Buoyancy of Atlantic cod larvae in relation to developmental stage and maternal influences

The lack of success in predicting the recruitment of fish may to some extent arise from insufficient knowledge of processes affecting egg and larval survival (Chambers and Trippel, 1997). In this context, the quality of eggs and larvae [cf. definitions given in Brooks et al. (Brooks et al., 1997)] and their capacity to develop properly are often mentioned as important factors along with the corresponding oceanographic spatial distribution and transport. Buoyancy, defined as the ability of these pelagic early life stages to float in sea water, is often used as a positive criterion of quality (Brooks et al., 1997), but is also as a key variable in drift models [see, for instance, Sundby (Sundby, 1991)]. For Atlantic cod (Gadus morhua L.), a highly fecund, multiple batch spawner, egg buoyancy has been extensively reported on by many research groups [see (Kjesbu et al., 1992; Nissling et al., 1994; Ouellet, 1997) and references therein]. Obvious stock differences seem to exist, exemplified most clearly by the extreme high water content of the Baltic Sea cod eggs resulting in low-density eggs (Thorsen et al., 1996). In contrast to this wealth of information, little published buoyancy data exist for the larval Atlantic cod stage (Sclafani et al., 1997), especially in relation to traits of the individual, mother fish (maternal effects) or time of spawning (seasonal effects) (Solemdal, 1997). For instance, larger cod females are known to produce larger eggs and thereby larger larvae (Marteinsdottir and Begg, 2002), while there is a significant depletion of body reserves during the course of spawning and a general reduction in egg size (Kjesbu et al., 1996; Lambert and Dutil, 2000). In the work of Yin and Blaxter, buoyancy patterns from hatching to point-ofno-return of benzocaine-anaesthetized larvae of cod, Atlantic herring (Clupea harengus) and flounder (Platichthys flesus), measuring rising and sinking rates, are described pointing to large inter-specific differences (Yin and Blaxter, 1987a). One partial reason for this shortage in comprehensive fish larval buoyancy data might be that any precise recordings are complicated by the fact that the larvae need to be adequately immobilized as they are normally actively swimming around in the density(salinity-) graduated column (Coombs apparatus) typically used today for any buoyancy work. In Sclafani et al. (Sclafani et al., 1997), MS-222 was added to anaesthetize pooled samples of cod larvae while they were being kept inside this column.