vol. 158, no. 5 the american naturalist november 2001 Complex Mate Searching in the Satin Bowerbird Ptilonorhynchus violaceus J. Albert C. Uy, * Gail L. Patricelli, and Gerald Borgia † Department of Biology, University of Maryland, College Park, Maryland 20742 Submitted October 19, 2000; Accepted June 5, 2001 abstract: Mate-choice studies typically focus on male traits af- fecting female mating decisions, but few studies seek to identify the behavioral rules females use when searching for mates. Current mod- els suggest that females may either directly compare a set of males (“pooled comparison”) or compare each male to an internal standard (“sequential-search rule”) when judging the suitability of potential mates. Models also differ in other specific aspects, such as the pre- dicted number of sampling bouts initiated and the tendency of fe- males to return to males after previous visits. We monitored 63 female satin bowerbirds, Ptilonorhynchus violaceus, during mate sampling to reconstruct their search patterns. We found that females typically sampled several males and returned to the most attractive male for mating: a behavior consistent with the pooled-comparison tactic. Females, however, varied in the number of males sampled; some visited only one male before mating. We found that this variation can be explained by differences among females in the number of mates, the date mate searching is initiated, and long-term experience with males. Further, females were observed to initiate two distinct sampling bouts, with the rejection of most of their potential mates occurring before the start of the second sampling bout. This suggests that the choices of potential mates are narrowed prior to the second sampling bout and that the later visits may function to reconsider preliminary decisions made during the first sampling bout or to resolve decisions concerning the remaining potential mates. Our re- sults indicate that mate searching is a complex process in which females use multiple sampling bouts to find suitable mates and in which several different factors influence their search behavior. Keywords: mate searching, mate sampling, mate choice, satin bow- erbird, Ptilonorhynchus violaceus. * Present address: Department of Ecology, Evolution, and Marine Biology, University of California, Santa Barbara, California 93106; e-mail: uy@ lifesci.ucsb.edu. † E-mail: borgia@umail.umd.edu. Am. Nat. 2001. Vol. 158, pp. 530–542.  2001 by The University of Chicago. 0003-0147/2001/15805-0007$03.00. All rights reserved. Recent studies indicate that females can gain material (Gwynne 1988), proximate (Borgia 1995b), and genetic (e.g., Hasselquist et al. 1994; Wilkinson et al. 1997) benefits from choosing among males. The net benefits of mate choice, however, are reduced by the costs associated with searching for mates. Thus, to maximize the benefits of mate choice, females need to choose among males as well as use a mate-searching tactic that limits search costs (Ja- netos 1980; Parker 1983). Several theoretical studies ad- dress the behavioral search rules females should use to optimize this trade-off (Janetos 1980; Real 1990; Dom- brovsky and Perrin 1994; Luttbeg 1996; Wiegmann et al. 1996, 1999); however, empirical tests of these models are still needed (see Gibson and Langen 1996; Houde 1997; Widemo and Sæther 1999). Janetos (1980) was the first to explicitly explore female mate searching as an adaptive trait by simulating four mate-searching tactics and comparing the fitness advan- tages of each. Females could employ a random-mating, fixed-threshold, one-step-decision, or best-of-n (called “pooled comparison” by Wittenberger [1983]) search tac- tic. The best-of-n tactic involves females sampling and directly comparing a pool of males, then returning to mate with the male of the highest quality. In contrast, the fixed- threshold and the one-step-decision tactics involve females sequentially sampling males and deciding to either reject or accept males as they are encountered. The decision to accept males is based on a fixed internal standard for the fixed-threshold tactic or the expected benefits of continued sampling for the one-step-decision tactic. Janetos (1980) found that the best-of-n resulted in the highest fitness for the chooser. Parker (1983) expanded the ideas presented by Janetos (1980) by considering search-time costs in his models and found that with an increase in benefits (e.g., more variance in male quality) and a decrease in costs, females become choosier. Building on the efforts of Janetos (1980) and Parker (1983), Real (1990) considered direct (e.g., in- creased risks of predation) and opportunity (e.g., loss of potential mates) costs associated with mate searching. He discovered that, when costs are considered, the best-of-n