(larvae 2.8 to 3.4 mm. long) shows no increase 

 in premaxillary length (« = 0.000), but the 

 second (larvae 3.8 to 4.4 mm. long) and third 

 curves (larvae 4.5 to 6.2 mm. long) indicate 

 positive allometry (a = 9.250 and 2.944, re- 

 spectively), i.e., the premaxillary grows at a 

 higher rate than standard length or a is greater 

 than 1.0. The fourth curve (larvae 6.6 to 10.7 

 mm. long) shows only slight positive allometry 

 (a = 1.273) and the fifth curve (larvae 13.2 to 

 23.7 mm. long) represents negative allometry 

 {a = 0.276), i.e., the premaxillary grows at a 

 lower rate than standard length or « is less 

 than 1.0. 



Growth of the snout less premaxillary can be 

 described by two curves. The first curve (larvae 

 2.8 to 4.5 mm. long) shows positive allometry 

 (a = 1.969) , but the second curve (larvae 4.6 to 

 23.7 mm. long) shows negative allometry (a = 

 0.771). 



Because the very rapid growth of the snout 

 (premaxillary included) would mask the 

 growth of the remainder of the head and unduly 

 influence head-length measurements, the head- 

 less-snout length was examined. Regression 

 lines for head-less-snout length are markedly 

 different from those for snout-less-premaxillary 

 and for premaxillary length (fig. 4). In larvae 

 smaller than 4.1 mm., the head-less-snout 

 length increases relatively slower than standard 

 length (a = 0.727), but, in larvae between 4.2 

 and 4.4 mm. long, it increases much faster 

 (a = 3.500). Abov( 4.5 mm. the growth rates 

 of head less snou and standard length are 

 nearly alike (« = 0.951). 



The sharp inr.ease in the premaxillary and 

 snout affects the appearance of the jaws con- 

 siderably. In larvae shorter than 4.3 mm., the 

 lengths of the upper and lower jaws are equal 

 (table 5) ; this condition is typical of larvae of 

 yellowfin tuna, skipjack tuna, and frigate 

 mackerel. At a length of 4.4 mm., however, the 

 upper jaw becomes longer than the lower and 

 remains longer throughout the rest of the larval 

 stages (the difference in length is still apparent 

 in the 23.7-mm. juvenile). The greatest differ- 

 ence in jaw lengths is in larvae between 9.0 and 

 9.2 mm. long, whose ratios of upper to lower jaw 

 are 1.25:1. This inequality diminishes progres- 

 sively in larger individuals; in the 23.7-mm. 



juvenile, the upper jaw is only slightly longer 

 than the lower (ratio, 1.07:1). In the adults 

 the lower jaw is slightly longer than the upper. 



Table 5. — Ratios of upper jaw length to lower jaw length, 

 upper jaw length to head length, and lower jaw length to head 

 length by standard length for larval wahoo collected in the 

 central Pacific Ocean, 1950-62 



The length of each jaw relative to head length 

 also changes significantly. In larvae smaller 

 than 4.3 mm., the ratios of upper and lower 

 jaws to head length are between 0.508:1 and 

 0.600:1 (table 5). In larvae larger than 4.4 

 mm., however, the length of the upper jaw 

 increases sharply to a maximum ratio of 

 0.769:1 in the 10.7-mm. larva, and the lower 

 jaw attains a maximum ratio of 0.626:1. In 

 larvae above 10.7 mm., the ratio of jaw to head 

 length decreases for each jaw; at a length of 

 23.7 mm. the ratios of the upper and lower jaws 

 to head length are only 0.578:1 and 0.542:1, 

 respectively. 



Jaw development is also unequal in larval 

 and juvenile Scomberomorus (Hildebrand and 

 Cable, 1938 ; Eckles, 1949) . As in larval wahoo, 

 the longer upper jaw of Scomberomorus can be 

 attributed to increased growth of the premaxil- 

 lary. The premaxillary and head lengths of the 

 wahoo were compared with the same measure- 

 ments of sierra mackerel {S. sierra) and skip- 

 jack tuna from the eastern Pacific Ocean 

 (Carlsberg Foundation's "Dana" Expedition 

 1928-30 and Inter-American Tropical Tuna 

 Commission) and central Pacific Ocean (Bu- 

 reau of Commercial Fisheries Biological Lab- 



MORPHOLOGY AND DISTRIBUTION OF LARVAL WAHOO 



309 



