calculated dry weight of the initial diet (subtracting 1.93 (+ 0.26)% moisture), acounting for the average percent of food leached (38 (+ 9)% of the dry weight). The efect of diet on the percentage of food consumed was examined by ANOVA. Results Out of 144 crabs, 25 crabs died during the experiment. Fourten of these died during a molt, and 11 died for unknown reasons. Five crabs escaped, and one crab was taken for use in another experiment. When their data was available, crabs that died or escaped were included in al of the following data analyses because they were considered healthy before their date of death or disappearance. The molt frequency and percent 55 increase in carapace width decreased with each succesive molt (Table 12). Variability in animal size was high throughout the experiment (Figure 12). Figure 12. Wet weight of all juvenile C. sapidus during the experiment, measured 24 hours post-molt. Vertical division in the plot represents switching of diets at day 88. Linear regresions are plotted for each dietary treatment. Table 12. Mean growth values for all juvenile C. sapidus, + SEM. Power regresions provided the best fit for the alometric data when plotted. Wet weight appeared to be a more acurate predictor of dry weight than carapace width, with an r 2 value of 0.82, compared to 0.75. Regresion relationships for calculating dry weight were as follows: DW = 0.402(+ 0.044)W 0.863(+ 0.051) 56 DW = 0.00055(+ 0.00035)CW 2.2(+ 0.17) Wet weight and carapace width, however, had a very close relationship, with an r 2 of 0.94 and the following regresion equation: W = 0.00038 (+ 0.00009) CW 2.6(+0.1) Growth: prior to switching diets Significant diet efects were found (F (4,109) = 18.05, p < 0.0001) using ANCOVA. Crabs in the 20% chitin diet (4.14 + 0.25 g) were significantly smaler than crabs in the 20% celulose diet (9.34 + 0.73 g), the 14% celulose / 6% chitin diet (9.00 + 0.72 g), the 14% chitin / 6% celulose diet (7.47 + 0.55 g), and the control diet (8.78 + 0.87 g) (Table 13). Table 13. ANCOVA performed on final ln(wet weight) using initial ln(wet weight) as a covariate for the first tie period (prior to switching diets). Specific growth rate was significantly diferent among diets (F (4,64) = 4.22, p = 0.004), and the diference existed betwen the 20% celulose (0.0127 + 0.00091) and the 20% chitin diets (0.00794 + 0.00046). There was also a significant diference betwen the 14% celulose/ 6% chitin diet (0.0110 + 0.00037) and the 20% chitin diet (Figure 13). 57 Figure 13. Specific growth rate (log(final wt.) ? log (initial wt.) days -1 ) of juvenile C. sapidus in fed diferent diets for the first half of the experiment. Eror bars = SEM. Bars with diferent leters = sig. dif. at p < 0.05 (Tukey's HSD). When comparing diet efects on molt frequency, a significant diet efect on the time to molt 2 was found (F (4,16) = 4.64, p = 0.002). During the second molt, crabs in the 20% chitin treatment (31 + 2 days) molted les frequently than crabs in the 20% celulose treatment (21 + 1 days), the 14% celulose / 6% chitin treatment (23 + 2 days), and the control (19 + 2 days). In addition, the 14% chitin / 6% celulose treatment (26 + 2 days) was significantly diferent than the control. Although there were no significant diferences for the other molts, a trend was present for molts 1 and 3, in which the time to molt was longer for the 20% chitin diet than for the other diets (Figure 14). In addition, no crabs reached molt 4 for the 20% chitin diet. Only 2 crabs in the control group reached molt 5. 58 Figure 14. Mean time to molt for juvenile C. sapidus fed diferent diets for the first half of the experiment. Eror bars represent standard eror of the mean. For molt 2, bars with diferent leters = sig. dif. at p < 0.05 (Tukey's HSD). There were significant diferences betwen the molt increments of crabs fed diferent diets for molt 2 (F (4,120) = 5.15, p = 0.0007) (Figure 15). The 20% chitin-fed crabs had a significantly smaler molt increment than any of the other treatments (0.19 + 0.02). Although there were no significant diferences for the other molts, the values of the molt increments during molts 1 (0.23 + 0.02) and 3 (0.20 + 0.02) for the 20% chitin diet were les than for the other diets. There was no significant diet efect on the total number of molts per crab (F (4,15) = 2.25, p = 0.07). 59 Figure 15. Mean increment (increase in carapace width)(carapace width prior to molt) -1 ) for juvenile C. sapidus fed diferent diets for the first half of the experiment. Eror bars = SEM. For molt 2, bars with diferent leters = sig. dif. at p < 0.05. Growth: after switching diets There was no significant diference in final wet weights (F (2,38) = 1.20, p = 0.31) betwen treatments using ANCOVA. The size of the animals at the end of the first half of the experiment (range 25-59 m carapace width) acounted for the majority of the variance in the second half of the experiment. There was a significant diet efect on specific growth rate (ANOVA, F (2,37) = 15.04, p < 0.0001). Crabs fed the 20% celulose diet had significantly higher specific growth (0.00668 + 0.00054) than either crabs fed 20% chitin (0.00273 + 0.00065) or the control diet (0.00318 + 0.00120) (Figure 16). 60 Figure 16. Specific growth rate (log(final wt.) ? log (initial wt.) time -1 (days) and total molts crab -1 of juvenile C. sapidus in diferent diet treatments for the second half of the experiment. Eror bars = SEM. Bars with diferent leters = sig. dif. at p < 0.05 (Tukey?s HSD). There was no significant diference betwen diets for frequency (ANCOVA, F (2,26) = 2.56, p = 0.10) or increment (ANCOVA, F (2,29) = 0.05, p = 0.95) during molt 1. There was insufficient data to compare means for molt 2, since only the crabs fed 20% celulose molted a second time. But because of this, the total number of molts per crab was significantly diferent betwen diets (ANOVA, F (2,37) = 11.52, p = 0.0001). Crabs fed the 20% celulose diet had a significantly higher number of molts (1.2 + 0.1) than crabs fed the 20% chitin (0.5 + 0.2) or control diets (0.6 + 0.2) (Figure 16). 61 Consumption There were no significant diet efects on the proportion of food consumed (F (4,125) = 1.71, p = 0.15). The average proportion of food consumed (taking leaching into acount) was 0.73 + 0.03. Discusion Altering the components of a commercialy available fed was a feasible method for diet studies of juvenile blue crabs. The combination of high survival and good growth and molting of the animals showed the diet to be adequate for the animals? needs. By combining water stability and palatability of the diet, we efectively provided for the animals? metabolic requirements. Although the diet was not initialy water-stable, providing a nutritional oil coating improved the stability to a certain extent. Normaly, higher growth rates are correlated to firmer diets (New, 1976), but we actualy observed higher growth rates in this experiment with les stable diets (stability lasting approx. 30 minutes) than in the initial experiment (stability > 4 hours) (Chapter 2). For crabs with the same approximate starting size and the same total number of molts, the average increment was 0.26 + 0.01 for this experiment versus 0.22 + 0.01 for the initial experiment, and the average time to molt was 25 + 1 days for this experiment versus 31 + 1 days for the initial experiment. Since the proximate analyses of the Melick control diet and the Ziegler 1/8? diet (Chapter 2) were virtualy the same aside from fat content, this provides evidence that the nutritional content of the oil coating over-compensated for lack of water-stability. The oil coating improved palatability so that the animals consumed the diet more quickly. It also provided an extra energy source to counteract 62 the loss of nutrients due to leaching, and in fact increased the overal available energy to enhance growth. The two types of nutritional oils were chosen in combination because the menhaden oil has a high omega-3: omega-6 faty acid ratio (approximately 4:1 by weight) while the lecithin has a low ratio (approximately 1:4 by weight). Combining the two ensures that the animals receive a proper balance of omega-3: omega-6 highly unsaturated faty acids, containing slightly more omega-3 than omega-6 faty acids, which is beneficial for growth of shrimp (Guilaume, Kaushik, Bergot, Metailer, 1999; New, 1976). For our study, the ratio of omega-3: omega-6 faty acid was approximately 1.4: 1. The rations fed were sufficient to maintain normal metabolic functions such as growth and molting, but were minimal enough to show diet efects. This confirms that the intermediate ration tested in the initial experiment (Chapter 2) is aceptable for diet experiments. The proportion of the diet consumed was the same among treatments, showing that the growth diferences betwen diets are a result of diet composition, not palatability. In general, crabs in this experiment consumed a higher proportion of their diet (0.73 + 0.03) than in the initial experiment (0.44 + 0.03) (Chapter 2). This may be a result of improved palatability as a result of the addition of nutritional oil to the diet, which may also have resulted in the higher growth rates observed. Diet afected molt increment, frequency, and the total number of molts (Figure 14, Figure 15, Figure 16). For blue crabs, previous studies have determined only frequency and not increment to vary in response to diet (Milikin, Biddle, Siewicki, Fortner, Fair, 1980). In temperature and salinity studies that determined both frequency and increment 63 to vary, molt frequency was more important in determining the overal growth rates of blue crabs than increment (Cadman, Weinstein, 1988; Lefler, 1972). It is posible that the previous diet studies were unable to detect the more subtle diferences in molt increment because the animals were being fed ad libitum instead of specific rations, making the studies les sensitive to treatment efects. The inclusion of 20% and 14% celulose had no efect on growth of juvenile C. sapidus when compared to the control diet (Figure 13, Table 13). The acid-detergent fiber content in the oil-coated control diet is 7.72%, compared to 13.39-19.90% in the experimental diets (Table 11). If the celulose were indigestible, the inclusion of twice as much fiber in the experimental diets would depres growth. In addition, the portion of the diet that we replaced had a high starch component, which blue crabs have the capacity to digest (McClintock, Klinger, Marion, Hsueh, 1991), so we replaced a portion of the diet containing energetic value with a potentialy indigestible substance. This indicates that blue crabs can, in fact, digest celulose. In shrimp, starch and carbohydrates are hydrolyzed to form glucose for use in the Krebs cycle, glycogen storage, chitin synthesis, and the formation of sterols and faty acids (New, 1976). It is possible that juvenile blue crabs utilize complex carbohydrates for these purposes as wel, and is a potential subject for further study. In the wild, juvenile crabs les than 80 m carapace width consume detritus and plant material that constitutes approximately 20% of their diet (Laughlin, 1982; Stoner, Buchanan, 1990). In light of this feding experiment, the consumption of plant material containing starch and celulose may be an energeticaly important part of a juvenile blue crab?s diet in the wild. In a study involving juvenile C. sapidus fed 100% detrital 64 material, the animals grew, however slowly, and had 90% survival in the 3 wek time- frame of the experiment (Ditel, Epifanio, Schwalm, Fantle, Fogel, 2000). Even though these crabs did not grow as quickly as those fed diets containing protein, the study suggests that detritus may provide a secondary source of nutrition when food is scarce. From the wide range of food items consumed by juvenile crabs and the broad descriptions of their dietary niche (Laughlin, 1982; Stoner, Buchanan, 1990), we can safely describe them as omnivorous. Therefore it is not surprising that the animals would grow slowly on 100% detritus. Although they may derive energetic value from celulose and starch in detrital material, detritus does not provide enough of the necesary proteins for growth and building of body mas needed by an omnivorous animal. In Biddle et al. (1978), growth was not depresed at 3% or 9% inclusion of celulose, but was inhibited at 27% inclusion. It is possible that the reduction in growth with a 27% celulose diet was a result of problems with the water-stability of the diet. When the diet contained 27% celulose it fel apart rapidly in water, which may have introduced some bias into the study. This was a problem with the extrusion-based production of the diet, and it may be possible to include more celulose if the procesing is improved. Crabs fed 20% chitin had lower mean wet weights (Table 13), specific growth rates (Figure 13), increments (Figure 15), and frequencies (Figure 14) than crabs fed 20% celulose, while the crabs fed a combination of chitin and celulose had growth parameters in betwen the 20% chitin and 20% celulose treatments. The addition of chitin in the diet has been found to depres growth in tilapia (Shiau, Yu, 1999). A possible reason for this is that chitin is known to interfere with fat digestion and absorption (Deuchi, Kanauchi, Imasato, Kobayashi, 1994). In the current study, that is of 65 particular importance since chitin may have behaved diferently towards the oil coating than celulose. As a result, some bias may have ben introduced into the experiment, and the oil coating may have been diferentialy absorbed betwen treatments. Although the nutritional oil had the advantage of compensating for the rapid leaching and instability of the diet, a neutral stabilizer that truly provides water stability (on the order of hours) should be used in the future. If chitin is indigestible, this provides litle explanation for why crabs consume their shed exuviae. Blue crabs have been observed consuming their own exoskeletons after molting, from the larval stages (Sandoz, Rogers, 1944) to juveniles and adults (A. Alman, pers. obs.). Chitin is also consumed when eating other crustaceans and when cannibalizing each other (Laughlin, 1982; Stoner, Buchanan, 1990). Since salts such as calcium carbonate are absorbed directly from the surrounding water via the gils (Cameron, Wood, 1985), the potential value of consuming shed exuviae exists in the resorption of chitin and the various proteins asociated with the molecule. It is possible that C. sapidus consumes their exoskeleton for the nutritional value of the protein portion only, which is les than 20% of the dry weight (Vigh, Dendinger, 1982). Upon switching the diets, crabs in the 20% celulose treatment molted more and had a higher specific growth rate (Figure 16) than either the 20% chitin or control group. Since they were growing poorly on 20% chitin diet prior to the switch, the acelerated growth beyond that of the control treatment is indicative of compensatory growth. This has been indicated for C. sapidus by Wang and Stickle (1986), who found that the RNA: DNA ratio (starting at 2.92 + 0.09) decreased to 1.15 + 0.09 after 30 days of starvation and then increased to 1.74 + 0.13 within 5 days of feding. The rapid increase in the RNA: DNA 66 ratio after 5 days of feding indicates that the animals have a high capacity for recovery, and can recover at a more rapid rate than the decline in protein production that they experienced upon starvation. Compensatory growth is common in fish (Ali, Nicieza, Wootton, 2003) and has also ben identified in the Chinese shrimp Fenneropenaeus chinensis (Wu, Dong, 2002; Wu, Dong, Wang, Tian, Ma, 2001). That juvenile C. sapidus appears to ?make up? for loss of growth from previous nutrient limitation is an interesting avenue for future research, and may have management implications. 67 Chapter 4. The efect of dietary fiber on growth of the blue crab, Callinectes sapidus (Rathbun), in a recirculating system: using alginate to improve water stability of a peletized diet Sumary The addition of a coating of nutritional oil may have introduced bias into the previous experiment, so we replicated the experiment using a neutral stabilizer instead of oil. We performed a 12-wek experiment, using a commercial peletized shrimp diet for which we replaced 20% of the wet weight (comprised of ground wheat derivatives) with either chitin or celulose, and coated with sodium alginate to provide water stability. We used a recirculating, compartmentalized system that we developed to maintain near optimal water quality values while preventing cannibalism. Specific growth rate (log (final weight) ? log (initial weight) time -1 ) was significantly lower in the chitin-fed crabs (3.77x10 -3 + 0.18x10 -3 ) than in either the celulose-fed crabs (5.81x10 -3 + 0.22x10 -3 ) or the control group (6.20x10 -3 + 0.26x10 -3 ). There was a substantial diference in food conversion eficiency (dry weight gained (dry weight fed) -1 ) betwen crabs fed chitin (0.0750 + 0.0056) and either celulose (0.144 + 0.009) or the control diet (0.161 + 0.017). Al of the growth parameters investigated (molt increment, molt frequency, and total number of molts per crab) were lower in the chitin-fed treatment than the other two treatments. These findings are similar to those of the previous experiment, suggesting that celulose is digestible in blue crabs, while chitin is not. There was no bias introduced in the previous experiment by the usage of nutritional oil, however alginate is the prefered stabilizing additive due to its higher water-stability and energeticaly neutral character. 68 Introduction Chitin is known to interfere with fat digestion and absorption (Deuchi, Kanauchi, Imasato, Kobayashi, 1994), and may have behaved diferently towards the nutritional oil coating in the previous experiment (Chapter 3) than celulose, introducing bias into the experiment. In addition, the oil-coated pelets were only water stable for approximately 30 minutes. Therefore, we set out to replicate the experiment with a coating of some non-nutritive, water-stable substance instead of oil. We performed a 12-wek experiment, using the same commercial peletized shrimp diet, which we coated with sodium alginate and calcium chloride instead of oil. Divalent calcium ions form ionic bonds betwen adjacent alginate molecules, forming a water-stable complex (Meyers, Butler, Hastings, 1972). After eliminating the potential bias from the oil coating, we set out to determine if the addition of fiber in the diet afects growth and survival of juvenile C. sapidus. If digestible, the addition of dietary chitin and celulose should have no deleterious efects on growth of juvenile crabs in the laboratory. Methods Ninety juvenile crabs from a single hatchery-reared brood (167 days old) were obtained from the Center of Marine Biotechnology. The crabs averaged 3.99 (+ 0.11) grams wet weight and 34 (+ 0.3) m carapace width. We used a compartmentalized recirculating system consisting of seven tanks (described in Chapter 2). The crabs were placed in 15x15x10 cm ?large? compartments with wals constructed of PVC board in 5 of the 6 available tanks. Treatments were randomly asigned to the compartments within each tank, acording to the random complete block 69 design (with the tanks as blocks). Wet weights and carapace widths were measured 24 hours after each molt, and the date of each molt was recorded. Water temperature was maintained at 23 + 0.2 o C as a compromise betwen rapid growth and high survival (Cadman, Weinstein, 1988; Lefler, 1972). Saltwater was formulated the same way as in the previous experiments (se Chapter 2). Salinity was maintained at 29 + 1 gL -1 , near the approximate isoosmotic point for C. sapidus, 27 gL -1 (Colvocoreses, Lynch, Webb, 1974). Fluorescent lights in the ceiling were on a controlled photoperiod equivalent to the photoperiod in Baltimore, Maryland from March 14, 2005 to June 28, 2005. Amonia and nitrite were measured daily, while pH, salinity and temperature were measured wekly from the sump. Disolved oxygen was measured on June 8, 2005 in one compartment, chosen randomly, within each tank (N=5) (Table 14). Table 14. Water quality parameters during the course of the experiment. The experiment was started on March 14, 2005. The crabs had been in the experimental system since December 18, 2004, at the same salinity and temperature. Prior to the experiment, they were fed a commercial shrimp diet (Ziegler 3/32? pelets) ad libitum daily. 70 Diet Formulations Formulated diets fel apart within 5 minutes of contact with the tank water. I set out to find a coating for the pelets that provided water stability but was metabolicaly neutral. After testing diferent concentrations of sodium alginate, gum Arabic, gum Ghati, and gelatin, I developed a protocol for spraying diets with sodium alginate and calcium chloride based on the methods of Meyers (1972) to visibly maintain their water stability for a minimum of 8 hours, providing enough time for the animals to consume the entire diet that they were given without losing the soluble nutrients in the diet to leaching. The crabs consumed it readily, showing it to be a palatable diet. Sodium alginate was disolved in deionized water overnight to make a 2% solution (by weight). Calcium chloride was also disolved in deionized water to make a 0.5% solution. Pelets were weighed, aranged in a single layer on a flat baking tray, sprayed on both sides using compresed nitrogen with a paint sprayer for a specific amount of time with sodium alginate solution followed by calcium chloride solution, weighing after each addition. The sodium alginate solution was highly viscous and had to be warmed to 50 o C in order for it to spray evenly on the diets. The pelets were then placed in a drying oven at 55 o C for 8 hours. The alginate and calcium chloride coating comprised 27 (+ 1)% of the wet weight of the coated diets, prior to placing in the drying oven. There were 3.67 x 10 -3 (+ 2.8 x 10 -4 ) g sodium alginate added to every gram of peletized diet, and 9.88 x 10 -4 (+ 9.1 x 10 -5 ) g calcium chloride added to every gram of peletized diet. The diet was sprayed in batches, approximately once a month, and the coated diet was kept refrigerated at 4 0 C to prevent mold spoilage. Triplicate samples of each of the experimental diets were sent to New Jersey Fed Lab, Inc. (Trenton, NJ) for performing 71 proximates and acid-detergent fiber (ADF) analyses (Table 15). Crabs readily consumed the alginate-coated diet. To increase the amount of replicate samples per treatment, I reduced the number of treatments to 3 (compared to Chapter 3), using only the 20% celulose, 20% chitin, and control diets. Table 15. Proximate analyses of alginate-coated diets, expresed in percent composition + standard eror. N = 3 for each diet. The animals were fed the median ration from a range of recommended rations for shrimp of the genus Penaeus (Guilaume, Kaushik, Bergot, Metailer, 1999) (se Chapter 2) 6 days a wek. Crabs were fed double rations the following day when a day was mised. Rations are expresed in grams dry weight per 100 grams body weight per day, and vary acording to the crabs? wet (body) weight. The weight of the pelets fed was corrected to acount for an estimated 10% water weight, which was later found to be 7.96% (determined from the proximates). This was considered unimportant since the rations were fed consistently acording to the estimated water weight, and the diference in rations acording to the estimated versus actual wet weights was never more than 0.01 g day -1 . On June 8, al crabs were weighed and measured to determine the relationship betwen wet weight and carapace width. On June 28, thirty-eight crabs (randomly chosen from each treatment) were weighed, measured, and sacrificed and the experiment terminated. The remaining crabs were used in another experiment. Crabs were sacrificed by submersion in ice water for 5 minutes. Seventen of the sacrificed crabs were freze- 72 dried to obtain dry weights. Twenty-one of the sacrificed crabs (7 from each treatment) were used for gut and hepatopancreas sample collection for future enzyme asays. Dry weight values were used to make linear regresions determining the relationship betwen dry weight (DW) and wet weight (W), and dry weight and carapace width (CW). Values for alometric relationships were collected at the time of sacrifice, therefore no repeated measures were used. Growth Al measures of growth were performed on the data collected from March 14 to June 8, when al of the animals were measured for final values. The SAS system V.8 software was used for data analyses (SAS 2001). When asumptions of normality and homogeneity of variance were not met, data was transformed as appropriate. Tukey?s HSD was performed post-hoc to examine specific diferences betwen treatments, with an experimentwise eror rate of p < 0.05. There was no significant diference in wet weight and carapace width betwen diets at the beginning of the experiment (ANOVA, F (2,86) = 0.49, p = 0.61 and F (2,86) = 0.55, p = 0.58, respectively). Throughout the experiment, the majority of the variance in wet weight and carapace width after each molt was related to the size prior to molting. Because of this, analysis of covariance was performed using the previous size as a covariate for the size at each succesive molt. Statistical results for wet weight and carapace width were redundant, so only the results for wet weight wil be reported. A linear measure of growth rate, specific growth rate (log W 1 ? log W 0 ) time -1 ) was also used to discern diferences betwen treatments. Wet weights measured 73 on June 8 were used as the final wet weights for this parameter. Since specific growth rate inherently includes the initial size in the values, it was analyzed by ANOVA. When dietary treatments significantly afected growth, the diferences in growth were further investigated. Growth can difer as a result of molt increment (increase in carapace width) (previous carapace width -1 ) or molt frequency (time to each molt in days). Both of these parameters were also found to covary with size prior to molt, so they were also analyzed by ANCOVA. The total number of molts per crab was analyzed by ANOVA. Since the instar is unknown, molts are counted relative to the start of the experiment. Conversion Eficiency We calculated food conversion eficiencies by subtracting the initial wet weights from the final wet weights (measured on June 8), and then dividing that by the total amount of food given per animal over the course of the experiment. An ANOVA was performed on the raw (wet weight) values, followed by pairwise comparisons betwen diets (Tukey?s HSD). The conversion eficiencies are also reported as dry weight gained and dry weight fed (DW) values since these values are more relevant from an energetic viewpoint. Dry weight gained was calculated from the regresion relationship found betwen wet weight and dry weight, using the equation below: DW = 0.0922(+ 0.0314)(W) 1.43(+0.13) This regresion had an r 2 of 0.89. Final calculated dry weight was subtracted from initial calculated dry weight to estimate the dry weight gained. Dry weight fed was calculated from the 7.96% water weight value found in the proximate analysis. 74 Consumption On May 23, an experiment was conducted to determine if al of the crabs were consuming the same proportion of their diet (se Chapter 2 for methods). The percent of diet consumed was calculated by subtracting the final oven-dried diet from the calculated dry weight of the initial diet (subtracting 4.62 (+ 0.81)% moisture), acounting for the average percent of food leached (25 (+ 3)% of the dry weight). The efect of diet on the percentage of food consumed was examined by ANOVA. Tukey?s HSD was performed post-hoc to examine specific diferences betwen treatments, with an experimentwise eror rate of p < 0.05. Results No crabs died during the course of the experiment. One crab escaped on June 8. Data prior to June 8 for this crab was used in the data analysis because the crab was considered healthy before it?s date of disappearance. Variability in size of the individuals was high throughout the experiment (Figure 17). When comparing overal average values, the molt frequency decreased betwen the first (20 + 2 days) and second molts (40 + 1 days) and then leveled out at 38 + 1 days during molt 3, while the percent increase in carapace width increased from molt to molt (Table 16). 75 Figure 17. Wet weight of all juvenile C. sapidus throughout the experiment, measured 24 hours post-molt. Linear regresions are plotted for each dietary treatment. Table 16. Mean growth values for juvenile C. sapidus, + standard eror. Based on r 2 values, power regresions provided the best fit for the data. Wet weight appeared to be a more acurate predictor of dry weight than carapace width, with an r 2 value of 0.89, compared to 0.72. Regresion relationships for calculating dry weight were as follows: DW = 0.0922(+ 0.0314)W 1.43(+ 0.13) DW = 4.8x10 -6 (+ 1.0x10 -5 )CW 3.4(+ 0.5) 76 Wet weight and carapace width, however, had a very close relationship, with an r 2 of 0.97 and the following regresion equation: W = 0.0003 (+ 6.4x10 -5 ) CW 2.7(+ 0.1) Growth Significant diet efects were found using ANCOVA (F (2,85) = 26.78, p < 0.0001). Crabs in the 20% chitin diet (8.73 + 0.45 g) were significantly smaler than crabs in the 20% celulose diet (12.64 + 0.61 g) and the control diet (13.63 + 0.91 g) (Table 17). Table 17. ANCOVA for final wet weight, using initial wet weight as a covariate. There was a significant diet efect on specific growth rate (log(final weight) ? log(initial weight) time -1 ) (ANOVA, F (2,85) = 34.5, p < 0.0001). Specific growth rate was significantly lower in the chitin-fed crabs (3.77x10 -3 + 0.18x10 -3 ) than in either the celulose-fed crabs (5.81x10 -3 + 0.22x10 -3 ) or the control group (6.20x10 -3 + 0.26x10 -3 ) (Figure 18). 77 Figure 18. Mean specific growth rate (log (final wt.) ? log (initial wt.) time -1 (days) for experimental diets. Eror bars = SEM. Bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD). Molt increment (increase in CW)(CW prior to molt -1 ) was significantly afected by diet during molt 2 (ANCOVA, F (2,73) = 13.5, p < 0.0001). Crabs in the 20% chitin diet had a significantly lower increment (0.14 + 0.01) than crabs in either the 20% celulose (0.19 + 0.01) or the control treatment (0.20 + 0.01). There was no significant diference in increment betwen treatments during molt 1 (ANCOVA, F (2,84) = 1.18, p = 0.31), but a trend was present where the 20% chitin diet had a smaler increment (0.16 + 0.01) than either the 20% celulose (0.18 + 0.01) or the control (0.17 + 0.01) (Figure 19). 78 Figure 19. Mean molt increment (increase in carapace width)(carapace width prior to molt) -1 ) for experimental diets, measured molt-by-molt. Eror bars = SEM. Within each molt: bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD). Time to molt (days) was significantly diferent among diets during molts 1 and 2 (ANCOVA, F (2.85) = 6.43, p = 0.003 and F (2,69) = 8.16, p = 0.0007, respectively). The post-hoc test indicated that crabs in the 20% chitin diet took significantly longer to reach molt 1 (29 + 4 days) than in either the 20% celulose (18 + 2 days) or the control diets (15 + 2 days). For molt 2, the same patern appears: the 20% chitin-fed crabs took significantly longer to molt (46 + 2 days) than either the 20% celulose-fed (39 + 1 days) or the control-fed crabs (38 + 1 days) (Figure 20). 79 Figure 20. Mean time to molt for experimental diets, measured molt-by-molt. Eror bars = SE. Within each molt: bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD). No crabs in the 20% chitin group molted a third time. This resulted in a significant diference betwen diets in the total number of molts per animal (ANOVA, F (2,87) = 20.4, p < 0.0001). The average number of molts in the 20% chitin group (1.6 + 0.1) was significantly lower than in the 20% celulose (2.2 + 0.1) or the control groups (2.3 + 0.1) (Figure 21). 80 Figure 21. Mean total molts per crab for experimental diets. Eror bars = SEM. Bars with diferent leters denote significant diferences betwen treatments at p < 0.05 (Tukey?s HSD). Conversion Eficiency A significant diference betwen diets was found for conversion eficiency (F (2,86) = 33.9, p < 0.0001). There was a substantial diference in food conversion eficiency (dry weight gained (dry weight fed) -1 ) betwen crabs fed chitin (0.0750 + 0.0056) and either celulose (0.144 + 0.009) or the control diet (0.161 + 0.017) (Figure 22). The average food conversion eficiency in wet weight gained (wet weight fed) -1 values was 0.231 (+ 0.014) and 0.403 (+ 0.019) for the chitin-fed and celulose-fed animals, respectively (Figure 22). 81 Figure 22. Mean food conversion eficiency (dry weight gained (dry weight fed) -1 ) for experimental diets. Eror bars represent standard eror of the mean. Bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD). Consumption Diet had a significant influence on the proportion of food consumed (F (2,85) = 5.51, p = 0.006). Tukey?s HSD indicated that the diference existed betwen the 20% celulose (98.7 + 0.7 %) and 20% chitin (91.1 + 2.2 %) diets (Figure 23). 82 Figure 23. Mean proportion of diet consumed in 4 hours for diferent diets, taking leaching into account. Eror bars = SEM. Bars with diferent leters denote sig. dif. betwen treatments at p < 0.05 (Tukey?s HSD). Discusion Spraying diets with a combination of alginate and calcium chloride was an efective method of improving the water stability of a peletized diet for juvenile C. sapidus. The high survival and growth rates showed the diet to be adequate for the animals? needs. In the literature protocol, the dry diet mixture was submerged in an alginate solution before peletizing (Meyers, Butler, Hastings, 1972), but for our purposes we needed to develop a protocol for coating the diets after they had already been peletized. Using our protocol, the diet retained water stability for at least 8 hours, while remaining palatable to the animals. In this way, we achieved the balance betwen retaining the physical structure of 83 the diet and releasing smal amounts of substances atractive to the crabs, alowing for the crabs? gradual feding behavior with minimal los of nutrients due to leaching. The average proportion of the diet consumed was the highest for this experiment (0.94 + 0.01) than any of the previous experiments, showing the highest palatability within four hours of feding. By improving the stability of the diet, water quality in the recirculating system could also be retained. While the water quality was aceptable in the previous experiment (Chapter 3), there is a relatively low biomas for the volume of water in the system, and use of oil-coated pelets in a more intensive system could potentialy cause problems with water fouling. The use of a dry, commercialy-formulated diet was convenient and practical, and coating the diets with alginate could be performed on large quantities as infrequently as once a month, provided the coated diet was refrigerated. Juvenile C. sapidus growth was suppresed by the addition of chitin, but there were no negative efects upon inclusion of 20% celulose to the diet. Al measures of growth (wet weights, specific growth rate, molt increment, molt frequency, and the total number of molts per crab) were significantly decreased by a 20% inclusion of dietary chitin, but not celulose, in the diet. Since the animals had a larger starting size in this experiment (3.99 + 0.11 g wet weight and 34 + 0.3 m carapace width) and the experiment lasted for 86 instead of 149 days, we could not compare specific values for growth parameters betwen this experiment and the previous experiment. However, when comparing treatment efects, there was not much of a diference in results when using either the oil- coated diet in the previous experiment (Chapter 3) or the alginate-coated diet. The results of the ANCOVA showed the weight of crabs fed 20% chitin to be significantly lower than for 20% celulose or the control in both this experiment and the one previous (Table 84 13, Table 17). In both experiments, the 20% celulose-fed crabs had higher specific growth rates than the 20% chitin-fed crabs, but the same as the control (Figure 13, Figure 16, Figure 18). The time to molt was significantly lower for crabs fed the 20% celulose and control diets than the 20% chitin diet during at least one molt (Figure 14, Figure 20), and the increment was significantly higher for crabs fed the 20% celulose and control diets than the 20% chitin diet during the second molt (Figure 15, Figure 19). This indicates that the oil coating did not obscure or bias the results of the previous feding experiment. Conversion eficiency was significantly decreased in crabs fed the chitin-containing diet (Figure 22). The average food conversion eficiency in wet weight gained (wet weight fed) -1 values was 0.231 (+ 0.014) and 0.403 (+ 0.019) for the chitin-fed and celulose-fed animals, respectively. These values are similar to the conversion eficiency values in a feding experiment with varying proportions of celulose (range 0.23 ? 0.77 wet weight gained (wet weight fed) -1 ) (Biddle, Milikin, Fair, Fortner, 1978). Conversion eficiency appears to be an efective tool in determining diferences betwen dietary treatments, which is another reason to deliver and record specific rations when performing a feding experiment. Much data was lost in the previous experiment as a result of feding the animals ad libitum on the wekend. Unlike in the previous experiment, the proportion of diet consumed was significantly lower for the 20% chitin diet (91.1 + 2.2 %) than the 20% celulose diet (98.7 + 0.7 %) (Figure 23). While this may have acounted for some of the diference betwen dietary treatments, neither treatment had a significantly diferent proportion of the diet consumed than the control. Therefore, this diference in consumption could not acount for the 85 diference betwen the 20% chitin-fed and control-fed crabs regarding wet weight, specific growth rate, molt increment, molt frequency, conversion eficiency, or the total number of molts per crab. The decline in growth and conversion eficiency in the chitin- fed crabs must therefore be atributed to diet composition rather than palatability. The results indicate that celulose is digestible in blue crabs, while chitin is not. While there is evidence that utilization of celulose provides an ecological advantage for juvenile C. sapidus (se discusion Chapter 3), it may not be necesary for them to utilize dietary chitin, if chitin can be synthesized from glucose monomers like it is in shrimp (New, 1976). However, al crustaceans produce some chitinase for reabsorption of chitin from the exoskeleton prior to molting (Warner, 1977). In the shrimp Palaemon seratus, chitinase activity in the integument of the gut is very low except imediately prior to a molt (Spindler-Barth, Van Wormhoudt, Spindler, 1990). However, the chitinase activity remains at a constant, higher level in the shrimp?s hepatopancreas regardles of molt. In order to further investigate this, enzyme asays wil be performed on gut and hepatopancreas samples taken from animals sacrificed at the termination of this experiment. 86 Chapter 5. Activity of chitinase and celulase in the gut and hepatopancreas of the juvenile blue crab, Callinectes sapidus (Rathbun), fed diferent diets Sumary To confirm that the diferences in growth rate were a result of diferences in enzyme activity, chitinase and celulase asays were performed on gut tisue samples from the previous experiment. Specific activity was highest in the foregut and hepatopancreas, with an average of 3.52 + 0.16 (standard eror) mU mg -1 min -1 for celulase, and 0.072 + 0.159 mU mg -1 min -1 for chitinase. The midgut, hindgut and gastric ceca had substantialy lower specific activities overal, and were variable from tisue to tisue. Celulase was higher in the foregut (4.54 + 0.31 mU mg -1 min -1 ) than the hepatopancreas (2.50 + 0.30 mU mg -1 min -1 ), whereas chitinase was higher in the hepatopancreas (0.117 + 0.018 mU mg -1 min -1 ) than in the foregut (0.0270 + 0.0074 mU mg -1 min -1 ). There was no efect of diet on specific activity of the enzymes. The levels of celulase are similar to those of other primarily carnivorous crustaceans. Gastric juice from wild-caught adult C. sapidus had specific activities (celulase = 1.92 mU mg -1 min -1 ; chitinase = 0.104 mU mg -1 min -1 ) similar to those of the juvenile hepatopancreas samples. The significantly lower chitinase activity suggests that chitin is not fully utilized in the diet. Both celulase and chitinase specific activities were substantialy lower in pooled samples from adult C. sapidus, indicating the posibility of ontogenetic shifts in enzyme activity. Introduction Endogenous celulases and chitinases have been discovered in a variety of crustaceans and mollusks, including the prawns Macrobrachium rosenbergii, Penaeus 87 japonicus and Pandalus borealis (Esaiasen, Myrnes, Olsen, 1996; Gonzalez-Pena, Anderson, Smith, Moreira, 2002; Kono, Matsui, Shimizu, Koga, 1990; Watanabe, Kono, Aida, Nagasawa, 1998), six species of Tasmanian crabs (Johnston, Freman, 2005), the lobsters Jasus edwardsi and Homarus americanus (Johnston, 2003; Lynn, 1990), the red claw crayfish Cherax quadricarinatus (Byrne, Lehnert, Johnson, Moore, 1999; Xue, Anderson, Richardson, Xue, Mather, 1999), the abalone Haliotis discus (Suzuki, Ojima, Nishita, 2003), and the blue mussel, Mytilus edulis (Xu, Janson, Selos, 2001; Xu, Helman, Erson, Janson, 2000). While these enzymes have been positively identified in many species, very few studies examine the efects that the presence of these enzymes have on growth rates when the substrate is present in the diet. In a study involving the mud crab Scyla serata, an endogenous celulase was identified that, when fed a diet containing 47% celulose, alowed for the same growth rates as when fed a diet high in protein and lacking celulose (Pavasovic, Richardson, Anderson, Mann, Mather, 2004). To further complicate the isue, there is often an interaction betwen the diet offered and the specific activity of the enzyme. Scyla serata had a fourfold higher specific activity for celulase when fed 47% starch or celulose than when fed 29% starch (Pavasovic, Richardson, Anderson, Mann, Mather, 2004). In the prawn Macrobrachium rosenbergii, celulase activity positively correlated to the level of dietary celulose (Gonzalez-Pena, Anderson, Smith, Moreira, 2002). The hepatopancreas is the site of carbohydrase synthesis in other crustaceans, which is then released into the foregut in the form of gastric juice (Johnston, Freman, 2005; Johnston, Yelowles, 1998). We asayed for chitinase and celulase in the hepatopancreas and diferent portions of the gut of intermolt crabs fed 20% chitin and 20% celulose diets (described 88 in the previous chapter). We then determined the efect of diferent diets on specific activity, the localization of the enzymes in the gut and hepatopancreas, and compared specific activities of chitinase and celulase. We believe we wil find activity for both enzymes, since al crustaceans synthesize chitinase for molting, and crabs had similar growth rates when fed celulose-containing diets as when fed the control. We expect these enzymes to be localized in the hepatopancreas and foregut, with higher activity in the hepatopancreas. We predict higher specific activity of enzymes in those diets containing the substrata that the enzyme degrades. We expect to find more celulase activity than chitinase, since crabs had higher growth rates when fed 20% celulose than when fed 20% chitin diets (Chapters 3 and 4). Methods Ninety juvenile crabs from a single hatchery-reared brood (167 days old) were obtained from the Center of Marine Biotechnology. We performed a 12-wek feding experiment (starting March 14, 2005: se Chapter 4) by using a commercial peletized shrimp diet for which we replaced 20% of the wet weight (comprised of ground wheat derivatives) with either chitin or celulose (se Chapter 4). On June 28, twenty-one crabs (7 intermolt crabs randomly chosen from each treatment) were weighed, measured, and sacrificed for tisue sample collection. Crabs were sacrificed by submersion in ice water for 5 minutes. The following tisue samples were collected from each animal under a disecting microscope: hepatopancreas, foregut, midgut, hindgut, anterior gastric ceca, and posterior gastric cecum. Gut tisues were carefully removed of al contents before collection. Tisues were placed in 1.5 mL Eppendorf tubes and stored at ?80 o C. 89 Frozen tisues were weighed and combined with extraction buffer (80 mM sodium acetate buffer, pH 4.5). The solutions were homogenized with Eppendorf micropestles for 1.5-2 mL tubes for 5 minutes and then centrifuged at 16.1K g for 10 minutes. The supernatant was placed in a clean tube and served as the extract for both the chitinase and celulase asays. For the hepatopancreas samples, a thin layer of lipid on top of the sample required pipeting out the middle layer as the extract. Extracts were used for both enzyme asays the same day that they were made. For the midgut, hindgut, anterior and posterior gastric ceca, the samples were too smal (< 10 mg) to detect enzyme activity and protein for the quantity of extract required in the following asays, so tisues were combined to make one extract per treatment. For the hepatopancreas and foregut, only 5 of the 7 animals disected in the 20% celulose and 20% chitin-fed treatments were used for the folowing enzyme asays. Celulase activity was quantified using a dye-linked substrate, Azo-CM-Celulose (Megazyme, Ireland). Triplicate samples of enzyme extract were added to extraction buffer, making a total of 100 uL of enzyme solution. Substrate solution (100 uL) was then added to the extract, and the mixture was shaken at 25 o C for 2 hours on a thermomixer at 1400 cycles per minute. The reaction was terminated by addition of 500 uL of precipitant solution (prepared acording to the asay protocol), stired on the thermomixer at 1400 cycles per minute for 10 seconds, and alowed to equilibrate at room temperature for 10 minutes. The tubes were then centrifuged at 16.1K g for 10 minutes and 200 uL of supernatant from each tube was placed in a clear 96-wel flat- bottom microplate. The blank was prepared the same as for the above samples, with the exception that the precipitant solution was added imediately after addition of the 90 substrate. The plate was then read on a Vmax Kinetic Microplate Reader (Molecular Devices Corp., Sunnyvale CA) at 595 nm. Optical density (O.D.) values were corrected by subtracting the mean value for the blank from the sample O.D values. Corrected O.D. values were converted to activity units in reference to a standard curve for purified T. resei celulase. One unit of activity (U) is equal to 1 uM of glucose liberated from celulose under the conditions described above. MiliUnits of activity for the 100uL asay volume were equivalent to 148.64(O.D.). Preliminary studies were performed to ensure that the rate of reaction was proportional to time and enzyme concentration at the temperature tested. Chitinase activity was quantified using tritium-labeled chitin as substrate, based on the methods of Molano (1977). The 100ul asay solution consisted of 5 ul of 1M amonium acetate pH 4.5, 15 ul chitin suspension, extract and extraction bufer. Samples were run in triplicate, and the blank excluded extract. The samples were incubated on a thermomixer at 1400 cycles per minute and 25 o C for 2 hours. The constant action of the thermomixer kept the chitin constantly suspended and therefore available for interaction with the asay sample. The enzymatic reaction was stopped and insoluble chitin is precipitated with 300 ul of 10% TCA. The mixture was centrifuged for 5 minutes at 20.8K g and 200 ul of the supernatant was carefully removed so as not to disturb the chitin pelet. The supernatant was combined with 400 uL of BioSafe I cocktail (RPI, IL) in a liquid scintilation tube. Disintegrations per minute (DPM) for each sample was read on the Liquid Scintilation Counter 1801 (Beckman Coulter Inc., CA). DPM was then converted to ug N-acetyl-D-glucosamine produced by measuring the DPM for a known amount of tritium-labeled chitin incubated with enzyme overnight, so that al of the 91 substrate was digested. One unit of N-acetyl-D-glucosamine was equivalent to 1036 DPM. One unit of activity (U) is equal to 1 uM of chitobiose liberated from chitin under the conditions described above. Protein was quantified using the Micro BCA Protein Asay Kit (#23235; Pierce, IL). The 200ul asay solution consisted of 100 ul working reagent, extract, and extraction buffer. Samples were run in triplicate, and the blank excluded extract. A set of BSA standards was run on each plate, ranging from 4-20 ug protein. After addition of the working reagent, the plate was mixed on a plate shaker for 30 seconds, covered and incubated at 37 o C for 2 hours. It was then cooled to room temperature and measured on the Vmax Kinetic Microplate Reader (Molecular Devices Corp., Sunnyvale CA) at 540 nm. Raw O.D. values were corrected by subtracting the average value for the blank from the sample O.D value. Specific activity is expresed in mU mg protein -1 min -1 . The means of the triplicate samples were used for the following data analyses. A factorial ANOVA was performed on specific activity of hepatopancreas and foregut samples, with tisue type (foregut versus hepatopancreas), diet (20% celulose versus 20% chitin), and enzyme (celulase and chitinase) as factors. A factorial ANOVA was performed on the pooled samples for the other four tisue types, comparing the efects of diet and enzyme on specific activity. For both factorial ANOVAs, specific activity was log-transformed to met homogeneity of variances and normality asumptions. Due to log-transformations, one sample from the foregut and hepatopancreas samples and two samples from the midgut, hindgut, and gastric ceca samples obtained values that were negative, and so were omited from the data analysis. 92 Results Specific activity was highest in the foregut and hepatopancreas, with an average of 3.52 + 0.16 (standard eror) mU mg -1 min -1 for celulase, and 0.072 + 0.159 mU mg - 1 min -1 for chitinase. The midgut, hindgut and gastric ceca had substantialy lower specific activities overal, and were variable from tisue to tisue (Table 18). This may have occurred because the total amount of material was so smal (< 2 mg after combining samples) that the activities and protein levels were about equal to the blank, and were therefore dificult to measure acurately. Table 18. Specific activity (mU mg -1 min -1 ) of celulase and chitinase in the gut of C. sapidus fed 20% celulose vs. chitin diets, + SEM. N=5 for hepato., foregut samples; N=1 for other tisues. There was a significant model efect in the factorial ANOVA run for hepatopancreas and foregut samples (F (7,31) = 100, p < 0.0005; Table 19). A significant diference existed betwen tisue types (F (1,31) = 6.94, p = 0.013), but this was confounded by a significant interaction betwen tisue type and enzyme type (F (1,31) = 43.4, p < 0.0005). The interaction efect is due to converse activities of the enzymes in the diferent tisues. Celulase was higher in the foregut (4.54 + 0.31 mU mg -1 min -1 ) 93 than the hepatopancreas (2.50 + 0.30 mU mg -1 min -1 ), whereas chitinase was lower in the foregut (0.0270 + 0.0074 mU mg -1 min -1 ) than the hepatopancreas (0.117 + 0.018 mU mg - 1 min -1 ). Table 19. Statistical results from factorial ANOVA performed on log-transformed specific activity (mU mg -1 min -1 ) of foregut and hepatopancreas samples. There was a significant model efect in the factorial ANOVA for midgut, hindgut, and gastric ceca pooled samples (F (3,10) = 8.15, p = 0.005; Table 20). The model efect was atributed to a significant main efect of enzyme type (F (1,10) = 21.0, p = 0.001). Celulase had a mean specific activity of 0.636 + 0.205 mU mg -1 min -1 , while chitin had a mean specific activity of 0.0206 + 0.205 mU mg -1 min -1 . Table 20. Statistical results from factorial ANOVA performed on log-transformed specific activity (mU mg -1 min -1 ) of pooled midgut, hindgut, and gastric ceca samples. We were also able to isolate gastric juice from the foreguts of 5 crabs (purchased at a local farmers market), which we pooled and analyzed for specific activities of chitinase and celulase. Celulase activity (1.92 mU mg -1 min -1 ) was approximately 94 twenty-fold the level of chitinase activity (0.104 mU mg -1 min -1 ) in the gastric juice. Specific activity was 1.47 mU mg -1 min -1 for celulase and 0.0638 mU mg -1 min -1 for chitinase in the hepatopancreas of the same animals. Discusion We found both chitinase and celulase activity in al of the tisues examined, as wel as in the hepatopancreas and gastric juice of adult wild-caught crabs (Table 18). Previously, McClintock et al. (1991) found low but detectable amounts of B-glucosidase activity in C. sapidus hepatopancreas (0.0007 + 0.0001 mU mg -1 min -1 ). While celobiose was used as a substrate in the study, larger oligomers and celulose chains were not tested. Our study shows a much higher specific activity for celulase (Table 19, Table 20), indicating that the celulase(s) in C. sapidus cleave glucose monomers from a mixture of complex substrates. To our knowledge, no previous studies have asayed for chitinase activity in blue crabs. Celulase activity was comparable to levels for other crustaceans. In hepatopancreas samples from six species of Tasmanian crabs ranging from carnivorous to herbivorous, celulase activity ranged from 1.4 ? 19 mU mg -1 min -1 respectively (Johnston, Freman, 2005). Celulase activity in juvenile C. sapidus hepatopancreas (2.50 + 0.30 mU mg -1 min -1 ) fals within this range, slightly higher than the celulase activity of primarily carnivorous crabs. However, C. sapidus celulase activity is approximately equal to celulase activity in the hepatopancreas of the juvenile spiny lobster Jasus edwarsi (2.1 + 0.6 mU mg -1 min -1 ), also a consumer of primarily animal material (Johnston, 2003). In the spiny lobster, protease activities were 3-4 orders of magnitude higher than carbohydrases. Conversely, celulase activity in the crayfish 95 Cherax quadricarinatus had a specific activity of 70 mU mg -1 min -1 in the digestive gland (homologous to the hepatopancreas), indicating that crayfish are more capable of utilizing lower trophic levels as a food source (Xue, Anderson, Richardson, Xue, Mather, 1999). What may be regarded as surprising is that primarily carnivorous species (determined by gut contents) have any celulase activity at al. The conclusion that the celulase activity is simply a byproduct of the organisms consumed may be incorrect. It was previously asumed that higher animals did not produce endogenous celulases, and that animals with this activity must acquire it from symbiotic microorganisms (Watanabe, Tokuda, 2001). However, this has been proven false for many species of crustaceans and mollusks (Byrne, Lehnert, Johnson, Moore, 1999; Suzuki, Ojima, Nishita, 2003; Xu, Janson, Selos, 2001; Xu, Helman, Erson, Janson, 2000; Xue, Anderson, Richardson, Xue, Mather, 1999). Molecular characteristics of the majority of these celulases are very similar, suggesting that they evolved from a common ancestor (Watanabe, Tokuda, 2001). Therefore, functionaly carnivorous species may posses celulase as a periphery enzyme. This can prove to be an advantage when resources are scarce (Ditel, Epifanio, Schwalm, Fantle, Fogel, 2000), so perhaps defining C. sapidus as carnivorous may be too limited a characterization of its trophic level. This is particularly true for juvenile C. sapidus, for which approximately 20% of the diet (< 80 m carapace width) consists of detritus and plant material (Laughlin, 1982; Stoner, Buchanan, 1990). The diet experiments (Chapters 4 and 5) showed that the inclusion of 20% celulose produced the same growth rates as the control diet. Therefore, even the low specific activity that we found in this study is enough to hydrolyze that percentage of celulose in the diet. 96 Celulase specific activity was higher in the foregut (4.54 + 0.31 mU mg -1 min -1 ) than in the hepatopancreas (2.50 + 0.30 mU mg -1 min -1 ). These findings are contradictory to those of McClintock et al. (1991) for C. sapidus carbohydrases, but consistent with celulase in the prawn Macrobrachium rosenbergii (Gonzalez-Pena, Anderson, Smith, Moreira, 2002). The specific activities of C. sapidus carbohydrases did not include general celulases, investigated in M. rosenbergii, which may acount for this inconsistency. Carbohydrases originate in the hepatopancreas of crustaceans, shown by mRNA in situ hybridization in Penaeus monodon (Lehnert, Johnson, 2002). In addition, the hepatopancreas has the highest specific activity of al gut tisues in six species of crabs (Johnston, Freman, 2005). Celulase is then secreted into the foregut in the form of gastric juice, which has a higher specific activity than the hepatopancreas in Cherax quadricarinatus (Xue, Anderson, Richardson, Xue, Mather, 1999). We also found higher celulase activity in the gastric juice than in the hepatopancreas of pooled samples for adult C. sapidus. The elevated specific activity in the foregut and gastric juice suggests that celulase is concentrated in the foregut to aid in digestion of raw materials. Celulase activity was significantly higher than chitinase for al tisue types investigated (Table 19, Table 20). Chitinase activities in other crustaceans are variable, and les asociated with trophic level than celulase. In the same six species of Tasmanian crabs, chitinase activity ranged from 23 ? 41 mU mg -1 min -1 in the hepatopancreas (Johnston, Freman, 2005). We did not find such high specific activity in C. sapidus hepatopancreas (0.117 + 0.018 mU mg -1 min -1 ). Considering the inhibitory afect that dietary chitin had on growth in the previous experiments (Chapters 3 and 4), it appears to have no functional importance to digestion. In the shrimp Palaemon seratus, 97 chitinase activity is very low in the gut except imediately prior to a molt (Spindler- Barth, Van Wormhoudt, Spindler, 1990). Since we ensured that the animals sacrificed for tisue samples were al intermolt, this patern sems to apply to C. sapidus. From this perspective, we can conclude that the primary purpose of chitinase in the blue crab is for molting. Chitinase specific activity was higher in the hepatopancreas (0.117 + 0.018 mU mg -1 min -1 ) than in the foregut (0.0270 + 0.0074 mU mg -1 min -1 ) at intermolt. In Palaemon seratus, chitinase activity remains constant in the midgut gland, and high relative to the integument, except during a molt (Spindler-Barth, Van Wormhoudt, Spindler, 1990). For both of these species, it appears that chitinase is more important for molting than for daily digestion. The only inconsistency is for the gastric juice of adult C. sapidus, which had a higher chitinase activity (0.104 mU mg -1 min -1 ) than the hepatopancreas (0.0638 mU mg -1 min -1 ). The adult crabs were not molt-staged, however, and it is possible that a couple of the adult males sampled were premolt. We suggest that chitinase is stored in the hepatopancreas until molting is iminent, at which point it is released into the foregut in concentrated amounts as gastric juice to aid in the degradation of the old cuticle at the anterior and posterior ends of the gut. We chose to focus on the chitinase enzyme that uses an endo-mechanism to hydrolyze polysacharides because many of the mollusks and crustaceans previously investigated had incomplete enzyme systems consisting solely of endoglucanases and endochitinases (Byrne, Lehnert, Johnson, Moore, 1999; Kono, Matsui, Shimizu, Koga, 1990; Suzuki, Ojima, Nishita, 2003; Xu, Helman, Erson, Janson, 2000). The limited hydrolytic capacity of these invertebrates may have occurred because they do not need 98 the complete system, since symbiotic microflora in the gut synthesize the mising enzymes, cleaving smal oligomers into individual sugar units (Tomme, Waren, Gilkes, 1995). Future work should be performed to determine whether blue crabs posses the enzymes to cleave N-acetylchitobiose. There was no diference in specific activities of either enzyme in relation to diet (Table 19, Table 20). In the prawn Macrobrachium rosenbergii and the mud crab Scyla serata, celulase activity was higher with increased proportions of dietary celulose (Gonzalez-Pena, Anderson, Smith, Moreira, 2002; Pavasovic, Richardson, Anderson, Mann, Mather, 2004). That there was no up- or down-regulation of enzymes suggests a lack of phenotypic plasticity in blue crabs. However, the rapid increase and decrease in digestive enzymes sen in these two species could reflect fluctuations in populations of endosymbionts in the gut flora. At least part of the celulase activity in M. rosenbergii was atributed to microorganisms in the gut, proven when the activity dropped upon treatment with antibiotics (Gonzalez-Pena, Anderson, Smith, Moreira, 2002). Whether celulase was endogenous or exogenous was not investigated for S. serata (Pavasovic, Richardson, Anderson, Mann, Mather, 2004). That neither celulase nor chitinase activity changed in relation to diet provides evidence that these enzymes are of endogenous origin. However, since we performed our asays on whole gut and tisue extracts of animals that were not treated with antibiotics, we cannot conclusively determine whether the celulase and chitinase activity is endogenous or exogenous. This isue is not of primary concern to our study, since we were more interested in the functional outcome of the enzymes. The determination of origin of the enzymes can be performed in future 99 studies through purification, cloning and sequencing procedures (Watanabe, Kono, Aida, Nagasawa, 1998; Xue, Anderson, Richardson, Xue, Mather, 1999) Compared to juveniles, both celulase and chitinase specific activities are lower in adult C. sapidus. Since the adult samples were pooled and there are no replicate samples, this data is tentative and any conclusions should be approached with caution. It is possible, however, that there are ontogenetic changes in enzyme activity, which should be further investigated. Since the percentage of detrital and plant material declines with increasing carapace width in wild blue crabs (Laughlin, 1982; Stoner, Buchanan, 1990), it is possible that the suite of carbohydrase activities may also decline. 100 Chapter 6: Discusion The experimental system Rearing juvenile blue crabs in a recirculating system was succesfully undertaken for multiple broods over the course of several months. This can be dificult to achieve with blue crabs, and high mortality and disease is common (Biddle, Milikin, Fair, Fortner, 1978; Guerin, Stickle, 1997; Winget, Epifanio, Runnels, Austin, 1976). We were able to maintain good water quality in our system throughout the feding experiments by using of mechanical and biofiltration, in addition to partial water changes. The 15 x 15 x 10 cm compartment size and mesh compartment floors alowed for enough water exchange with the tank that the disolved oxygen in each compartment remained high, while the mesh was smal enough that food particles were retained in the cages. There was some interaction efect that restricted growth and reduced conversion eficiency in the smal (10 cm 3 ) compartments when fed higher rations, so these compartments were not used after the initial experiment. The extremely low mortality in our studies and absence of any indication of disease demonstrated that this is an excelent system for future long-term feding experiments. In the future, a larger range of cage sizes should be studied to determine the point where compartment size no longer limits growth. Other parameters of the system, such as water depth, production costs, and optimal yield should be investigated in the future to optimize the system. Experimental diets Feding commercial peletized shrimp diets to juvenile C. sapidus proved a realistic option for the rearing of crabs. The combination of high survival and good growth and molting in the feding experiments showed the Ziegler (Ziegler Bros, Inc., Gardners, PA, 101 USA) formulated diet to be adequate for the animals? needs. The Melick (Melick Aquafed, Catawisa, PA) formulated diets, once coated with water-stabilizing additives, were also adequate for maintaining juvenile crabs. It is easier to achieve a balanced diet including al esential amino acids, lipids, sterols, vitamins and minerals in a multi- ingredient diet containing diferent protein sources, such as the commercial diets we used (New, 1976). The practicality of having an easily produced, inexpensive, and storable diet proved prevailingly important for a long-term feding experiment. Water stability is desirable not only because of the crabs? feding behavior of gradualy consuming it over several hours, but it also helped to preserve water quality in the recirculating system (New, 1976). Although the Melick diets were not initialy water- stable, providing an oil coating improved the stability to a certain extent, and compensated for its lack of water-stability by the added nutritional content. The oil coating improved palatability so that the animals consumed the diet more quickly, and it provided an extra energy source to compensate for nutrients lost due to leaching. Spraying the Melick diets with a combination of alginate and calcium chloride was a more efective method of improving the water stability. The combination of water stability and palatability of the diets proved efective in providing for the animals? needs, and resulted in high survival and growth rates. The average proportion of the diet consumed was the highest for the alginate-coated diets (0.94 + 0.01), showing the highest palatability within four hours of feding. In the future, various aspects of the diet composition should be investigated so that the diet is not providing more protein content than is needed, as this is wasteful and fish meal is expensive (Naylor, Goldburg, Primavera, Kautsky, Beveridge, Clay, Folke, 102 Lubchenco, Mooney, Troel, 2000). In shrimp, it is posible to decrease the protein content if the caloric content is increased (New, 1976). This should be investigated for blue crabs in the future. However, using a commercial shrimp diet from the beginning gives us a jump-start for achieving the goal of economic feasibility. Rations Depending on the priorities of the grow-out facility, optimal rations can be determined in various ways. From an energetic viewpoint, the ration resulting in the highest conversion eficiency is the best ration to deliver. In Chapter 2, the food conversion eficiency (dry weight gained (dry weight fed) -1 ) increased rapidly as ration increased up to the 50% ration, and then declined (Table 21). This conversion eficiency continued to decline from the low to the high rations in Experiment 1, showing a continuum of conversion eficiency betwen the two experiments. We can therefore conclude that the optimal ration to use (C opt ) is the 50% ration in Experiment 2. This results in the most economical use of diets, when diet is the highest cost involved in grow-out. However, we have determined diet to be a fairly inexpensive commodity, while electrical costs including water heating, aeration and water pumps for running the recirculating systems are relatively high. To reduce these costs, the priority is to grow animals as quickly as posible to maturity, and to perform experiments with equal sped. Therefore, the 50% ration is not the ideal ration for our purposes. Instead, we should use the intermediate or high ration from Experiment 1 for future diet studies, as they resulted in the highest growth rates. Even though the diferences in growth were not as obvious for the rations delivered in Experiment 1, they could stil be statisticaly distinguished from one another, so any treatment efects should be apparent in future dietary studies. 103 The intermediate ration we used for the diet experiments, based on recommended rations for penaeid shrimp, was sufficient for the crabs? metabolic needs. Rations for the crayfish Cherax quadricarinatus are around 4% of the animals? body weight day -1 , starting at 10g body weight (Manor, Segev, Leibovitz, Aflalo, Sagi, 2002). The shrimp rations we chose for our diet studies had a range of 2.8-3.9% body weight day -1 at 10g body weight, and then reached an asymptote at about 2% body weight day -1 (Guilaume, Kaushik, Bergot, Metailer, 1999), which is approximately the rate of ingestion for the juvenile shrimp Penaeus setiferus (Sick, White, Baptist, 1973). This appears to be a more fine-tuned method of ration calculation, since the rations continuously change as the animal increases in size. It alows the aquaculturist to be more eficient, reducing costly and wasteful overfeding. We did not find significant diferences in ingestion rates when rations increased because the rations we delivered were low enough that the animals were food-limited for al of them. Feding specific rations helps prevent confounding of treatment efects in nutritional experiments (Tenore, Hanson, McClain, Macubbin, Hodson, 1984). Table 21. Growth parameters for juvenile C. sapidus in large cages of experimental system, + standard eror. From Chapter 2. 104 Measures of Growth Growth parameters were analogous to literature values for blue crabs, showing the experimental system, diets, and rations to be aceptable for juvenile C. sapidus. Molt frequency was comparable to other studies with similar-sized crabs, both in actual values (range: 12.4 - 50 days; Table 22) and in the trend of frequency decreasing with each succesive molt (Cadman, Weinstein, 1988; Guerin, Stickle, 1997; Lefler, 1972; Milikin, Biddle, Siewicki, Fortner, Fair, 1980) (Table 7, Table 8). Percent increase in carapace width is comparable to published values at 20-26 o C, which fal betwen 11.8- 26% (Table 22). Values were also comparable to published data for similar-sized juvenile C. similis, a close relative to C. sapidus (Table 22). The comparison to previous laboratory studies shows that our crabs exhibited normal growth and molting paterns. Among the feding trials we performed, the average intermolt period ranging from 18-32 days was related to the length of the study, due to shorter intermolt periods in smaler crabs. But when comparing our experiments to other experiments of the same duration, the intermolt periods were similar. The molt increments (x 100%) for our feding trials, ranging from 18-26%, were higher than most of the other studies performed (Table 22). 105 Table 22. Intermolt periods and percent increase in carapace width (CW) for juvenile C. sapidus and C. similis: Comparative data from previous and curent studies. Growth rates are substantialy lower in laboratory experiments than they are in the field. Crabs in field enclosures at Point Lookout, Maryland starting at approximately 4.0 g and 35 m carapace width (determined using our alometric relationship betwen wet weight and carapace width) molted twice in 35 days (Brylawski, Miler, 2003). This is almost double the molt frequency than we found for animals of the same size in our experiments. There are various reasons why animals grow faster in the field. Among them is temperature, which reaches several degres (celcius) higher in the Chesapeake Bay in summer than in our laboratory experiments (Cowan, Boynton, 1996). Preventing activity in the compartmentalized system also decreases metabolic rates, which may depres growth rates in the laboratory compared to the field (Brylawski, Miler, 2003). A more widely varied diet in the field may supply additional esential nutrients to C. sapidus that are largely unknown, which limits our ability to properly formulate diets for 106 laboratory experiments. These various obstacles may be dificult to overcome in rearing studies, and should be further investigated to improve growth in the laboratory. Dietary Celulose In both dietary fiber experiments, there was no diference in growth betwen juvenile C. sapidus fed the control diet and the diets containing 14% or 20% celulose. The acid- detergent fiber content in the oil- and alginate-coated control diet is 7.72% and 8.20% respectively, compared to a range of 13.39-19.90% in the experimental diets (Table 11, Table 15). If the celulose were indigestible, the inclusion of twice as much fiber in the experimental diets would depres growth. In addition, the portion of the diet that we replaced had a high starch component, which blue crabs have the capacity to digest (McClintock, Klinger, Marion, Hsueh, 1991), so we replaced a portion of the diet containing energetic value with a potentialy indigestible substance. This, combined with the presence of celulase in the hepatopancreas samples, shows that blue crabs can digest celulose. In shrimp, starch and carbohydrates are hydrolyzed to form glucose for use in the Krebs cycle, glycogen storage, chitin synthesis, and the formation of sterols and faty acids (New, 1976). It is possible that juvenile blue crabs utilize complex carbohydrates for these purposes as wel, and is a potential subject for further study. In the wild, juvenile crabs les than 80 m carapace width consume detritus and plant material that constitutes approximately 20% of their diet (Laughlin, 1982; Stoner, Buchanan, 1990). In light of these experiments, the consumption of plant material containing celulose is an energeticaly important part of a juvenile blue crab?s diet in the wild. In hepatopancreas samples from six species of Tasmanian crabs ranging from carnivorous to herbivorous, celulase activity ranged from 1.4 ? 19 mU mg -1 min -1 107 respectively (Johnston, Freman, 2005). Celulase activity in juvenile C. sapidus hepatopancreas (2.50 + 0.30 mU mg -1 min -1 ) fals within this range (Table 18), slightly higher than the celulase activity of the most carnivorous crab, Carcinus maenas, for which algae comprises approximately 10% of its diet. C. sapidus celulase activity is approximately equal to celulase activity in the hepatopancreas of the juvenile spiny lobster Jasus edwarsi (2.1 + 0.6 mU mg -1 min -1 ), which also has a minor algae constituent in its diet (Johnston, 2003). From the wide range of food items consumed by juvenile crabs and the broad descriptions of their dietary niche (Laughlin, 1982; Stoner, Buchanan, 1990), we can safely describe them as omnivorous. Although the majority of their diet consists of animal material, they are also able to derive energetic value from celulose and starch in plant and detrital material, similar to other omnivorous crustaceans. Although the specific activity that we found in this study is low compared to that of the omnivorous scavenger Cherax quadricarinatus (70 mU mg -1 min -1 ) (Xue, Anderson, Richardson, Xue, Mather, 1999), it is at least enough to provide the same level of growth in a diet containing 20% celulose as a control diet. Dietary Chitin Suppresion of growth by inclusion of dietary chitin and low chitinase activity suggest that chitin has litle dietary contribution in C. sapidus. In both dietary fiber experiments, crabs fed 20% chitin had lower mean wet weights, specific growth rates, increments, and frequencies than crabs fed 20% celulose or a combination of chitin and celulose (Chapter 3, 4). While there is evidence that utilization of celulose provides an ecological advantage for juvenile C. sapidus, it may not be necesary for them to utilize dietary chitin, if chitin can be synthesized from glucose monomers like it is in shrimp 108 (New, 1976). In six species of Tasmanian crabs, chitinase activity ranged from 23 ? 41 mU mg -1 min -1 in the hepatopancreas (Johnston, Freman, 2005). We did not find such high specific activity in C. sapidus hepatopancreas (0.117 + 0.018 mU mg -1 min -1 ) (Table 18). Considering the inhibitory afect that dietary chitin had on growth in the previous experiments (Chapter 3, 4) and the extremely low chitinase activity, it appears to have no functional importance to digestion. However, al crustaceans produce some chitinase for reabsorption of chitin from the exoskeleton prior to molting (Warner, 1977). In the shrimp Palaemon seratus, chitinase activity in the integument of the gut is very low except imediately prior to a molt (Spindler-Barth, Van Wormhoudt, Spindler, 1990). However, the chitinase activity remains at a constant, higher level in the shrimp?s hepatopancreas regardles of molt. In our samples, chitinase specific activity was higher in the hepatopancreas (0.117 + 0.018 mU mg -1 min -1 ) than in the foregut (0.0270 + 0.0074 mU mg -1 min -1 ) at intermolt (Table 18). For both of these species, it appears that chitinase is more important for molting than for daily digestion. Endogenous vs. Exogenous Origins There was no diference in specific activities of either enzyme in relation to diet (Table 19, 20). In the prawn Macrobrachium rosenbergii and the mud crab Scyla serata, celulase activity was higher with increased proportions of dietary celulose (Gonzalez-Pena, Anderson, Smith, Moreira, 2002; Pavasovic, Richardson, Anderson, Mann, Mather, 2004). That there was no up- or down-regulation of enzymes suggests a lack of phenotypic plasticity in blue crabs. However, the rapid increase and decrease in digestive enzymes sen in these two species could reflect fluctuations in populations of endosymbionts in the gut flora. At least part of the celulase activity in M. rosenbergii 109 was atributed to microorganisms in the gut, proven when the activity dropped upon treatment with antibiotics (Gonzalez-Pena, Anderson, Smith, Moreira, 2002). Whether celulase was endogenous or exogenous was not investigated for S. serata (Pavasovic, Richardson, Anderson, Mann, Mather, 2004). That neither celulase nor chitinase activity changed in relation to diet provides evidence that these enzymes are of endogenous origin. However, since we performed our asays on whole gut and tisue extracts of animals that were not treated with antibiotics, we cannot conclusively determine whether the celulase and chitinase activity is endogenous or exogenous. This isue is not of primary concern to our study, since we were more interested in the functional outcome of the enzymes. The determination of origin of the enzymes can be performed in future studies through purification, cloning and sequencing procedures (Watanabe, Kono, Aida, Nagasawa, 1998; Xue, Anderson, Richardson, Xue, Mather, 1999) Other Discoveries In searching for the presence of chitinase and celulase in the diet and whether the presence of these enzymes afects growth, we had some additional unlooked-for discoveries that could become topics of further research. Previously, evidence of multiple matings was discovered by the presence of two ejaculates in 12.4% of wild females. Physiologicaly, both ejaculates have aces to the eggs, since the seminal fluid disappears over time, alowing spermatophores to mix in the spermathecae (Jivoff, 1997). However, the following fertilization of the eggs, sponge production, hatching, and survival of offspring have never ben monitored to determine if the two fathers both played a role in brood production. The initial experiment proved that two males could both have offspring from a dual-inseminated female, and in a single brood. The half- 110 siblings had significantly diferent conversion eficiencies, indicating that genetics plays a role in growth (Table 9). In addition to this finding, we provide the first evidence of compensatory growth in C. sapidus. In the first dietary fiber experiment, after switching diets, crabs in the 20% celulose treatment molted more and had a higher specific growth rate than either the 20% chitin or control group (Figure 16). Compensatory growth is common in fish (Ali, Nicieza, Wootton, 2003) and has also been identified in the Chinese shrimp Fenneropenaeus chinensis (Wu, Dong, 2002; Wu, Dong, Wang, Tian, Ma, 2001). That juvenile C. sapidus appears to ?make up? for loss of growth from previous nutrient limitation is an interesting avenue for future research, and may have management implications. 111 References Abbe, G.R., 2002. Decline in Size of Male Blue Crabs (Callinectes sapidus) from 1968- 2000 near Calvert Clifs, Maryland. Estuaries 25, 105-114. Ali, M., Nicieza, A., Wootton, R.J., 2003. Compensatory growth in fishes: a response to growth depresion. Fish and Fisheries 4, 147-190. Biddle, G.N., Milikin, M.R., Fair, P.H., Fortner, A.R., 1978. The efects of dietary fiber on survival, growth, and feding eficiency of juvenile blue crabs (Callinectes sapidus). In: Nickelson, R. (Ed.), Procedings of the 3rd Annual Tropical and Subtropical Fisheries Technological Conference of the Americas. Texas A. & M. University, College Station, Texas, pp. 285-296. Brylawski, B.J., Miler, T.J., 2003. Bioenergetic modeling of the blue crab (Callinectes sapidus) using the fish bioenergetics (3.0) computer program. Bulletin of Marine Science 72, 491-504. Byrne, K.A., Lehnert, S.A., Johnson, S.E., Moore, S.S., 1999. Isolation of a cDNA encoding a putative celulase in the red claw crayfish Cherax quadricarinatus. Gene 239, 317-324. Cadman, L.R., Weinstein, M.P., 1985. Size-weight relationships of postecdysial juvenile blue crabs (Callinectes sapidus Rathbun) from the lower Chesapeake Bay. Journal of Crustacean Biology 5, 306-310. Cadman, L.R., Weinstein, M.P., 1988. Efects of temperature and salinity on the growth of laboratory-reared juvenile blue crabs Callinectes sapidus Rathbun. Journal of Experimental Marine Biology and Ecology 121, 193-207. Cameron, J.N., 1985. Post-moult calcification in the blue crab, National Symposium of the soft-sheled blue crab fishery, pp. 31-35. Cameron, J.N., Wood, C.M., 1985. Apparent H + excretion and CO 2 dynamics acompanying carapace mineralization in the blue crab (Callinectes sapidus) following moulting. Journal of Experimental Biology 114, 181-196. Clark, M.E., Wolcott, T.G., Wolcott, D.L., Hines, A.H., 1999. Foraging and agonistic activity co-occur in fre-ranging blue crabs (Callinectes sapidus): observation of 112 animals by ultrasonic telemetry. Journal of Experimental Marine Biology and Ecology 233, 143-160. Colvocoreses, J.A., Lynch, M.P., Webb, K.L., 1974. Variations in serum constituents of the blue crab, Callinectes sapidus: major cations. Comparative biochemistry and physiology 49A, 787-803. Commision, C.B., 2005. Blue Crabs 2004: Status of the Chesapeake population and its fisheries, Annapolis, MD. Cowan, J.L.W., Boynton, W.R., 1996. Sediment-water oxygen and nutrient exchanges along the longitudinal axis of Chesapeake Bay: seasonal paterns, controlling factors and ecological significance. Estuaries 19, 562-580. Das, T., Stickle, W.B., 1994. Detection and avoidance of hypoxic water by juvenile Callinectes sapidus and C. similis. Marine Biology 120, 593-600. Davis, J.L.D., Young-Wiliams, A.C., Hines, A.H., Zohar, Y., 2005. Asesing the potential for stock enhancement in the case of the Chesapeake Bay blue crab (Callinectes sapidus). Canadian Journal of Fisheries and Aquatic Sciences 62, 109-122. Deuchi, K., Kanauchi, O., Imasato, Y., Kobayashi, E., 1994. Decreasing efect of chitosan on the apparent fat digestibility by rats fed on a high-fat diet. Bioscience, Biotechnology and Biochemistry 58, 1613-1616. Ditel, A.I., Epifanio, C.E., Schwalm, S.M., Fantle, M.S., Fogel, M.L., 2000. Carbon and nitrogen sources for juvenile blue crabs Callinectes sapidus in coastal wetlands. Marine Ecology Progres Series 194, 103-112. Esaiasen, M., Myrnes, B., Olsen, R.L., 1996. Isolation and substrate specificities of five chitinases from the hepatopancreas of northern shrimp, Pandalus borealis. Comparative biochemistry and physiology 113B, 717-723. Francis, G., Makkar, H.P.S., Becker, K., 2001. Antinutritional factors present in plant- derived alternate fish fed ingredients and their efects in fish. Aquaculture 199, 197-227. 113 Gonzalez-Pena, M.d.C., Anderson, A.J., Smith, D.M., Moreira, G.S., 2002. Efect of dietary celulose on digestion in the prawn Macrobrachium rosenbergii. Aquaculture 211, 291-303. Guerin, J.L., Stickle, W.B., 1997. A comparative study of two sympatric species within the genus Callinectes: osmoregulation, long-term aclimation to salinity and the efects of salinity on growth and moulting. Journal of Experimental Marine Biology and Ecology 218, 165-186. Guilaume, J., Kaushik, S., Bergot, P., Metailer, R., 1999. Nutrition and feding of fish and crustaceans. Praxis Publishing, Chichester, U.K. Heck, K.L.J., Thoman, T.A., 1984. The nursery role of seagras meadows in the upper and lower Chesapeake Bay. Estuaries 7, 70-92. Henrisat, B., 1991. A clasification of glycosyl hydrolases based on amino acid sequence similarities. Biochemistry Journal 280, 309-316. Hil, J., Masters, B., Place, A.R., In Review. Multiple paternity in the blue crab (Callinectes sapidus) asesed with microsatelite markers. Aquaculture Research. Jivoff, P., 1997. The relative roles of predation and sperm competition on the duration of the post-copulatory asociation betwen the sexes in the blue crab, Callinectes sapidus. Behavioral Ecology and Sociobiology 40, 175-185. Johnston, D., 2003. Ontogenic changes in digestive enzyme activity of the spiny lobster, Jasus edwardsi (Decapoda; Palinuridae). Marine Biology 143, 1071-1082. Johnston, D., Freman, J., 2005. Dietary preference and digestive enzyme activities as indicators of trophic resource utilization by six species of crab. Biological Bulletin 208, 36-46. Johnston, D.J., Yelowles, D., 1998. Relationship betwen dietary preferences and digestive enzyme complement of the slipper lobster Thenus orientalis (Decapoda: Scyllaridae). Journal of Crustacean Biology 18, 656-665. Ju, S.-J., Secor, D.H., Harvey, R., 1999. Use of extractable lipofuscin for age determination of blue crab Callinectes sapidus. Marine Ecology Progres Series 185, 171-179. 114 Kono, M., Matsui, T., Shimizu, C., Koga, D., 1990. Purifications and some properties of chitinase from the liver of a prawn, Penaeus japonicus. Agricultural and Biological Chemistry 54, 2145-2147. Laughlin, R.A., 1982. Feding habits of the blue crab, Callinectes sapidus Rathbun, in the Apalachicola Estuary, Florida. Bulletin of Marine Science 32, 807-822. Le, K.J., Watson, R.D., 2002. Antipeptide antibodies for detecting crab (Callinectes sapidus) molt-inhibiting hormone. Peptides 23, 853-862. Le, K.J., Watson, R.D., Roer, R.D., 1998. Molt-inhibiting hormone mRNA levels and ecdysteroid titer during a molt cycle of the blue crab, Callinectes sapidus. Biochem Biophys Res Comm 249, 624-627. Le, P.G., Smith, L.L., Lawrence, A.L., 1984. Digestive proteases of Penaeus vannamei Boone: relationship betwen enzyme activity, size and diet. Aquaculture 42, 225- 239. Le, R.F., Noone, T., 1995. Efect of reproductive toxicants on lipovitelin in female blue crabs, Callinectes sapidus. Marine Environmental Research 39, 151-154. Lefler, C.W., 1972. Some efects of temperature on the growth and metabolic rate of juvenile blue crabs, Callinectes sapidus, in the laboratory. Marine Biology 14, 104-110. Lehnert, S.A., Johnson, S.E., 2002. Expresion of hemocyanin and digestive enzyme mesenger RNAs in the hepatopancreas of the black tiger shrimp Penaeus monodon. Comparative Biochemistry and Physiology Part B. 133, 163-171. Lipcius, R.N., Stockhausen, W.T., 2002. Concurrent decline of the spawning stock, recruitment, larval abundance, and size of the blue crab Callinectes sapidus in Chesapeake Bay. Marine Ecology Progres Series 226, 45-61. Lynn, K.R., 1990. Chitinases and chitobiases from the American lobster (Homarus americanus). Comparative biochemistry and physiology 96B, 761-766. Manor, R., Segev, R., Leibovitz, M.P., Aflalo, E.D., Sagi, A., 2002. Intensification of redclaw crayfish Cherax quadricarinatus culture I. Growout in a separate cel system. Aquacultural Engineering 26, 263-276. 115 McCarthy, J.F., Skinner, D.M., 1979. Changes in the ecdysteroids during embryogenesis of the blue crab, Callinectes sapidus Rathbun. Developmental Biology 69, 627- 633. McClintock, J.B., Klinger, T.S., Marion, K., Hsueh, P., 1991. Digestive carbohydrases of the blue crab, Callinectes sapidus (Rathbun): implications in utilization of plant- derived detritus as a trophic resource. Journal of Experimental Marine Biology and Ecology 148, 233-239. McDonald, P., Edwards, R.A., Grenhalgh, J.F.D., Morgan, C.A., 2002. Animal Nutrition. Pearson Education Ltd., Harlow, England. Mesick, G.A., Sindermann, C.J., 1992. Synopsis of principal diseases of the blue crab, Callinectes sapidus, NOA Technical Memorandum. NMFS-F/NEC-88, pp. 24. Meyers, S.P., Butler, D.P., Hastings, W.H., 1972. Alginates as binders for crustacean rations. The Progresive Fish Culturist 34, 9-12. Milikin, M.R., Biddle, G.N., Siewicki, T.C., Fortner, A.R., Fair, P.H., 1980. Efects of various levels of dietary protein on survival, moulting frequency and growth of juvenile blue crabs (Callinectes sapidus). Aquaculture 19, 149-161. Moksnes, P.O., Lipcius, R.N., Pihl, L., Van Montfrans, J., 1997. Cannibal-prey dynamics in young juveniles and postlarvae of the blue crab. Journal of Experimental Marine Biology and Ecology 215, 157-187. Molano, J., Duran, A., Cabib, E., 1977. A rapid and sensitive asay for chitinase using tritiated chitin. Analytical Biochemistry 83, 648-656. Naylor, R.L., Goldburg, R.J., Primavera, J.H., Kautsky, N., Beveridge, M.C.M., Clay, J., Folke, C., Lubchenco, J., Mooney, H., Troel, M., 2000. Efect of aquaculture on world fish supplies. Nature 405, 1017-1024. New, M.B., 1976. A review of dietary studies with shrimp and prawns. Aquaculture 9, 101-144. Panel, F.E.P.T.A., 2004. Fisheries Ecosystem Planning for Chesapeake Bay, NOA, Chesapeake Bay Ofice Prepublication, Annapolis, MD, pp. 374. 116 Pavasovic, M., Richardson, N.A., Anderson, A.J., Mann, D., Mather, P.B., 2004. Efect of pH, temperature and diet on digestive enzyme profiles in the mud crab, Scyla serata. Aquaculture 242, 641-654. Perkins-Viser, E., Wolcott, T.G., Wolcott, D.L., 1996. Nursery role of seagras beds: enhanced growth of juvenile blue crabs (Callinectes sapidus Rathbun). Journal of Experimental Marine Biology and Ecology 198, 155-173. Rugolo, L.J., Knotts, K.S., Lange, A.M., 1998. Historical profile of the Chesapeake Bay blue crab (Callinectes sapidus Rathbun) fishery. Journal of Shelfish Research 17, 383-394. Sagasti, A., Schafner, L.C., Duffy, J.E., 2001. Efects of periodic hypoxia on mortality, feding and predation in and estuarine epifaunal comunity. Journal of Experimental Marine Biology and Ecology 258, 257-283. Sandoz, M., Rogers, R., 1944. The efect of environmental factors on hatching, moulting, and survival of zoea larvae of the blue crab Callinectes sapidus Rathbun. Ecology 25, 216-228. Shiau, S.-Y., Yu, Y.-P., 1999. Dietary supplementation of chitin and chitosan depreses growth in tilapia, Oreochromis niloticus X O. aureus. Aquaculture 179, 439-446. Sick, L.V., White, D., Baptist, G., 1973. The efect of duration of feding, amount of food, light intensity, and animal size on rate of ingestion of peleted food by juvenile penaeid shrimp. The Progresive Fish Culturist 35, 22-26. Spindler-Barth, M., Van Wormhoudt, A., Spindler, K.D., 1990. Chitinolytic enzymes in the integument and midgut-gland of the shrimp Palaemon seratus during the moulting cycle. Marine Biology 106, 49-52. Steven, C.R., Hil, J., Masters, B., Place, A.R., 2005. Genetic markers in blue crabs (Callinectes sapidus) I: Isolation and characterization of microsatelite markers. Journal of Experimental Marine Biology and Ecology 319, 3-14. Stoner, A.W., Buchanan, B.A., 1990. Ontogeny and overlap in the diets of four tropical Callinectes species. Bulletin of Marine Science 46, 3-12. 117 Suzuki, K.-i., Ojima, T., Nishita, K., 2003. Purification and cDNA cloning of a celulase from abalone Haliotis discus hannai. European Journal of Biochemistry 270, 771- 778. Tenore, K.R., Hanson, R.B., McClain, J., Macubbin, A.E., Hodson, R.E., 1984. Changes in composition and nutritional value to a benthic deposit feder of decomposing detritus pools. Bulletin of Marine Science 35, 291-311. Tomme, P., Waren, R.A.J., Gilkes, N.R., 1995. Celulose hydrolysis by bacteria and fungi. Advances in microbial physiology 37, 1-67. Van Olst, J.C., Carlberg, J.M., 1978. The efects of container size and transparency on growth and survival of lobsters cultured individualy. In: Avault, J.W., Jr. (Ed.), The ninth annual meting of the World Mariculture Society. World Mariculture Society, Atlanta, GA, pp. 469-479. Vigh, D.A., Dendinger, J.E., 1982. Temporal relationships of postmolt deposition of calcium, magnesium, chitin and protein in the cuticle of the Atlantic blue crab, Callinectes sapidus Rathbun. Comparative Biochemistry and Physiology Part A. 72, 365-369. Wang, S.Y., Stickle, W.B., 1986. Changes in nucleic acid concentration with starvation in the blue crab Callinectes sapidus Rathbun. Journal of Crustacean Biology 6, 49-56. Warner, G.F., 1977. The Biology of Crabs. Van Nostrand Reinhold Company, New York, 197 pp. Watanabe, H., Tokuda, G., 2001. Review: Animal Celulases. Celular and Molecular Life Sciences 58, 1167-1178. Watanabe, T., Kono, M., Aida, K., Nagasawa, H., 1998. Purification and molecular cloning of a chitinase expresed in the hepatopancreas of the penaid prawn Penaeus japonicus. Biochimica et Biophysica Acta 1382, 181-185. Watson, R.D., Le, K.J., Borders, K.J., Dircksen, H., Lily, K.Y., 2001. Molt-inhibiting hormone imunoreactive neurons in the eyestalk neuroendocrine system of the blue crab, Callinectes sapidus. Arthropod Structure & Development 30, 69-76. 118 Winget, R.., Maurer, D., Anderson, L., 1973. The feasibility of closed system mariculture: preliminary experiments with crab molting, Procedings of the National Shelfisheries Asociation, pp. 88-92. Winget, R.., Epifanio, C.E., Runnels, T., Austin, P., 1976. Efects of diet and temperature on growth and mortality of the blue crab, Callinectes sapidus, maintained in a recirculating culture system, Procedings of the National Shelfisheries Asociation, pp. 29-33. Wootton, R.J., 1998. Ecology of Teleost Fishes. Kluwer Academic Publishers, Boston, 386 pp. Wu, L., Dong, S., 2002. The efects of repetitive 'starvation-and-refeding' cycles on the compensatory growth response in Chinese shrimp, Fenneropenaeus chinensis (Osbeck, 1765) (Decapoda, Penaeidae). Crustaceana 74, 1225-1239. Wu, L., Dong, S., Wang, F., Tian, X., Ma, S., 2001. The efect of previous feding regimes on the compensatory growth response in Chinese shrimp, Fenneropenaeus chinensis. Journal of Crustacean Biology 21, 559-565. Xu, B., Janson, J.-C., Selos, D., 2001. Cloning and sequencing of a molluscan endo-B- 1,4-glucanase gene from the blue mussel, Mytilus edulis. European Journal of Biochemistry 268, 3718-3727. Xu, B., Helman, U., Erson, B., Janson, J.-C., 2000. Purification, characterization and amino-acid sequence analysis of a thermostable, low molecular mas endo-B-1,4- glucanase from blue mussel Mytilus edulis. European Journal of Biochemistry 267, 4970-4977. Xue, X.M., Anderson, A.J., Richardson, N.A., Xue, G.P., Mather, P.B., 1999. Characterization of celulase activity in the digestive system of the redclaw crayfish (Cherax quadricarinatus). Aquaculture 180, 373-386. Zmora, O., Findiesen, A., Stubblefield, J., Frenkel, V., Zohar, Y., 2005. Large-scale juvenile production of the blue crab Callinectes sapidus. Aquaculture 244, 129- 139.