Alternative reproductive tactics in male freshwater fish influence the accuracy of species recognition

Abstract

Sexual conflict can result in coercive mating. Because males bear low costs of heterospecific mating, coercive males may engage in misdirected mating attempts toward heterospecific females. In contrast, sexual selection through consensual mate choice can cause mate recognition cues among species to diverge, leading to more accurate species recognition. Some species show both coercive mating and mate choice-associated courtship behaviors as male alternative reproductive tactics. We hypothesized that if the selection pressures on each tactic differ, then the accuracy of species recognition would also change depending on the mating tactic adopted. We tested this hypothesis in the guppy (Poecilia reticulata) and mosquitofish (Gambusia affinis) by a series of choice experiments. Poecilia reticulata and G. affinis males both showed imperfect species recognition and directed all components of mating behavior toward heterospecific females. They tended to direct courtship displays more frequently toward conspecific than heterospecific females. With male P. reticulata, however, accurate species recognition disappeared when they attempted coercive copulation: they directed coercions more frequently toward heterospecific females. We also found that heterospecific sexual interaction had little effect on the fecundity of gravid females, which suggests that prepregnancy interactions likely underpin the exclusion of G. affinis by P. reticulata in our region.

Keywords: Fisherian process; Poeciliidae; male mate choice; mate recognition; reproductive interference.

FIGURE 1

Proportion of association time with Poecilia reticulata female and Gambusia affinis female (a) Association preference of P. reticulata males. (b) association preference of G. affinis males

FIGURE 2

Frequencies of male behaviors toward Poecilia reticulata females and Gambusia affinis females. (a) gonopodial thrust, (b) gonopodial swing, (c) follow, (d) sigmoid display (d), and (e) jolts

FIGURE 3

Size–fecundity relationship in (a) Poecilia reticulata females and (b) Gambusia affinis females. The symbols represent the three treatments (○, control, female alone; +, P. reticulata male; △, G. affinis male), and the lines represent the prediction from the generalized linear model (solid line, control; dotted line, P. reticulata male; dashed line, G. affinis male)

FIGURE A1

Schema of Experiment 1. (A) Experimental setup; (B) procedure. The marked areas on either side of the center tank in A are the preference zones. The thick vertical lines in B are opaque dividers.

FIGURE A2

Schema of Experiment 2. (A) Experimental setup; (B) procedure. The thick line in B is an opaque divider.

FIGURE A3

Schema of Experiment 3, experimental setup.

FIGURE A4

Mating behavior transition network based on the 8 replicates. The line width was corresponded to the average numbers of transition: (A) P. reticulata male and P. reticulata female; (B) P. reticulata male and G. affnis female; (C) G. affinis male and P. reticulata female; (D) G. affinis male and G. affnis female.