«Acta Herpetologica 3(2): 129-137, 2008 ISSN 1827-9643 (online) © 2008 Firenze University Press Disappearance of eggs during gestation in a ...»
Acta Herpetologica 3(2): 129-137, 2008
ISSN 1827-9643 (online) © 2008 Firenze University Press
Disappearance of eggs during gestation in a viviparous snake
(Vipera aspis) detected using non-invasive techniques
Xavier Bonnet1, Serge Akoka2, Richard Shine3, Léandre Pourcelot4
CEBC, CNRS, 79360, Villiers en Bois, France. Corresponding author. E-mail: email@example.com
LAIEM - Faculté des Sciences, Nantes cedex 3, France
3 Biological Sciences A08, University of Sydney, NSW, Australia
4 Université François Rabelais, UFR de Médecine, TOURS, France Submitted on 2008, 19th May; revised on 2008, 22nd August; accepted on 2008, 26th August.
Abstract. The number of eggs released at ovulation may be greater than the number of offspring born, if some of these ovulated eggs and/or embryos disappear during gestation. Although this process can potentially exert significant effects on reproductive output, logistical problems have discouraged studies on the disappearance of eggs and embryos in most kinds of vertebrates. Nuclear Magnetic Resonance (NMR) imaging and ultrasound Doppler-imaging have not been applied previously to such questions.
Using these techniques, we monitored changes in the female’s oviduct through gestation in a viviparous snake. We documented a case of disappearance of two ovulated eggs (from a litter of four) in the aspic viper, Vipera aspis. The female ovulated four normalsized eggs, two of which contained living embryos when examined by NMR and ultrasound Doppler-imaging early in gestation. Subsequent NMR imaging midway through gestation showed the same situation, but a third imaging session immediately prior to parturition revealed that the oviducts contained only the two live embryos. The two nonviable eggs had disappeared. The female gave birth to the two live offspring, with no evidence of any additional material. These data thus offer the strongest evidence so far available for egg disappearance (resorption?) during gestation in reptiles. More generally, NMR imaging offers a valuable tool for investigating processes inside the body cavity, where direct observation is otherwise difficult or impossible. The technique does not require sacrifice of the animals, and hence allows dynamic investigations over time.
Keywords. Ultrasonography, gestation, Nuclear Magnetic Resonance imaging, reproduction, reproductive output, snake, ultrasound, Vipera.
INTRODUCTIONLife-history theory suggests that the organisms should divide their available resources between maintenance, growth, storage and reproduction in a way that maximizes total 1
lifetime reproductive output (e.g., Stearns, 1989; Roff, 1992). Expenditure of resources on reproduction takes many forms in males, but in females of most animal taxa the primary form of maternal investment involves the production of eggs. Thus the control of egg production (i.e., the number and size of those eggs) is one of the most immediate targets of natural selection. Broadly, such control involves two major physiological mechanisms acting in opposition to each other: recruitment versus regression (follicular atresia, resorption, etc.). A large theoretical, empirical and experimental literature has accumulated on the recruitment of follicles and the maintenance of developing offspring (Lack, 1947, 1954;
Jones, 1978; Skinner, 1985; Duellman and Trueb, 1986; Godfray et al., 1991; Sinervo and Licht, 1991a, b; Thibault and Levasseur, 1991; Jorgensen, 1992; Morris, 1992), but egg resorption has received less attention (Rosenheim et al., 2000). This dearth of attention partially reflects logistical problems: it is difficult to study phenomena that are hidden within the female’s body. Nonetheless, follicular atresia, resorption of ovulated eggs and abortion of embryos have been documented in a wide variety of taxa (Edwards, 1954; Bell and Bohm, 1975; Byskov, 1978; Wilson, 1985; Duellman and Trueb, 1986; Gosling, 1986;
Fox and Guillette, 1987; Thibault and Levasseur, 1991), and have been interpreted through both adaptive and non-adaptive explanations. For example, selective resorption of embryos enables females to manipulate sex ratio in mammals (e.g., Gosling, 1986; Forbes, 1997), and factors such as stress, poor body condition or diseases can induce females to terminate their reproductive effort (Boué and Boué, 1973; Dollander and Fenart, 1979; Hattel et al., 1998). Some species show highly specialized adaptations to reduce egg numbers subsequent to ovulation, including cases of intrauterine cannibalism (oophagy or adelphophagy) in sharks and amphibians (Springer, 1948; Vilter and Vilter, 1960; Hourdry and Beaumont, 1985). These phenomena are somewhat functionally equivalent to direct resorption, in that they notably (but not exclusively) provide mechanisms by which females can reduce their overall litter size.
Although both the ultimate selective advantages and the proximate mechanisms underlying post-ovulatory reduction in litter size are undoubtedly complex, technical difficulties preclude studies of this topic in many wild animal species. In most cases, the existence of intrauterine resorption has been inferred from autopsy. For this reason, the most extensive data sets on resorption have come from species that are killed in very large numbers for other reasons (e.g., coypu Myocastor coypus: Gosling, 1986). Such killing clearly cannot be justified on either ethical or ecological grounds for most species of wild animals. Furthermore, methods based on autopsy of sacrificed individuals do not allow the investigator to monitor resorption processes dynamically, by following individual females through the gestation process (Weintraub et al., 2004). In this paper we present an alternative, non-invasive, method that we have used with snakes, and that could potentially be applied to a wide range of animal species.
Although scientific interest in reptilian reproduction has increased dramatically in recent decades, a surprisingly high number of basic questions remain to be answered. One such topic is the question of embryonic resorption (Blackburn, 1998a). Despite frequent anecdotal reports, and occasional claims in the scientific literature, there has been no reliable documentation of ovulated eggs or embryo resorption in any reptile species. Intuition and logic suggest that reproducing females might often benefit by resorbing nonviable eggs or embryos. Not only may resources be recovered in this way, but females may be Disappearence of eggs during gestation 131 able to remove dead embryos that would otherwise block the oviduct (Blackburn, 1998a, b). Such blockages can have fatal consequences for both the female and the more anteriorly-positioned offspring (pers. obs.). The ability to resorb embryos is well-documented in viviparous mammals (e.g., Westlin et al., 1995), but anatomical differences between the oviducts of reptiles and mammals may preclude resorption in the former group (Blackburn, 1998a, b).
Obviously, much of the difficulty in evaluating reports of egg/embryo resorption in reptiles (as in other groups) involves the fact that the process (if it occurs) is hidden inside the female’s oviducts. Hence, most of the kinds of evidence that have been used to infer resorption are indirect, and open to other interpretations. For example, the production of smaller-than-average nonviable neonates at parturition does not necessarily mean that these offspring were normal-sized at ovulation. Ideally, we need to quantify numbers and sizes of “eggs” at the beginning of embryogenesis (i.e., immediately post-ovulation), and then monitor these variables through the period of gestation. Imaging techniques developed for medical uses are well-suited to this purpose, and in this paper we describe the application of two such techniques (Nuclear Magnetic Resonance imaging and Doppler ultrasounds) to visualizing ovulated (intra-uterine) eggs and developing embryos in a viviparous snake. This technique provides the first direct evidence for the disappearance of ovulated eggs during gestation in any squamate reptile (although we cannot claim any certitude about the mechanisms involved; resorption versus leakage from the oviduct versus expulsion into the peritoneal cavity).
MATERIALS AND METHODS
We used two techniques to examine intra-uterine embryos in gravid snakes. The first of these (Nuclear Magnetic Resonance, or NMR) relies upon the magnetic properties of atomic nuclei when they are exposed to a strong magnetic field. NMR imaging is a non-invasive technique that does not harm the subject, and hence the same animal can be examined on several occasions (Lauterbur, 1973). NMR image acquisition was carried out using a Biospec BMT 24/40 spectro-imager (Brucker) operating at 2.35 Tesla and using a 20 cm Alderman-Grant resonator. Animals were not anaesthetized, but were cooled to 15 oC. The snake was gently restrained using a purpose-built device consisting of a hollow foam-plastic tube cut lengthways and positioned on a Plexiglas rule; the animal was held in position by velcro strips. The time required to obtain an image was 2 min 34 s.
For large snakes, we then moved the animal to another position and repeated the procedure (2 to 4 times depending upon SVL), to ensure that the imaging procedure covered the entire abdomen. The second technique (ultrasonography) is based upon detection of movement by red cells in the circulatory system. We performed these tests using a linear array transducer probe with electronic focusing, at a frequency of 10 MHz and connected to a color-coded Doppler echograph (ESAOTE AU4).
With this equipment, it was possible not only to visualize areas of blood flow, but also to record the velocity of that flow pattern using the Doppler technique.
Data in this paper were obtained from three successive imaging sessions with a gravid female aspic viper, Vipera aspis. The snake was collected in late April from a large wild population (Château d’Olonne, Vendée) in western central France. She measured 54.5 cm snout-vent length (62 cm total length), and weighed 113.8 g at the time she was first examined. She was maintained in captivity throughout vitellogenesis and gestation, first in an outdoor enclosure (6×3 m) during vitellogenesis and the beginning of gestation (April to late July), and then (just before the second imaging sesX. Bonnet et alii sion, until the third imaging session and parturition) in a small cage (40×40×40 cm) with the floor covered with a plastic sheet. Water was provided ad libitum but the female did not accept any food (mice killed by the snake by provoking the bite to induce feeding behavior), as often observed in pregnant individuals. Parturition occurred on 21st August. The cage was regularly inspected (1 to 5 times a day), and we saw no evidence of any expelled ova or other materials (e.g., yolk, membranes, blood, etc. that are very often associated with the expulsion – either of living neonates, stillborn and/or unfertilized eggs) on the floor of the cage at any stage throughout the study. Other reproductive females from the same population (n = 10) have been examined during the same sessions, but the disappearance of eggs was detected in only one. Our sample size is thus very small, and we have no precise idea about the occurrence egg disappearance in this species. However, the fact that we observed the disappearance during pregnancy of ovulated items shows that this remarkable phenomenon can indeed occur in snakes. Furthermore, the main aim of this paper is precisely to present techniques that may be useful to study and quantify such phenomenon.
Palpation on 14th May 1996 enabled us to detect five developing follicles. The first imaging session (19th June) took place soon after ovulation, which occurs in the first two weeks of June in western central France (Naulleau and Bidaut, 1981). NMR revealed the presence of four spherical objects in the oviduct (Fig. 1a). These four objects were very similar in size (length 44 to 47 mm; width 16 to 18 mm). Their size and location made it obvious that these were oviductal “eggs” (although the oviduct[s] that contained the eggs was not identified). This inference was confirmed by direct visual inspection of the embryos using NMR; as expected, the embryos (including embryonic membranes, etc.) were very small at this early stage of development (9 to 12 mm, within the size range we have observed in autopsied gravid females collected in mid June). Only two of the four “eggs” contained discernible embryos. Doppler imaging confirmed the presence of viable embryos in these two “eggs”, and the absence of any heartbeat in the other two “eggs”.
Heartbeats were clearly discernible in the two living embryos, despite their small size.
Comparisons with maternal heartbeat rates recorded either in the oviductal blood vessels or in the heart showed that the signals from the embryos were not artifacts of signals from the maternal heart; the hearts of the embryos were beating faster than the heart of their mother (75 beats m-1 for the embryos versus 45 beats m-1 for the mother). The cranialcaudal sequence of the four live and dead eggs was as follows: 1 dead egg – 2 living eggs – 1 dead egg.
The second imaging session took place 22 days later (10th July), midway through gestation. By this time the two living embryos had increased appreciably in size (19 mm), but no details of their internal anatomy could be clearly resolved (Fig. 1b). The two non-viable eggs had decreased only slightly in size (35 to 47 mm in length and 13 to 14 mm width).