Mate choice in zebrafish (Danio rerio) analyzed with video-stimulus techniques.

Abstract

In many species, individuals of both sexes have developed a variety of visual signals and behavioral patterns with which to broadcast their quality as mating partners (1). The complexity of these signals makes it diffi cult to distinguish those that are most important in mate selection. Animated models offer a solution to this problem by allowing for the alteration of single parameters in the complex stimulus presented. In this study, we have tested the use of computer animated three-dimensional models to analyze mate choice criteria in zebrafi sh; we applied this tool to examine the roles played by two visual characteristics in mate selection. Sexually mature wild-type zebrafi sh, aged at least 8 months, were kept in 9-1 aquaria under a day/night cycle of 14/10 h. Subjects were tested in a 50 cm 32 cm tank; a vertical line drawn on the wall of the tank divided it into two equal sections, A and B. The tank was placed between two 17-inch Dell computer monitors on which animated models of swimming zebrafi sh were displayed. The models were created using 3D Studio Max 1.0 (Kinetix) on a Dell Optiplex GXPro computer and a Targa 1000 board for digital/analog conversion of video signals, as described by Rosenthal (2, 3). Before each trial, an individual fi sh was placed in the tank and allowed to acclimate for 5 min. During each trial, the subject was simultaneously shown two different animated stimuli. Each trial consisted of four 5-min viewing periods separated by 1-min intervals during which black covers were gently slid in front of the monitors. Three pairs of stimuli were shown to the subjects: [1] male versus female body shape (both with natural horizontal stripes), [2] vertical versus horizontal stripes (both with female shape), and [3] vertical versus horizontal stripes (both with male shape). The vertical and horizontal stripe patterns had equal amounts of blue coloration. Female-shaped images differed from male-shaped images only in that their bellies were 10% larger in side view. Stimulus pairs [1] and [2] were presented to both male and female subjects, while stimulus pair [3] was presented to female subjects only. In addition, stimulus pair [1] was shown to females who had spawned on the day of the trial. (The females used in all other trials did not spawn on the day of the trial.) The two animated stimuli were alternated between the monitors to balance for side effects. During each viewing period, the location of the fi sh was recorded every 10 s. The percentage of time the subject spent in proximity to each stimulus (presence on side A or B) was calculated and compared using a Wilcoxon matched-pairs signed-ranks (WSR) test. The results and statistical evaluations are shown in Figure 1. Males did not differentiate between the male and female-shaped images but showed a signifi cant preference (19.6%) for the horizontal stripe pattern over the vertical stripe pattern when the images were female-shaped. Female zebrafi sh preferred a male-shaped stimulus over a female-shaped stimulus by 20.3%. However, females that had just spawned eggs on the morning of the trial did not show a preference for either the male or the female shape. Females also showed a signifi cant preference (10.7%) for the horizontal stripe pattern over the vertical stripe pattern when the images were male-shaped, but did not differentiate between stripe patterns when the images were femaleshaped. The preference of females for the male-shaped stimulus over the female-shaped stimulus indicates that belly size alone allows female zebrafi sh to distinguish between sexes. The indifference of the females who had recently laid eggs suggests that females’ interest in males correlates with their reproductive stage. The signifi cant bias of males and females against vertical stripe pattern in the opposite-sex animation may result from selection against mating with heterospecifi cs.