WOLFF: BEAK KKY FOR EIGHT CEPHALOPOD SPECIES 



Table 5.— Regression equations and r 2 values for mantle length and body weight, 

 lower beak regression equations in centimeters, asterisk indicates best regression 

 based on r 2 . 



damaged margins (e.g., RW, WW upper beak) 

 were excluded from consideration when con- 

 structing the key, even though they might show 

 very good separation between species' means 

 when used in a ratio (e.g., RL/RW upper beak). 

 Larger dimensions which have easily damaged 

 margins (e.g., CL/HL) can still provide a reliable 

 dimension within the variability of the sample 

 simply because the eroded margin represents 

 less of the overall dimension compared with the 

 smaller dimension with similar properties. 



Accurately determining which cephalopods 

 are abundant in an area and which of these 

 might be important in a predator's diet are diffi- 

 cult problems to solve. The abundance of a spe- 

 cies in a trawl sample is not necessarily an accu- 

 rate reflection of its relative abundance in the 

 field (Wormuth 1976) or in a predator's stomach 

 (Clarke 1977). In an attempt to reduce this 

 sampling bias the cephalopods in this study were 

 chosen on the basis of their abundance in trawl 

 samples (Young 1972; Okutani 1974), in collec- 

 tions using alternate sampling devices (e.g., dip 

 nets and jigs (Wormuth 1976)), and in stomach 

 content studies of cephalopod predators in the 

 same area (Pinkas et al. 1971; Perrin et al. 1973). 



The eastern tropical Pacific is the area for 

 which these beak characterizations were con- 

 structed. In many cases, large, pelagic cephalo- 

 pod predators in this area will contain a large 

 percentage of the species described in this study. 

 As one moves away from this area, however, less 

 can be said about the potential usefulness of this 

 key, since the species composition and morpho- 

 logical characteristics, including beak dimen- 

 sions, can change. As an example, 28 specimens 

 of O. bartramii from the Gulf of Mexico and 

 northwestern Atlantic have an upper rostral 



length to jaw width ratio mean (RL/JW) of 1.22 

 (CI05 = ±0.02); considerably greater than the 

 eastern tropical Pacific mean of O. bartramii (x 

 = 1.06, CI05 = ±0.03). This higher ratio value also 

 holds for three specimens from southeastern 

 Australia. 



Such geographical variation in species with 

 disjunct distributions is not uncommon and has 

 been noted in other body measurements for O. 

 bartramii by Young (1972). Additional measure- 

 ments must be made on remaining cephalopod 

 species in this key, particularly those with dis- 

 junct distributions, before this key can be reliably 

 used outside the eastern tropical Pacific area. 



There will be cephalopods in the stomachs of 

 predators which are not included in this work. In 

 order to reduce misidentifications, therefore, 

 full use should be made of the alternate ratio 

 means, the beak figures, and the descriptive 

 characteristics. 



In most beaks, the dimensions which resulted 

 in the best regression equations for mantle 

 length and body weight were those that were 

 close to the overall length of the beak (CL, HL, 

 RC). In badly damaged beaks, however, these 

 dimensions are often in poor conditon. The pairs 

 of regression equations for each of the eight spe- 

 cies represent an effort to increase the flexibility 

 of estimating the size of a cephalopod. The re- 

 gression equations which use the RL dimension 

 variable will give less accurate estimates, but 

 can be used in all but the most severely damaged 

 beaks, as the RL is a very durable dimension. 



ACKNOWLEDGMENTS 



I thank J. H. Wormuth and A. D. Hart, Texas 

 A&M University, for providing many helpful 



369 



