This trend was the opposite of that of the aged sample (This trend is also discussed in 

 Section II). 



20 13 



CLBS 



Figmc 1: Cumulative distributions of ages (id GLGS) of 

 Texas January-June 1983-89 and 1990 bottlenose dolphin 

 strandings. 



M no IM laa it* im 2i« zm 



LEKTTM Ol 



290 ]7s no 3ia 



Fignre 2 Cumulative distributions of lengths of Tens 

 Januaiy-June 1983-89 and 1990 bottlenose dolphin 

 unndings. 



The aged samples are a sub-sample of the stranded animals with length data. If the 

 aged sample is representative of the length sample, the length structure of the aged samples 

 should reflect that of the overall length samples. To examine this, the cumulative 

 distributions of the lengths of the 1983-89 and 1990 aged samples were compared to the 

 1983-89 and 1990 length samples, respectively (Figure 3-4). The results of the Kohnogorov- 

 Smimov two-sample tests of these distributions indicate that the while the 1990 aged sample 

 was not significantly different from the 1990 length sample (DN=0.105, p>0.59), the 1983-89 

 aged sample was significantly different from the 1983-89 length sample (DN=0.245, p<0.01). 



*S IIS ISO ISO IT! IM 111 130 230 270 l*a III 



IINCTM Ol 



Figure 3: Cumulative distributions all lengths and lengths 

 of aged sample ofToas 1983-89 January-June bottlenose 

 dolphin strandings. 



M no ISO ISO 



170 Ita 211 230 290 171 2M 3l« 

 UNCTM CH 



Figiiic 4: Cumulative distributions 

 lengths of aged sample of Texas 

 bottlenose dolphin strandings. 



of all lengths and 

 1990 January-June 



These results indicate that the 1983-89 aged sample was biased in some way. It 

 appears that shorter (< 140cm) animals were under-represented in the 1983-89 aged sample. 

 The teeth of most of these shorter animals probably were not collected during 1983-89 

 because the teeth had not yet erupted. 



22 



