摘要:Two nonlethal alimentary tract lavage techniques were conducted on ten bonytail chub (Gila elegans, 211-241 mm TL) that had previously ingested 2-4 rainbow trout (Onchorynchus mykiss, ∼40 mm TL) to evaluate their capabilities of detecting piscivory in the field. Each fish was initially lavaged with water down its buccal cavity and out its anal vent using the Wasowicz and Valdez (W-V) technique and then immediately flushed in the opposite direction with the Baker and Fraser (B-F) method. Microscopic examination of evacuated material revealed only three identifiable rainbow trout scales and other assorted clumps of presumably well-digested body material. Overall, 30% of bonytail chub expelled rainbow trout scales andlor other body material during the initial W-V lavage, and 60% expelled material during the B-F flush. Dissections of four bonytail, seven humpback (G. cypha), and four roundtail chubs (G. robusta) revealed that their intestines were similarly curved and tapered to ≤50% of the anterior widths. Whereas the B-F method pushes ingested items backwards through the increasingly larger intestinal circumferences in these Gila fishes, it is more apt to retrieve greater quantities of food items. Food preference studies based on findings from a highly size-selective lavage technique will likely be biased in favor of smaller, less-digestible prey items, such as invertebrates with chitinous exoskeletons. Use of the B-F rather than W-V technique should lessen this bias.