Evolutionary Remarks

As defined and discussed in this chapter, the formation of true segments in a bound- ary-driven process comprises three steps: (1) elongation of the anteroposterior (AP) axis, (2) segregation of cells into metameric units within one or more germ layers, and (3) patterning within each segmental unit showing the same spatial frequency in two or more germ layers (Balavoine 2015). Note that the second step can be partially bypassed if patterning in one germ layer is imposed directly by inductive signaling from metamers in another germ layer. In lineage-driven segmentation, by contrast, the second step is obviated entirely, because metameric patterning arises directly from the stereotyped lineages of cells produced by the PGZ, as seen during teloblas- tic segmentation in clitellates and in early Platynereis development.

It is intriguing to consider that several annelids may transition from lineage-driven to boundary-driven segmentation mechanisms, for example, during post-embryonic segmentation after the teloblasts have exhausted their capacity for regular blast cell production, or during posterior regeneration, after complete removal of the PGZ (see Chapter 10). Under this scenario, the differences in segment development dynamics described for different annelid lineages could be explained by heterochronic shifts of this transition along developmental trajectories. This shift could also provide an evolutionary mechanism to adjust for changes in life history strategies, for example, changes in the amount of maternal yolk, switches from planktotrophy to lecithotrophy and back, evolution of direct development, and emergence of alternative maternal provisioning strategies such as albumenotrophy or adelphophagy (i.e., when egg sacs contain abortive nurse eggs that provide additional nutritional reserves to the remaining viable eggs).

Evolutionary plasticity in the timing of the switch between lineage- and bound- ary-driven segmentation could also provide a mechanism for evolving heteronomy, fixed segmental counts and even loss of segmentation. Heteronomy denotes the presence of larval segments whose morphology and development are proposed to be qualitatively different from that of most adult segments. In many annelids with a larval stage, embryos develop into planktonic larvae with three or more segments. In some groups these larval segments do not exhibit a strict anteroposterior sequence of development: in Hydroides elegans for example, segments 4 to 7 begin differentiating after segments 8 to 11 have already formed. And in myzostomid annelids, a highly modified parasitic lineage, segment number has become fixed (as in leeches); furthermore, segments do not develop following an anteroposterior sequence of development (Jagersten 1940). Still other annelid lineages have lost segmentation altogether: sipunculans, echiurids, and orthonectids show very limited evidence for segmentation during development, and almost no trace of it in adults (Hessling 2003; Kristof. Wollesen, and Wanninger 2008; Boyle and Rice 2014; Tilic, Lehrke, and Bartolomaeus 2015; Schiffer. Robertson, and Telford 2018; Zverkov et al. 2019). This enormous morphological and developmental diversity shows that the ancestral annelid segmentation mechanisms have evolved in many directions since their last common ancestor, and hint that excessive emphasis on reconstructing ancestral traits and trying to find the homologies across annelids, arthropods, and vertebrates can be misleading and can even blind us from appreciating the true evolutionary richness of diverse animal morphologies and the developmental pathways by which they arise.

< Prev   CONTENTS   Source   Next >