squid was immature but had developing gonads. 

 Based upon 10 growtli measurements (10 data 

 points) from dead squid and several measurements 

 obtained photographically from live squid (4 data 

 points), the following exponential growth equation 

 was generated for L. pealei (Fig. 1) during the first 

 six months: 



Mantle length = 1.30 e^o^os', r^ = 0.98. 



This indicates an approximate mean growth rate of 

 2% increase in mantle length per day. 



Swimming and social behavior were observed 

 carefully. A gentle circular flow (1-4 cm/s) was main- 

 tained in the CT system. Two squid 4 mm ML (41 

 d old) were able to maintain their position against 

 a current of 1.8 cm/s. One week later (day 49), the 

 squid could maintain their position in a current of 

 2.9 cm/s. On day 52, the two squid (approximately 

 5-8 mm ML) first showed schooling behavior by 

 swimming parallel to one another throughout the 

 day. They usually stayed in the darker areas of the 

 tank but frequently moved into the lighter areas, 

 probably to feed on the mysids that concentrated 

 there. The final mortalities were caused by trauma 

 and secondary infections that resulted from bump- 

 ing the walls of the tank (Hulet et al. 1979). 



Discussion 



Since our first small-scale experiments with Loligo 

 pealei (Yang et al. 1980), we have enlarged and re- 

 fined our squid culture methodology enough to allow 

 us to grow L. forbesi and L. opalescens through the 

 life cycle (Hanlon et al. 1985; Yang et al. 1986). 

 Loligo pealei, with its very small hatchling size (1.8 

 mm ML vs. 4.0 and 2.7, respectively for L. forbesi 

 and L. opalescens), continues to be difficult to rear. 

 Although our culture results in this report were poor 

 numerically, they were a vast improvement over the 

 many attempts in the past 100 yr to rear L. pealei 

 (Verrill 1881; Williams 1909; Arnold et al. 1974). 

 Significant improvements over our past L. pealei 

 experiment include 1) increasing the culture tank 

 size from 66 or 99 L to 1,600 L, 2) improved filtra- 

 tion capacity, and 3) feeding the squid many times 

 daily (compared to twice) on different types and sizes 

 of wild-caught zooplankton (compared to small cope- 

 pods only). We cannot explain the high early mor- 

 tality in the present experiment even though it is 

 characteristic of all Loligo spp. rearing experiments 

 thus far; feeding was observed often and the water 

 quality remained in an acceptable range. Predation 

 by wild zooplankton on squid hatchlings was possi- 



ble (e.g., crab megalops), but this factor alone did 

 not cause the high mortality. The small hatching size 

 of L. pealei may partly explain the greater initial 

 mortality (compared with L. opalescens and L. 

 forbesi) since providing food organisms within the 

 proper size range was more difficult. The zooplank- 

 ton offered to the squid during the first week was 

 composed primarily of copepods ranging from 0.5 

 to 2.5 mm long, i.e., 25 to 110% the length of the 

 squid. Nevertheless, squid hatchlings captured and 

 fed upon copepods, often the largest ones. Occa- 

 sionally, however, hatchlings avoided copepods and 

 appeared startled by their jerky movements. Curi- 

 ously, it is this same jerky motion that provides the 

 behavioral stimulus for all Loligo hatchlings to feed 

 upon copepods. However, if enough cannot be cap- 

 tured by the small L. pealei, they may not be able 

 to meet the high energetic costs of pursuit, capture, 

 and digestion of a mobile, armored prey. Another 

 possible contributor to mortality may have been 

 reduced levels of dissolved organic nutrients in our 

 artificial seawater (which was physically, chemically, 

 and biologically filtered; cf. Manahan and Stephens 

 1983) combined with a qualitatively restricted diet 

 compared to nature. 



The value of laboratory data is its potential to 

 verify or refute hypothesized descriptions based on 

 limited or discontinuous fisheries data. Although the 

 number of individuals studied was low (only two 

 squid after day 23), several growth and behavioral 

 patterns can be described. For example, schooling 

 behavior, which depends partly upon size and swim- 

 ming strength, was observed at a similar size for 

 cultured L. pealei (4-6 mm ML, 50-60 d; this report), 

 cultured L. vulgaris (5-10 mm ML, 20-40 d; Turk 

 et al. 1986), and cultured L. opalescens (8-11 mm 

 ML, 40-50 d; Yang et al. 1986). The appearance of 

 schooling behavior may be related to the transition 

 from the planktonic phase to the juvenile and adult 

 demersal (neritic) phase of the life cycle. Increased 

 swimming ability associated with schooling would 

 allow the young squid to migrate vertically and ex- 

 ploit other food sources during the night. Squid size 

 seems to be a key, with all three species first ex- 

 hibiting schooling behavior when the hatchlings are 

 5 to 10 mm ML. Age at schooling is more variable, 

 as early as 20 d for L. vulgaris (the largest hatch- 

 ling) and as late as 60 d for L. pealei (the smallest 

 hatchling). These observations, while limited to two 

 individuals, conform generally to estimates of the 

 end of the planktonic period for L. pealei in nature. 

 Vecchione (1981) estimated that a distinct change 

 in morphometric growth in L. pealei, especially ten- 

 tacle growth, occurred at 4.5 mm ML. A change in 



165 



