504 



Fishery Bulletin 98(3) 



Table 6 



Results of fixed-factor ANOVAs relating (A) sex ratio (females/total) versus period (spawning = Mar^Jul, nonspawning = Aug-Feb). 

 latitude class (North: >27°N, South: <27°N) and longitude class (East: <160'W. West: >160-Wl; (B) the relative number of fish 

 smaller than the 25th length percentile (RNQl, see text! versus period and latitude class (Far S: <22=N, Other: >22"N); and (C) 

 the relative number offish larger than the 75th percentile (RNQ3, see textl vs sex and latitude class (Far N: >35-N. Other: <35"N) 

 on sets with swordfish present, for swordfish, Xiphias gladius. caught by the Hawaii-based pelagic longline fishery during March 

 1994-June 1997. NS terms (e.g. longitude! are unspecified. SS = sum of squares; MS = mean square; LSMEAN = least square mean; 

 SE LSMEAN = Standard error of least square mean. 



A Source 



df 



SS 



MS 



f-value 



Pr>F 



Latitude 



Period ^ Latitude 



Case 



Period x Latitude 



Nonspawning 

 Nonspawning 

 Spawning 

 Spawning 



B Source 



df 



SS 



MS 



F-value 



Pr>F 



Period 



Latitude 



Period > Latitude 



Case 



O Source 



df 



SS 



MS 



F-value 



Pr>F 



Sex 



Latitude 



Sex X Latitude 



Case 



Acknowledgments 



We thank Don Peterson, Tim Price, and Lyie Enriquez 

 of the NMFS, Southwest Region (SWR, Long Beach, CA) 

 office and Tom Shcarei-, Kevin Bu.sscher, and numerous 

 shipboard observers (anonymous for confidentiality) of the 

 SWR, Honolulu Office, for the organized collection of spec- 



imens and data management necessary to complete our 

 study. Eric Yamanoha (Pathology Department, Queen's 

 Hospital, Honolulu) ably supervised the histological work. 

 We also thank Chns Hoggs, Gerard DiNardo, Kurt Schae- 

 fer, and two anonymous reviewers for constructive criti- 

 cisms of manuscript drafts and Debra Yamaguchi for help 

 with the figures. 



