Viviparity is a form of reproduction in which embryonic development is supported by nutrients in the body of the mother and has evolved multiple times independently in multiple lineages of animals (reviewed in Blackburn, 2015; Ostrovsky et al., 2016). This suggests a convergent evolution event along the animal phylogenetic tree, likely in response to similar selective forces driving similar adaptations. Several environmental factors driving the evolution of viviparity have been assumed in animals (Hogarth, 1976). The best-studied viviparous animals for understanding the evolution of viviparity are squamate reptiles (Blackburn, 1998; Van Dyke et al., 2014). In squamate reptiles, the driving force is believed to be cold climate. This hypothesis is underpinned by the habitat of viviparous species and phylogenetic analyses indicating that most resent transitions occurred in cold climates (Hodges, 2004; Lambert and Wiens, 2013; Watson et al., 2014). However, whether viviparity arose from a common reproductive system and its driving force in animal lineages are unclear.

Nematodes are excellent model organisms for studying viviparity evolution, given their ecological diversity and ease of genetic manipulation with a short life cycle. Most nematodes are oviparous: females/hermaphrodites lay eggs, which hatch into the external environment. Oviparous mothers retain eggs in the uterus under some stressful conditions. The larvae hatched in the uterus consume the body contents of the mother, and emerge through the mother’s body wall, generally resulting in the mother’s death. This facultative vivipary is known as “bagging” (also known as endotokia matricida) (Lordello, 1951; Johnigk and Ehlers, 1999; Chen and Caswell-Chen, 2004). In contrast, females/hermaphrodites of some nematode species obligately retain fertilized eggs within their reproductive track and give birth to larva; these are described as ovoviviparous or viviparous. Here, we define ovoviviparity as birth following intrauterine hatching from an egg from a rigid eggshell without embryonic growth, whereas viviparous animals gestate the embryo in the uterus until the embryos become larvae (Balinsky, 1970). Since nutrition is supplied by the mother in viviparous animals, the embryo typically does not have a rigid eggshell in viviparous animals, such as insects (Tworzydlo et al., 2013), squamate reptiles (Blackburn, 1993a; Blackburn, 1993b; Blackburn, 1998), and sharks (Lombardi and Files, 1993; Heiden et al., 2005). Some nematode species have been reported to be viviparous nematodes, e.g., Trichinella spiralis, filarial nematodes (Smyth, 1994; Hugot et al., 2001). However, in most cases where these features have been adequately described, this vivipary, which is ovovivipary as defined above, is the retention of eggs, wherein embryonic development proceeds as in oviparous species. In nematodes, no examination of nutrient supply and eggshell structure has made a clear distinction between ovoviviparous and viviparous species. A new species of nematode, Tokorhabditis tufae, was discovered in the alkaline, hypersaline, and arsenic-rich environment of Mono Lake, California; being obligately live-bearing, T. tufae is likely viviparous (Shih et al., 2019; Kanzaki et al., 2021). Kanzaki et al. (2021) observed the embryos with differential interference contrast (DIC) microscopy and showed that T. tufae embryos increase in size during embryonic development, suggesting nutrient supply from mother to embryo. Although the magnitude of the increase in size and nutrient provisioning are unclear, the form of reproduction in T. tufae, based on the definition above, is consistent with viviparity rather than ovoviviparity.

To confirm the viviparity of T. tufae and to demonstrate its developmental and morphological uniqueness, we measured embryo size at various developmental stages and calculated the growth rate. Subsequently, we investigated the permeability of substances of various molecular weights in T. tufae and its closely related egg-laying species, Auanema rhodensis. Auanema is the sister group of Tokorhabditis and shares similar features, such as a trioecious mating system (see Figure 9 in Kanzaki et al., 2021 for their phylogenetic relationship). Given the high substance permeability of T. tufae, the ultrastructure of the eggshell and permeable barrier was visualized by transmission electron microscopy (TEM).

We have reported that T. tufae has a viviparous reproductive mode because embryos increase in size during embryonic development. However, we were concerned that the embryos were deformed due to spatial constraints in the gonad and uterus. Here, we measured egg and embryo sizes (area) at three developmental stages in T. tufae and A. rhodensis. Single-cell-stage embryos of T. tufae were smaller than those of A. rhodensis. Lecithotrophic females, which have oviparous embryos, obtain nutrients from the yolk of the ovum and allocate all nutrient resources to the embryo before fertilization, resulting in larger eggs than matrotrophic females, which allocate resources to offspring throughout gestation (Trexler and DeAngelis, 2003; Buddle et al., 2019). Furthermore, in oviparous animals, egg size does not change significantly after fertilization because the mother no longer supplies nutrients during pregnancy. In viviparous animals, egg enlargement occurs after fertilization, which is associated with nutrient supply from the mother (Wourms, 1981; Huveneers et al., 2011). In this study, the egg and embryo sizes of A. rhodensis did not change much during embryonic development, although significant differences were observed in the size of the threefold stage embryos. By contrast, significant differences in egg and embryo size were observed among all stages of T. tufae. T. tufae eggs and embryos at the threefold stage were respectively 2.6- and 3.6-fold larger than at the single-cell stage, suggesting nutrition supply by the mother in the uterus. Notably, removal from the uterus halted embryonic development at the single-cell and lima bean stages in T. tufae irrespective of the solution used for incubation. This indicates that the embryos, at least from the single-cell to the lima bean stage, receive nutrients essential for growth from the mother.

Regarding egg permeability, fluorescence was observed only in the region between the eggshell and the embryo in A. rhodensis (Figure 3). By contrast, all of the fluorescent substances permeated T. tufae embryos (Figure 3). Olson et al. (2012) reported that in C. elegans, high-molecular-weight substances did not permeate embryos because of a permeability barrier rather than the eggshell. In this study, ≤70,000 MW molecules penetrated the eggshells of A. rhodensis and T. tufae. However, in A. rhodensis, the permeability barrier surrounding the embryo inhibited this permeation. Indeed, TEM demonstrated that T. tufae lacks a permeability barrier (Figure 4D), whereas A. rhodensis does not (Figure 4B). The permeability barrier acts as an osmotic barrier to prevent large molecules or toxins in the external environment from transmit into embryos (Olson et al., 2012). In T. tufae, the absence of the permeability barrier resulted in increased molecular permeability. Nutrient substances, e.g., vitellogenin, are typically of high molecular weight. Therefore, the absence of the permeability barrier may be linked to nutrient supply by the mother.

Eggshells have different functions in oviparous and viviparous nematodes. The eggshell of A. rhodensis was composed of VL, ML, and IL (Figure 4A). By contrast, the eggshell of T. tufae comprised only a VL (Figure 4C). Our permeability tests indicated that this difference in eggshell structure is not due to increased permeability, as the fluorescent substances passed through the A. rhodensis and T. tufae eggshells regardless of size. The simplification of eggshell structure is typically associated with the evolution of viviparity (Blackburn, 1993a; Blackburn, 1993b; Heiden et al., 2005). In the eggshell of oviparous animals, a hard layer comprising chitinous or calcareous materials protects the embryos from external stresses. In viviparous animals, the hard layer is absent because there is no need to protect the egg from external stressors. The ML contains chitin and acts as a framework to maintain egg shape (Olson et al., 2012; Stein and Golden, 2018). The absence of the chitin-containing ML likely allows an increase in egg size during embryogenesis. Furthermore, a thinner or absent eggshell promotes gas exchange in the embryo. In reptiles, dissolution or absence of the structure surrounding the embryo, the shell membrane, during pregnancy enhances gas exchange in viviparous taxa (Blackburn, 1998). The absence of the rigid eggshell in T. tufae may facilitate embryonic gas exchange; this warrants further investigation.

Although the nutrients supplied by the mother to the embryo in T. tufae are unknown, the developmental changes in structural and functional features of T. tufae are like those in other viviparous animals. Therefore, T. tufae is a viviparous rather than an ovoviviparous nematode species. Investigation of the ecology of T. tufae is required to identify the evolutionary driver of the transition from oviparity to viviparity. Although T. tufae has only been found in Mono Lake, we recently isolated two other species of Tokorhabditis from dung beetles (Ragsdale et al., 2022). Further, Sudhausia spp. nematodes sometimes cohabit with Tokorhabditis spp. and so are likely to be viviparous (Herrmann et al., 2013; Kanzaki et al., 2017a). Identification of the ecological factors common to these viviparous nematode species may provide insight into the evolutionary forces that drive the transition from oviparity to viviparity. Comparison with the model organism C. elegans will clarify the evolution of the mechanism of viviparity.