MACY: CLASSIFYING LONG-FINNED SQUID INTO SEXUAL MATURITY STACKS 



Laboratory studies confirm evidence from the 

 field that spent squid cannot be reliably distin- 

 guished from spawning squid, even though the 

 majority of samples (Narragansett Bay) were 

 taken on the spawning grounds during the peak 

 of reproductive activity. Hixon (1980) was also 

 unable to find spent L. pealei or arrow squid, L. 

 plei, in the Gulf of Mexico, and he too document- 

 ed multiple spawning by L. pealei. Furthermore, 

 it has been shown histologically (Burukovski and 

 Vovk 1974) that egg development is highly asyn- 

 chronous among individuals, and that a series of 

 eggs at different stages of development in the 

 ovary is typical. Prolonged spawning by poorly 

 synchronized individuals of a population would 

 tend to extend spawning over time, and may well 

 account for the lack of reports of dead or 

 dying squid of this species on the spawning 

 grounds. 



Stage-3 individuals, particularly males, were 

 poorly sampled during 1976 (Table 3). This was 

 expected since only fully mature individuals 

 move inshore in large numbers, and reproduc- 

 tive development ceases or regresses in late fall. 

 These animals show the first obvious signs of ap- 

 proaching maturity: Developing eggs and sper- 

 matophores may be visible, and the nidamental 

 and accessory glands of females and the testes of 

 males (Table 3) have reached sizes comparable to 

 those of fully mature squid. The stage may be of 

 short duration, since gonad development appears 

 to be rapid offshore, with large squid, especially 

 males, maturing faster than the smaller ones 

 (Summers 1969; Macy 1980). In the March 1978 

 Argus samples, large numbers of stage-3 squid 

 were identified, and it is unfortunate that other 

 late winter or early spring offshore samples were 

 not available to better document the latter stages 

 of maturation. 



The relatively low classification accuracy of 

 the method for stage-2 squid (81% for males) 

 probably is due to the wide size range of these in- 

 dividuals in the fall and early winter, and to the 

 unusual gonad development of the larger indi- 

 viduals. Sexual regression by gonad resorption 

 to a neutral or inactive state, though relatively 

 common in bivalve molluscs (Sastry 1979), ap- 

 pears to be rare in cephalopods. Regression has 

 been suspected in L. pealei (Summers 1971; 

 Vovk 1972; Arnold and Williams-Arnold 1977) 

 and possibly also in European common squid, L. 

 vulgaris (Tinbergen and Verwey 1945), however. 

 The phenomenon could explain why only a few of 

 even the largest squid (20 cm and larger DML), 



thought to be in their second year (Summers 

 1971; Mesnil 1977), are mature in the lute fall 

 and early winter offshore samples. Lacking a re 

 liable means of aging this species, it is not cur- 

 rently possible to prove that the larger "re- 

 gressed" squid are in factolder than their smaller 

 developing stage-2 counterparts. It is also pos- 

 sible that sexual maturation merely halts at the 

 onset of winter and resumes again in January or 

 February prior to onshore migrations. This hy- 

 pothesis does not explain why the gonads appear 

 to be shrinking, nor would it account for the pres- 

 ence of both very large and small squid at the 

 same stage of development. Unfortunately, L. 

 pealei has not been held sufficiently long in cap- 

 tivity to confirm either supposition. 



Comparisons with Other 

 Classification Methods 



The main assets of the classification method 

 presented here are its objectivity and its ease and 

 speed of use. To be sure, the development took 

 considerable time, especially interpretation of 

 early cluster analyses, but the basic strategy is 

 relatively straightforward and does have inter- 

 nal accuracy checks. If a more detailed break- 

 down of the maturation process were desired, 

 e.g., to examine details of gametogenesis, the 

 same analysis techniques could be applied to ob- 

 jectively identify and separate the various 

 phases. Thus the methodology should be appli- 

 cable to a wide range of biological problems. 



When compared with several other classifica- 

 tion systems, the advantages of the present meth- 

 od become more evident. Two basic classes of sys- 

 tems exist. Those used by Tinbergen and Verwey 

 (1945), Holme (1974), Juanico (1979), and Hixon 

 (1980) are mainly qualitative, in that the pres- 

 ence or absence of one or more characters, con- 

 siderations of color or texture, and estimation of 

 relative sizes or gamete abundance are used. The 

 other group of classification schemes, typified by 

 those of Mangold-Wirz (1963), Hayashi (1970), 

 Vovk (1972), and Durward et al. ( 1978) are quan- 

 titative methods. These schemes may employ one 

 or more subjective judgments or estimates, but 

 rely mainly on objective characters such as rela- 

 tive organ lengths or weights, egg diameters, or 

 spermatophore lengths (absolute or relative) to 

 distinguish successive maturity stages. Both 

 types of classification systems are of value, but 

 only the latter group will be discussed further 

 because their methods are more objective and 



457 



