Klibansky and Juanes: Effects of preservation on oocytes of tfiree Atlantic fisfi species 



539 



so they could then be counted and measured more easily 

 (Simpson, 1951). 



In this analysis, the effects of four preservatives on 

 ovarian material weight and oocyte diameter of three 

 species of commercially important Northwest Atlantic 

 groundfish are evaluated. The preservatives are forma- 

 lin (0.037 formaldehyde, 0.015 methyl alcohol, <0.01 so- 

 dium phosphate dibasic, <0.01 sodium phosphate mono- 

 basic, 0.93 deionized water), modified Gilson's solution 

 (0.10 607f ethanol, 0.015 nitric acid, 0.008 glacial acetic 

 acid, 0.88 distilled water), 70% ethanol, and freezing 

 and are hereafter referred to simply as formalin, Gil- 

 son's, ethanol, and freezing. The three species used in 

 this project included two gadids, Atlantic cod iGadus 

 morhua) and haddock (Melanogrammus aeglefinus), 

 and one pleuronectid, American plaice (Hippoglossoides 

 platessoides). These species are hereafter referred to as 

 cod, haddock, and plaice. All are historically important 

 groundfish species in the Northwest Atlantic that have 

 been reduced to low numbers over the past century 

 (Boreman et al., 1997) and are now managed together 

 in the Northeast multispecies fishery. Reproductive 

 biology and ecology are similar among these species: 

 they are all iteroparous, determinate, batch spawn- 

 ers exhibiting group-synchronous ovarian organization 

 (Murua and Saborido-Rey, 2003). Although data are 

 abundant for these species on many characteristics used 

 to predict stock reproductive potential, such as maturity 

 and sex-ratio, data on fecundity and oocyte size remain 

 very limited (Tomkiewicz et al., 2003b). 



Materials and methods 



Ovary collection 



All ovarian specimens were taken from fish caught in 

 bottom trawls in the Northwest Atlantic during peak 

 spawning (Table 1). In February 2004 the ovaries of 

 19 ripening (classified as stage IV by the maturity key 

 developed by Tomkiewicz et al. [2003a]) cod and 16 had- 

 dock were collected by bottom trawling in the western 

 portion of Georges Bank (GB). In May 2004 the ovaries 

 of 23 ripening cod and 16 ripening plaice were collected 



by bottom trawling in the inshore waters of the Gulf of 

 Maine (GOM). 



Preservation and measurement 



For all specimens collected on GB (cod and haddock), 

 fresh weight of the left ovary was measured at sea, to 

 the nearest 1.0 g, with a Mare! marine balance (Marel 

 Food Systems, Gardabaer, Iceland). These ovaries were 

 then preserved in formalin. Specimens collected in the 

 GOM (cod and plaice) could not be weighed at sea and 

 therefore were packed in ice until they could be weighed 

 to the nearest 0.001 g in the laboratory 24 hours later. 

 The left ovarian lobes of 17 cod caught in the GOM were 

 reweighed after 48 hours on ice to determine if time on 

 ice affected ovary weight. 



Within minutes of weighing, the entire left lobe of 

 most specimens was placed in a 1-L jar containing a 

 volume of formalin approximately equal to four times 

 the volume of the ovary. When the left lobe was too 

 large (>250 g) to fit through the opening of the jar. a 

 large portion weighing 250 g or less was cut off and 

 weighed and preserved as above. These whole lobes and 

 lobe portions of ovarian material remained in formalin 

 for 158-175 (mean=169) days and were then reweighed 

 to the nearest 0.001 g. These are hereafter referred to 

 as lobe-formalin samples. 



From the center of the right ovary of each fish, four 

 1.5-mL subsamples of ovarian material (i.e., one sub- 

 sample per fish for each of four treatments: formalin, 

 Gilson's, ethanol, and freezing) almost entirely compris- 

 ing vitellogenic oocytes, were removed with a 3-mL 

 plastic syringe tube, the end of which was cut off at 

 the zero mark. The collection of subsamples from fresh 

 ovaries in this way was very easy, gave no evidence of 

 damaged oocytes, and repeatedly produced subsamples 

 with a mean of 1.54 g (/i = 155, coefficient of variation 

 [CV] = 3.70). Across fish species, differences in oocyte 

 size and density within and between left and right 

 ovarian lobes are uncommon (West, 1990), and oocyte 

 size tends not to vary among different locations in cod 

 ovaries (Kjesbu and Holm, 1994). Still, in this study 

 ovarian material was always taken from the same part 

 of each ovary in order to ensure that subsamples of 



