The absence of recognizable antecedents to most structural components of the egg of amniotes, particularly the extraembryonic membranes, has forced speculation on the origin of the amniote egg to focus on functional analogs among modern anamniotes and on the ecological and physiological advantages of specific structures to modern amniotes. The resulting scenarios have emphasized three questions: (1) Which came first, the amniote or the amniote egg (Romer, 1957; Carroll, 1970); (2) what was the order of appearance and functional characteristic of the extraembryonic membranes (Mitsukuri, 1891; Hubrecht, 1912; Szarski, 1968); and 3) what was the reproductive mode of the first amniotes (Hubrecht, 1912; Mossman, 1987).
In contrast to Amphibia, amniotes are terrestrial organisms and reproductive traits are among their specializations for terrestrial life. Romer (1957) considered that all stages of the life cycle likely did not evolve terrestrial specializations simultaneously; that based on the fossil record, the adults of the earliest amniotes were aquatic, and that predation pressure and risk of drying in unstable aquatic habitats would exert high risk on aquatic eggs. Given these three assumptions, Romer (1957, 1967) argued that the amniote egg evolved among aquatic organisms as a response to high egg mortality in aquatic environments. Tihen (1960) concurred on the sequence, but noted that the most likely source of egg mortality was predation, not drought. An alternative suggestion for the ecological conditions selecting for the ancestral amniote egg was proposed by Goin and Goin (1962). They argued that terrestrial breeding habits, specifically internal fertilization, large egg size, and oviposition in terrestrial sites are correlated with humid, montane habitats in modern Amphibia and, by analogy, similar ecological conditions could have selected for the amniote egg. This scenario followed the observation by Lutz (1948) that, based on conditions in Anura in which terrestrial habits are associated with large egg size, an increase in yolk reserves followed by gradual loss of the larval stage of development likely were critical steps in the evolution of the reptilian egg. Elinson (1989) noted that large size is a significant feature of the amniote egg, but because the pattern of cleavage limits egg size the evolution of meroblastic cleavage, followed by the evolution of large vascularized surfaces, must precede increases in egg size.
Carroll (1970) offered an imaginative argument to link reproductive characteristics of the early amniotes to body size and cranial morphology. Because size of terrestrial eggs is limited by respiratory constraints and adult body size of modern Caudata is correlated with egg size, the ancestors of amniotes must have had small body size (<80 mm snout-vent). Larger adult size could evolve only after the evolution of embryonic respiratory specializations released the size limitation on eggs. The evolution of small adult body size thus preceeded the evolution of the terrestrial amniote egg. Carroll's (1970) scenario was founded on an hypothesis that Captorhinomorpha is the sister taxon to all other amniotes and that the amniote egg evolved among primitive captorhinomorphs with small adult body size. Alternative phylogenetic hypotheses (see Lombard and Sumida, 1992) have placed Captorhinomorpha within Amniota. Further, according to recent phylogenetic analyses, the likely sister groups to amniotes are either Diadectomorpha (Gauthier et al., 1988b; Laurin and Reisz, this volume) or Seymouriamorpha (Berman et al., 1992; Lee and Spencer, this volume), both of which are relatively large bodied organisms. If the amniote egg evolved in organisms with small body size as Carroll (1970) suggested, this "terrestrial" egg must have preceeded the evolution of the Amniota (sensu Gauthier et al., 1988a, 1988b).
The amniochorion and allantois are synapomorphies for Amniota for which no clearly intermediate structures are known among Amphibia. Speculation on the evolutionary order of the appearance of these structures has been based on assessment of a possible preeminent functional significance to the ancestral amniote egg. Mitsukuri (1891) considered the frequent development of nonspecific folds in the spreading blastoderm, the universal occurrence of an amniotic headfold, the development of a seroamniotic connection, and the tendency of the head region of the embryo to sink into the large yolk mass to provide clues to both the formation and the function of the earliest amnion. In his view, the amnion arose as a consequence of the mechanics of growth of the blastoderm over the yolk mass and functioned to prevent the embryonic head region from sinking into the yolk. Primitively, the amnion consisted of ectoderm. The mesodermal contribution arose in association with a secondary event-the growth of a respiratory organ, the allantois. Fisk and Tribe (1949) also cite the great tendency for growth of extraembryonic ectoderm and support the view that the ancestral amnion was entirely ectodermal; the seroamniotic connection is the homolog of this structure. However, the evolutionary sequence proposed by Fisk and Tribe (1949) arises from a different functional concern. They argue that a shell membrane, and perhaps eggshell, preceeded the evolution of the amnion that arose to protect the developing embryo from adhering to the outer investment of the egg.
The evolution of the extraembryonic coelom and its relationship to the yolk sac, particularly the transition from a choriovitelline membrane to a splanchnopleuric yolk sac, was a primary event in the origin of the amniote egg (Mossman, 1987). Mossman (1987) also noted that the allantois, originally an embryonic urinary bladder, was the last of the extraembryonic membranes to appear. Szarski (1968) likewise placed primary emphasis on the evolution of the allantois as an excretory organ, but differed in the sequence of membrane appearance by suggesting that the evolution of an amnion and chorion were later events. In both scenarios, the respiratory function of the allantois was secondarily derived.
For Hubrecht (1912), the primitive amniote was viviparous and the key innovation was a chorion-like embryonic envelope. This structure anchored the eggs in the maternal oviduct and absorbed fluids from the uterine cavity. An embryonic vascular system such as the allantois developed later, and the amnion evolved still later as a protective water jacket. Mossman (1987) also believed that the amniote egg evolved as a specialization for viviparity but for different reasons. As in other scenarios (Lutz, 1948; Elinson, 1989), Mossman (1987) emphasized that large yolked eggs could not evolve in the absence of sufficient respiratory support. If a respiratory function for the allantois evolved prior to the advent of large eggs, then the products of these small eggs must have been immature, altricial young, thus requiring parental care. These reproductive characteristics would be most likely to occur in a viviparous species.
These scenarios offer much for consideration but, as reconstructions of historical events, most understandably lack the capacity to be constructed as testable hypotheses. This does not make them any less interesting nor thought provoking but does foster a sense of frustration that such a significant event in the evolution of vertebrates is so elusive. Based on modern species, the structure of the amniote egg differs markedly from that of anamniotes, but even similar characteristics such as oviductal secretions have not been carefully compared. Luckett (1977) predicted characteristics of the predecessor to the egg of amniotes based on the distribution of traits among modern species, and similarly, the pattern of character transformation among extant amniotes can be predicted as phylogenetic relationships become more clearly defined. Such a method was used by Iverson and Ewert (1991), for example. However, there is much that is not known about embryonic development and egg physiology for virtually every major amniote group. For example, details of the development of the nominal structure for the taxon, the amnion, are known for very few species. Reproductive and developmental characters can contribute importantly to phylogenetic analyses (Luckett, 1977, 1993). Unfortunately, the greatest impediment to this approach is the lack of detailed studies of egg development among amniotes.
Was this article helpful?