590 



Fishery Bulletin 103(4) 



Characters 



A subset of the molecularly identified istiophorid larvae 

 were examined to ascertain which morphological char- 

 acters might aid in specific identification and possibly 

 obviate the need for future molecular work. The measure- 

 ments made by Richards ( 1974) served as a starting point 

 for quantitative larval descriptions: standard length (SL); 

 snout length (SN); tip of the snout to the center of the 

 eyeball (SN-E); diameter of the eye (ED); diameter of the 

 eye orbit (OD); head length (HL); and difference in length 

 between the upper and lower jaws (JD). To this suite 

 were added measurements of the preopercular (PRO) 

 and pterotic (PTS) head spines. All measurements were 

 taken with Image-Pro Plus software (version 4.5, Media 

 Cybernetics, Silver Spring, MD), and each specimen was 

 viewed through a CoolSNAP-PROcf monochrome digital 

 camera (Media Cybernetics, Silver Spring, MD) which 

 was connected to a Leica MZ12 dissecting microscope (at 

 magnifications 0.8-10. Ox). Each larva was soaked in tap 

 water for one minute before measurements were taken, 

 to rehydrate the fish and facilitate handling. SL and 

 PRO measurements were made from the dorsal view, JD 

 measurements were made from the ventral view, and all 

 other measurements were made from the left lateral view 

 (Fig. 2). Because the preopercular spine often prevents 



an istiophorid larva from lying on its side, a side view 

 was obtained by using the surface tension of the still-wet 

 larva to adhere it to the side wall of a Petri dish. Care 

 was taken to maintain the two points of measurement 

 on a plane parallel to the microscope lens. 



Pigments observed on the ventral surface of the lower 

 jaw rami, gular membrane, and branchiostegal mem- 

 branes of each larva were drawn onto a generalized dia- 

 gram of the larval istiophorid lower jaw (Fig. 3). A grid 

 was then superimposed on the diagram, and the shape 

 (pointate or stellate) and number of chromatophores in 

 each grid cell were recorded. Pigment data were also 

 recorded as binary presence or absence per grid cell. 

 Two other categorical variables assessed were flexion 

 stage (i.e., preflexion, flexing, postflexion) and the posi- 

 tion of the tip of the snout with regard to a plane passing 

 through the center of the eye and the mid-line of the body 

 (i.e., below, even, above). Although the latter character is 

 useful for identifying Indo-Pacific istiophorids (Ueyanagi, 

 1963. 1964), in our collection it was highly variable with- 

 in species, and therefore it was not analyzed further. 



Month of capture was considered a partially discrimi- 

 nating character based on differences in the length and 

 timing of spawning seasons of local populations. Spawn- 

 ing seasons were determined by de Sylva and Breder 

 (1997) by gonad histology studies. 



