Evidence of Selection for Vertebral Number of Fry in PeamouthMylocheilus caurinus

Laboratory-reared peamouth fry (Mylocheilus caurinus) were exposed to predation by pumpkinseed (Lepomis gibbosus) or smallmouth bass (Micropterus dolomieui) both before and after vertebrae were developed (i.e. countable) in fry. Vertebral counts of survivors were compared with those of unexposed controls (after further laboratory rearing when experiments were conducted before vertebrae were countable). Predation was selective for vertebral number only in experiments at the larger sizes when vertebrae were well developed in fry. Selection for vertebral number did not vary between temperature (15 or 25 °C) nor between cover (presence or absence) treatments and was not attributable to selection for prey length. Vertebral counts of wild fry were monitored in Holden Lake, British Columbia, between May 14 and July 1, 1984. Differences in vertebral number between age or length classes of fry in the lake that were not attributable to variation in the vertebral counts of fry entering the lake indicated selection for vertebral number of fry over a narrow range of lengths (9.3–10.6 mm). Selection favoured a low number (44) at small lengths and a high number (45) at greater lengths in this range.