Does Less-than-Annual Production of Offspring by Female Vipers (Vipera aspis) Mean Less-than-Annual Mating?

Abstract

Copulation does not necessarily result in the production of offspring. Copulatory behavior unrelated to output has been recorded in a wide variety of species (e.g., Hrdy, 1986; De Ruiter et al., 1994; Poldmaa et al., 1995). These matings have been interpreted as nonadaptive by some workers (e.g., as consequences of high sex-steroid levels: e.g., Symons, 1979) but as adaptations by others. For example, females that mate more often than is necessary for reproduction may enhance their nutritional status through nuptial gifts or sperm acquisition (Ward and Landolt, 1995), or may be able to manipulate the subsequent behavior of their male partners (Hrdy, 1986; De Ruiter et al., 1994). Because studies on reptiles have generally interpreted copulations with females strictly in terms of sperm transfer, most herpetologists have tacitly assumed that mating only occurs in circumstances where the sperm that are transferred are likely to contribute to paternity of a female's offspring. The inference is logical and plausible, but definitive evidence is weak. Snakes with less-than-annual frequencies of female reproduction offer an exceptional opportunity to examine this question. In many species of snakes, especially viperids from temperate zone habitats, adult females do not produce offspring every year. Instead, females delay the production of their next litter until they have accumulated sufficient body reserves, a process that may take from one to several years (e.g., Saint Girons, 1957; Fitch, 1960a; Brown, 1991). In consequence, the population each year contains both and females. We define as reproductive those females which contained large (>10 mm diameter) vitellogenic follicles or embryos (detected by palpation or observed at parturition). Females were classified as Non-reproductive if they were caught without enlarged ova or embryos during the period of vitellogenesis and/or gestation (Bonnet and Naulleau, 1996a). This definition is strictly based on physiological events (Bonnet et al., 1994), and thus is somewhat different from a purely behavioral one (e.g., active versus non sexually active females). This clarification is important because the concepts of reproductive versus active although biologically overlapping and often used as synonyms, are not strictly identical. One obvious question is thus: do females mate as usual during their non-reproductive years (i.e., while they are accumulating energy stores), or do they mate only in years when they will undergo vitellogenesis and produce a litter? Surprisingly, there appears to be little reliable information with which to answer this simple question (Schuett et al., 1993). Some authors have inferred that females do not mate except in years (e.g., Macartney and Gregory, 1988), whereas others have concluded that females mate every year despite a less-than-annual schedule of offspring production (Saint Girons, 1949, 1952, 1957, 1996; Saint Girons and Duguy, 1982; Luiselli, 1993; Capula et al., 1995). In most cases, these inferences have been based on relatively limited data. The reason for the lack of information on mating frequency is straightforward; it is easier to determine a female's status (by body shape, etc.: Fitch, 1960a, 1987) than it is to detect whether or not she has mated. Snakes are often secretive at this time, and matings may be difficult to observe under field conditions. Nonetheless, there is a simple and longestablished technique by which prior mating can be determined, by sampling fluids from the female's tract and examining them for the presence of spermatozoa (Fukada, 1959; Fitch, 1960b, 1987; Naulleau 1992; Bonnet and Naulleau 1996b). We have gathered an extensive data set on female aspic vipers (Vipera aspis) using this method, to test Saint Girons' (1949, 1952, 1957, 1996) hypothesis that females of this species continue to mate even during years in which they do not produce offspring. If such matings occur in this species, it would offer an exciting opportunity to test ideas on the factors responsible for such behavior. Snakes were captured in the field in western central France (Loire Atlantique, Vend6e, Charente Maritime, and Deux Sevres) as part of extensive mark-recapture studies on biology of this species (Naulleau and Bonnet, 1996; Bonnet and Naulleau, 1996a; Bonnet, 1997). Since 1990, more than 340 adult females (>41 cm snout-vent length, 47 cm total length: Naulleau and Bonnet, 1996) have been tested for the presence of oviductal sperm following the methods of Fukada (1959) and Fitch (1960b). The posterior part of the abdomen was softly massaged to expel fluids from both oviducts and cloaca. The fluids were then spread across a glass microscope slide and allowed to dry. They were later examined microscopically (60x magnification). Dried spermatozoa are easily identified and no specific preservation method is needed. With this method, the presence of spermatozoa can only be detected for few days after mating. We base this conclusion on hundreds of examinations of cloacal swabs