44 



Fishery Bulletin 98(1) 



100 E 



110 



120 



130 140 



150 



160 



10S 



20 



30 



40 



50 



o Observed on longlines 

  Pregnant females 



10S 



20 



30 



40 



50 



100 E 



110 



120 



140 



150 



160 



130 



Figure 2 



Map of the Australian region showing start positions of tuna longline sets from which por- 

 beagles were recorded, and captui-e locations of pregnant females l7!=5.). The Exclusive Eco- 

 nomic Zone is also shown. 



classes. To assist modal discrimination, we limited our 

 length-frequency analyses to the period April-July, 

 during which Kl'''( of New Zealand and 86'^^^ of Austra- 

 lian length measurements were taken. The MIX com- 

 puter program (MacDonald and Pitcher, 1979; Mac- 

 Donald, 1987; MacDonald and Green, 1988) was ap- 

 plied separately to the New Zealand and Australian 

 length-frequency data for combined sexes to decompose 

 the distributions into their component age classes. The 

 program estimates the mean length, and the standard 

 deviation of the lengths, for each age class, and the 

 proportion of the sample in each age class. Length-fire- 

 quency data were grouped into 3-cm class intei'vals, 

 and truncated at 162 cm (New Zealand) and 150 cm 

 (Australia) before analysis because large sharks were 

 poorly represented in the samples. For each data set, 

 we fitted a MDC model with three age classes and then 

 progressively added extra age classes until there was no 

 significant improvement in the )[- goodness of fit (Mac- 

 Donald and Green, 1988). Occasionally, partially con- 

 strained fits ( alternately fixing the standard deviations 

 and proportions of one or two of the older age classes ) 

 were necessary for successfiil convergence (MacDonald 

 and Green, 1988). This had no effect on the estimates 

 of mean length, which were never constrained. 



Embryos and ova 



Embryos were placed in plastic bags and ft-ozen; some- 

 times only partial litters were retained. Rarely, the 

 uteri were removed intact and frozen with the em- 

 bryos still inside. In the laboratory, embryos were 

 thawed, sexed, and their jaws were examined for func- 

 tional teeth. The embryos were then weighed, and 

 measured (usually PL, FL, TL„.,,, and TL^^,^). FL was 

 estimated from TL for three embryos without FL mea- 

 surements (see "Results" section for regression equa- 

 tions). The liver and the contents of the stomach 

 and intestine were weighed separately. Stomach con- 

 tents were expressed as a percentage of total weight. 

 Liver weight, and the weight of the intestinal contents 

 were expressed as percentages of yolk-free embryonic 

 weight to avoid distortions caused by the large varia- 

 tion in the stomach contents. 



The uteri and right ovary- of the Macquarie Island 

 female were examined. The diameters of a subsample 

 of ovarian ova were measured using an image analy- 

 sis system attached to a binocular microscope, and an 



- The left ovary of lamnid sharks is vestigial iPratt, 1988). 



