30 



Fishery Bulletin 89(1), 1991 



would be characterized by relatively heavier calcifica- 

 tion compared with the more rapidly growing, less 

 calcified zones. In addition, the formation of the ring 

 takes at least 4 months, but during the first 10 years 

 of life the width of a ring is much less, in comparison 

 with adjacent growth bands, than it would be if the 

 growth rate were constant throughout the year. It is 

 concluded that in G. galeus the ring represents a period 

 of slow growth, and that mineralization is more intense 

 during this period. 



Vooren and Betito (In press) have shown that indi- 

 viduals of Galeorhinus galeus migrate northwards into 

 the study area starting in April, until their peak abun- 

 dance in September. At this time the school shark is 

 abundant on the shelf south of lat. 32 °S and scarce or 

 absent further north. After September, the fish mi- 

 grate southward from the study area, are scarce there 

 in November, and absent in January and February. At 

 the onset of winter, a marked change in temperature 

 preference occurs, with most fishes occurring at 

 18-20°C from April to June and at 11-15°C in August 

 and September, although water masses of higher 

 temperatures are available during the latter months. 

 The present data show that gravid females arrive first, 

 making up all the April and May catches. From June 

 onwards, other groups (adult males, non-gravid adult 

 females and juveniles of both sexes) migrate into the 

 area. The whole population experiences the decrease 

 in temperature from June to August (Vooren and 

 Betito In press). Thus, the period of slow growth and 

 ring formation in vertebral centra coincides with this 

 change in temperature preference of the population at 

 the onset of winter. Similar results have been reported 

 for other species (Stevens 1975, Jones and Geen 1977), 

 confirming the general view that ring formation is 

 associated with low temperature and slow vertebral 

 growth (Longhurst and Pauly 1987). 



As for the process involved, Casselman (1974) has 

 suggested that during slow-growth phases, the amount 

 of protein available for appositional growth might be 

 reduced although minerals would still be available. In 

 the shark vertebra, slow growth is evidently associated 

 with a reduced rate of deposition of matrix compo- 

 nents, which include both collagen fibrils and poly- 

 saccharides. Digestion and depletion of the carbo- 

 hydrate moiety of the matrix during the slow-growth 

 phase would facilitate interaction between collagen 

 fibrils and mineral ions, thus promoting calcification. 

 Mugiya (1987) has shown that in fish otoliths both 

 calcium uptake and protein synthesis vary in an en- 

 dogenous process controlled by hormones. In the school 

 shark vertebrae, a matrix less dense in protein formed 

 during slow-growth phases could be the cue to a higher 

 calcium uptake to fill the space available for mineraliza- 

 tion. Jones and Geen (1974) related wider rings with 



warmer years in spines oiSqualus acanthias, and this 

 relationship may explain the variation in the width of 

 the rings which is observed in vertebrae of adult female 

 school sharks (Fig. 15). Gravid females which migrate 

 into the study area during March to May show prefer- 

 ence for higher temperatures, and it is possible that 

 they remain in the warmer part of the species' tem- 

 perature range during the winter. If so, their rings 

 might grow faster than in fishes that remain at lower 

 temperatures. The distribution of the different com- 

 ponents of the population in the study area during the 

 winter should be investigated in detail to test this 

 hypothesis. 



Reading vertebrae is the most important tool for age 

 determination in many elasmobranchs. Length-fre- 

 quency analysis is most suitable for fast-growing spe- 

 cies because of the assumption that all fishes in the 

 sample have the same age at the same length (Long- 

 hurst and Pauly 1987). In slow-growing, long-lived 

 species, however, a given size class contains several 

 different age groups (Gruber and Stout 1983). The 

 Von Bertalanffy growth parameters estimated by 

 ELEFAN software (Pauly and David 1981), using 

 length-frequency data, overestimated the growth rate 

 determined by vertebral readings. These values were 

 approximately the same as those found by Olsen (1954) 

 when analyzing length-frequency data for the Aus- 

 tralian school shark. Later, Grant et al. (1979), using 

 tag-recapture methods, estimated lower values of 

 growth rates and concluded that length-frequency 

 analysis was impracticable in this species. 



Acknowledgments 



Special thanks are due to Dr. Daoiz Mendoza for all his 

 help and assistance with the histological aspects of this 

 work. We thank Monica B. Peres for the valuable 

 information and Mauro Maida, Dr. Garry R. Russ, 

 Marcus V.S. Ferreira, and Annadel Cabanban for their 

 constructive criticisms of the manuscript. We also 

 thank the people from Projeto Talude and RV Atlan- 

 tico Sul who helped with the field work. The research 

 was sponsored by the Brazilian Ministry of Science and 

 Technology (CNPq). 



Citations 



Applegate, S.P. 



1967 A survey of shark hard parts. In Gilbert, P.W., R.F. 

 Mathewson, and D.P. Rail (eds.). Sharks, skates and rays, p. 

 37-67. John Hopkins Press, Baltimore. 

 Beamish, R.J., and G.A. McFarlane 



1983 The forgotten requirement for age validation in fisheries 

 biology. Trans. Am. Fish. Soc. 112:735-743. 



