Theories on the Evolution of Seriality and Segmentation
Since the late 19th century a series of hypotheses and theories have been put forward on the evolutionary development of segments and segmentation. It is possible to distinguish between concepts that are more function-based and those that are more structure-based. At the same time, these theories are influenced by the respective view of the phylogenetic relationships of the major groups in the animal kingdom.
Function-based theories have a teleological component and view the reason for a segmental subdivision of a wormlike body as lying in adaptive or construction-dependent necessities. For example, this can be the need for an improved blood supply in connection with an increase in size. The subdivision of an originally uniform coelom space is seen in connection with the emergence of serially arranged blood vessels (Westheide 1997). Alternatively, the possibility of improving the mobility of a worm-shaped body is interpreted as the driving force for segmentation to take place (Korschelt and Heider 1890). The external rings of a hard exoskeleton are considered a functional adaptation (Hatschek 1888-1891). Other authors, on the other hand, emphasize the hydraulic advantages of serially arranged, paired, fluid-filled cavities for animals that burrow (Clark 1964) or exhibit undulatory locomotion (Gutmann 1972). In particular, the “hydroskeleton theory” of Wolfgang Friedrich Gutmann (1935-1997) attempts to explain segmentation and its variations in detail, based on biotechnologically functional plausibility considerations (Gutmann 1972). Gutmann develops hypothetical intermediate forms (Figure 1.6) that are construction-dependent and follow a principle of economy.
Structure-oriented theories of evolutionary segmentation begin with the historical question as to the original structures and the transformation processes that change them. As a rule they follow two patterns, which can be summarized by the conceptual pair: multiplication versus subdivision. The multiplication concepts assume that the parts of an originally uniform, short body are copied and then repeated in series like links in a chain. This can be vegetative budding when the multiplied bodies detach, as occurs in a series of flatworms and annelids, but also in the budding of jellyfish from polyps (Figure 1.7). According to the theory, the individuals generated in this way do not separate; instead they lose their autonomy in the course of evolution and are transformed into segments. This idea, referred to as the “corm or fission theory,” was supported, varying in details, by many influential 19th-century zoologists including Ernst Haeckel (1834-1919), Carl Gegenbaur (1826-1903), and Berthold Hatschek (1854-1941) (Haeckel 1866; Gegenbaur 1870; Hatschek 1878, 1888-1891). This view was recently revived through mathematic modeling of processes of pattern formation by the theoretical biologist Hans Meinhardt (1938-2016) (Meinhardt 2015) (Figure 1.7). Similarities between the corm theory and the pattern of sequential ontogenetic segment formation is evidently not a coincidence, but instead is marked by ideas of a parallelism of ontogenetic and phylogenetic processes as formulated by Haeckel’s theory of recapitulation (Haeckel 1866).
The theories based on subdivision as the mechanism for explaining segmentation view a wormlike form that already has some repeated structures along the body axis as the point of departure for the evolution of segments (Figure 1.8). These segments are ultimately formed in the course of evolution by integrating other repeating elements. The most elaborate theory along these lines is the “tropho- coel theory,” which was developed by Arnold Lang (1855-1914) on the basis of the
FIGURE 1.6 The “hydroskeleton theory” according to Gutmann (modified after Scholtz 2017). Sequence from left to right. From a gelatinous wormlike ancestor with a central intestine and small lateral intestinal diverticula, an organism with hydrostatic pressure opposing the external muscular system is formed by constricting the serial, fluid-filled gastric pouches. These diverticula ultimately detach from the intestine and form paired serial coelom cavities (secondary body cavities). This hydroskeleton serves to improve undulatory movement.
FIGURE 1.7 The “corm theory.” A. A flatworm (Microstomum lineare) that generates complete new worms through budding, which form a chain (modified after Scholtz 2017). B. The modern version of the “corm theory” according to Meinhardt (modified after Meinhardt 2015, with permission from Elsevier). Left: Schematic depiction of budding, as illustrated in the top sketch. The colored stripes represent the expression of the genes that determine the anteroposterior axis of the animal. Right: Depiction of the hypothesis as to how segmentation might have evolved from this mechanism. This occurs through modification of the gene (suggested by the switch from violet to red) with the budded sections remaining together.
FIGURE 1.8 Arnold Lang’s “trophocoel theory” (modified after Scholtz 2017). Top: Sketch of a flatworm (Gunda segmentata), with serial organs (nervous system [red], blind-ended diverticula [light gray], and gonads [grainy, light, oval structures]), the head is facing to the left. Bottom: A hypothetical sequence (left to right) of the evolutionary development of segmented coelom cavities from serial sexual organs. The longitudinal axis of the body is oriented horizontally. Between every two rounded gonads is in each case a short blind-ended diverticulum. These serial diverticula withdraw in the course of evolution, the gonads grow, and finally a series of large, adjacent coelom spaces forms to the left and right of the now- smooth intestines. The vascular system runs between the coelom spaces.
body organization of free-living (nonparasitic) flatworms (Lang 1881, 1903). These animals have no external rings, but some internal organs indicate repeated sections with a corresponding arrangement and rhythm. This pertains in particular to the central nervous system with numerous interconnections, the series of repeated sexual organs along the longitudinal body axis, and the numerous lateral intestinal diverticula (Figure 1.8).
Some theories combine the aspects of subdivision and duplication in that they assume an initial small number of metameres (segments), which then are supplemented and multiplied through replication. One example of this is the “enterocoel theory” (Figure 1.9) proposed by several authors in the 19th century (e.g., Sedgwick 1884; Salvini-Plawen 1998). Adolf Remane (1898-1976) took up these older ideas in the mid-20th century, developing and refining them further (Remane 1950). He assumed there are initially three serially arranged coelom spaces (the anterior one unpaired, the two posterior ones paired), which are derived from the intestinal diverticula (gastric pouches) as they occur in radially symmetrical jellyfish. As bilaterally symmetrical animals evolved, these gastric pouches transformed into coeloms that functioned as a hydroskeleton. This body structure, called archimetamerism, is first transformed into a deutometamere by reducing the two anterior metameres through minor replication of the third section. In a third step this leads to tritometamerism with the actual segments in the trunk of annelids and arthropods, and possibly also chordates.
All these scenarios are necessarily speculative. Also, they are sometimes based on different basic assumptions with respect to the phylogenetic relationships of the major animal groups and the original bilaterian body organization. There are nevertheless clear-cut differences in how well they correspond to existing organismic forms of organization and to today’s view of phylogenetic relationships. Furthermore, the quality of the homology assumptions differ, since not all of the compared structures exhibit sufficient complexity to make a multiple independent evolution improbable. In his hydroskeleton theory, Gutmann (1972), with his notion of construction-depen- dent transformation series, distanced himself too far from actually existing organismic organizational forms. Plausibility scenarios alone obviously do not describe the course of evolution.
As tempting as the corm theory might be due to its proximity to observable ontogenetic processes, a key weakness of it lies in the fact that in no case in the process of segmentation can the anlage of replicated complete organisms be verified. The inner germ layer that forms the digestive tract also does not show a serial ontogenetic subdivision. In addition, it is obvious that in all groups in which such chain formations appear, these depict evolutionarily secondary mechanisms of reproduction. The enterocoel theory, too, suffers under a large share of untestable premises and speculations. It is simply assumed that the last common ancestor of the bilaterians had a serial coelom, which still is a debated issue. In many animal groups, unverifiable structures such as protometameres or other coelom spaces must have been lost over the course of evolution, although there is no evidence whatsoever that they were ever present in the first place. The homology between such unspecific structures as the gastric pouches of cnidarians and coelomic
FIGURE 1.9 The “enterocoel theory” according to Remane (modified after Scholtz 2017). Top row: From left to right this top view shows the evolution of bilateral symmetry from a radially symmetrical jellyfish-like organism. The intestines show four diverticula, the mouth is shown as a small circle (left). As one axis is extended simultaneous to the elongation of the mouth, precursors to the coelom spaces (center) form along the longitudinal axis. Finally, the coelom spaces detach and form the three protometameres (the anterior one unpaired, the two posterior ones paired). The original mouth opening divides into mouth and anus, with a continuous intestine. Middle row: This depicts the stepwise loss of the protometameres (dashed lines, point clouds) and the formation of deutometameres (secondary subdivisions of the posterior coelom cavities). Bottom row: This shows the formation of the tritometameres by means of a budding zone posterior to the deutometameres (dotted line). The segmentation also becomes externally visible.
cavities of some bilaterally symmetrical animals is not well founded, and the distinction between deuto- and tritometameres lacks any foundation (Dohle 1979). The trophocoel theory, as well, is today no longer tenable in the form presented by Lang. There was certainly no linear transformation from a type of flatworm with numerous derived traits to an annelid or other segmented animal. Another weakness is the implicit assumption that the evolution of the coelom was associated with segmentation. All animals with nonserially arranged coeloms would then have experienced a reduction or in some other way would have evolved corresponding structures independently. In a certain way, the subdivision theories on the evolution of segments shift the problem to the level of the development of the original serial structure. In contrast to the multiplication theories, however, this initially has to do with individual serial structures or organs that occur recurrently and not integrated segments that have already been put together from a set of substructures.
Nevertheless, the theory of the origin of segmentation through the stepwise integration of serial structures is most compatible with the different organizational forms on the basis of the major bilaterally symmetrical animal groups. This concept can be easily reconciled with today’s perspective on the evolution of animals, and thus it has also been recently advocated (Budd 2001; Scholtz 2003). Supporting the view of the origin of segments from an initially small number of repeated structures is the fact that a bilaterally symmetrical, wormlike body is not at all conceivable without serial elements. This is already true at the cell level, but also the muscular system and, even more so, the nervous system requires an inherent serial structure with precise positioning in order to interact specifically along a body axis (see Deutsch and Le Guyader 1998). It seems understandable that based on this, other—and, depending on the animal group, different—organ systems would follow this series (see also Chipman 2019). Ultimately, these organs must also be controlled by the nervous system. Also supporting these theories, there are in fact structures that evolved later and which are additionally integrated into the serial rhythm of already existing repeated structures. A good example of this are the vertebrae of vertebrates, which first appear within the segmented chor- dates, and whose serial arrangement is based on serial coeloms, which evolved far earlier. The same is true regarding the external rings and serial nerve ganglia of arthropods. These structures evidently first appeared in the line to the arthropods, as they are absent in the closely related Onychophora (velvet worms), although like arthropods these have serially arranged extremities and other serial structures (Martin et al. 2017).
Such a view of the evolution of segmentation reveals a number of implications. Similar to the process of ontogenesis, in the evolution of segmentation two processes can be distinguished. One is longitudinal growth and the other is the sequential subdivision of the body in serial structures. Longitudinal growth is certainly the evolutionarily older of the two phenomena, since also an unsegmented worm must form its longitudinal body shape from a more spherical egg through directed growth. Furthermore, the aforementioned basic premise of initially homonomous segmentation is disputable, as it obviously conflicts with the at least partial heteronomy that was always observed. If different serial structures are integrated stepwise, then initial differences in the serial arrangement can also lead to an originally heteronomous segmentation.