TESTS FOR SEXUAL INCOMPATIBILITY BETWEEN TWO NEWT SPECIES, TRITURUS VULGARIS AND TRITURUS MONTANDONI: NO-CHOICE MATING DESIGN.

Abstract

The evolutionary origin of reproductive isolation is still a source of considerable debate, although the consensus is that it develops by a gradual accumulation of genetic differences between populations, mostly in allopatry (Dobzhansky 1937; Muller 1942; Mayr 1963; Rice and Hostert 1993; Wu and Davies 1993; Templeton 1996). Quantitative studies of sexual isolation usually go no further than the calculation of the frequencies of homoand heterospecific matings; direct observations of courtship are rarely made (Spieth and Ringo 1983; Verrell and Arnold 1989). Consequently, relatively little is known about the role played by sexual behavior in maintaining reproductive barriers between species. Sexual isolation occurs when individuals tend to avoid mating with those of another strain, race, or species (Gilbert and Starmer 1985). Isolation estimates can be effected by mating propensity, or sexual vigor: the likelihood that a male or female will mate regardless of partner type. Moreover, mating propensity is likely to obscure estimates of asymmetry in sexual isolation unless controlled by experimental design or statistical analysis. Asymmetry is present when females from population A mate more frequently with males from population B than females from population B do with males from population A. Another source of error in both isolation and asymmetry estimates is the effect of social interactions. The most commonly used designs of mating experiments, in which individuals are given simultaneous choice of mates, do not control for intrasexual interactions and therefore can produce biased results (discussed in Arnold et al. 1996). We conducted a laboratory experiment with two species of European newts, the smooth newt, Triturus vulgaris, and Montandon's newt, T. montandoni, in which single males were paired with single females in a variety of crosses. This permitted a comparison of mating probabilities for four possible crosses. Our main goal was to determine the duration of courtships and identify the behavioral causes of sexual isolation. We also sought to demonstrate asymmetries at different stages of behavioral isolation between these two species. Although mating behavior of Triturus newts has been carefully described (Halliday 1977; Arntzen and Sparreboom 1989), surprisingly little information is available concerning the basis of sexual isolation between species. The genetic distance between T. vulgaris and T. montandoni is the lowest among species of the genus Triturus (DNei = 0.16; Rafin'ski and Arntzen 1987), and they readily hybridize in montane areas wherever they coexist (Hofmann 1908; Rafin'ski, unpubl. data). Old World newt species, including T. vulgaris and T. montandoni, possess elaborate and complex mating behavior that involves a prolonged male display followed by spermatophore transfer. Display consists of three distinct tail-movements produced by the male: whip, fan, and wiggle (detailed descriptions in Halliday 1974, 1975, 1976, 1977; Pecio and Rafin'ski 1985; Arntzen and Sparreboom 1989; Michalak 1996). Despite large differences in morphology of secondary sexual characters between the two species, there is only a slight difference in male courtship behavior, primarily in relative frequencies and duration of tail movements. In most cases the male initiates the courtship. The duration of display varies considerably and depends mainly on the reaction of the female. During spermatophore transfer, the female follows the male, walks over the spermatophore, and if her cloaca comes into contact with it, the spermatophore is then taken up into her cloaca. Cooperation on the female's part is an essential prerequisite both for spermatophore deposition and successful transfer. The onset of display, spermatophore deposition and transfer constitute distinctive events in the courtship of newts. Thus the composite courtship of newts provided us with an opportunity to analyze sexual isolation separately for the different mating phases.