HOHN and HAMMOND: POSTNATAL GROWTH OF SPOTTED DOLPHIN 



ferences are more likely a result of variability in the 

 data. We suggest, therefore, that the mean of the 

 estimates from both growth curves be used for 

 estimates of lengths-at-age for the northern offshore 

 spotted dolphin. 



For the southern population, the mean lengths of 

 105.0 and 127.5 cm for January and May, respec- 

 tively, suggest that there may be two calving sea- 

 sons in this area. At present, the sample size is 

 too small to assess whether or not this is the 

 casa 



ESTIMATION OF 



LENGTH-AT-AGE USING 



GROWTH-LAYER-GROUP (GLG) AGING 



Increments of tissue are deposited in teeth as a 

 function of time The most important incremental 

 pattern in odontocete teeth is comprised of growth 

 layer groups (GLGs), defined as "a repeating or semi- 

 repeating pattern of adjacent groups of incremen- 

 tal growth layers within the dentine, cementum, or 

 bone which is defined as a countable unit" (Perrin 

 and Myrick 1980, p. 48-49). These GLGs are used 

 for age determination in many species of odon- 

 tocetes, as well as pinnipeds and sirenians (see 

 review by Scheffer and Myrick 1980), but in most 

 species no calibration of GLGs with absolute time 

 is available However, a few known-age captive and 

 minimum-known-age captive bottlenose dolphins 

 (Sergeant 1959; Sergeant et al. 1973; Hui 1978) and 

 captive tetracycline-marked specimens of other 

 species {Lagenorhynchus obscurus, Best 1976; 

 Delphinus delphis, Gurevich et al. 1980; 5. longi- 

 rostris, Myrick et al. 1984) have provided evidence 

 that the GLG as defined and calibrated by these 

 workers represents an annual deposition pattern. In 

 the absence of any known-age specimens of spotted 

 dolphins, we have assumed that a GLG pattern 

 similar to that described in the above species 

 represents the same amount of time 



Methods 



A sample of 800 males and 800 females, selected 

 randomly from the specimens collected between 

 1973 and 1978, and all 312 female specimens col- 

 lected in 1981 made up the sample of animals from 

 which teeth were aged. 



The teeth were decalcified in RDO^, a commercial 

 decalcifying agent, cut longitudinally into 24 ^m thin 

 sections using a freezing microtome, stained in 



■•Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



haematoxylin, and mounted in 100% glycerin. Detail- 

 ed procedures for the preparation technique and 

 interpretation of GLGs are described by Myrick et 

 al. (1983). 



Tfeeth from each of the 1,600 specimens collected 

 between 1973 and 1978 were "read" for age at least 

 three times, to the nearest 0.1 GLG in young animals, 

 by each of two readers over a period of 2 yr. The 

 series of age estimates was averaged for each reader, 

 and the resulting two mean age estimates were again 

 averaged to produce a pooled mean age estimate (see 

 Reilly et al. 1983). Only one age reading was made 

 by each reader for the 312 specimens collected in 

 1981, and the mean of these two readings deter- 

 mined. For lack of a preference for one reader's 

 estimates, growth rate analyses used the pooled 

 mean and mean age estimates. Growth models were 

 fit to the age-length data for males and females 

 separately. 



Growth rate was estimated by regressing length 

 on number of GLGs (age) using Laird's (1969) form 

 of the Gompertz model. The data were truncated at 

 <3.0 GLGs in order to reduce the effects of older 

 animals on our estimates of growth in the first 2 yr, 

 in case the chosen model failed to describe growth 

 adequately over a wider range of ages. Length at 

 birth was fixed at the independently estimated value 

 of 85.4 cm (see above). 



Results 



There was no difference in growth between males 

 and females at this age Therefore, the data were 

 pooled. Figure 7 shows the Gompertz model of 

 growth fitted to the pooled data. The model gives 

 a predicted length of 123 cm (SD = 0.7 cm) at 1 yr 

 and a predicted length of 143 cm (SD = 0.6 cm) at 

 2 yr. These standard deviations are underestimates 

 because they do not incorporate variability in the age 

 of individual specimens resulting from between 

 reader differences. 



Figure 7 also shows that lengths predicted by this 

 model may be underestimated up to about 8 mo and 

 overestimated from about 8 to 13 mo. For com- 

 parison with predicted length at 1 yr from the model, 

 the mean length of specimens aged between 0.9 and 

 1.1 yr (n = 24) is 121 cm with a range of 101-140 cm. 



The estimated monthly growth rate is 3.84 cm/mo 

 initially, falling to 3.11 cm/mo at 6 mo, 2.33 cm/mo 

 at 12 mo, 1.67 cm/mo at 18 mo, and 1.15 cm/mo at 

 24 mo. 



Discussion 



The accuracy of these length-at-age estimates 



563 



