1.07.1.1 Why Evolutionary Neuroethology?
The question of how nervous systems and behavior evolve is a core issue in both neurobiology and evolutionary biology. How can something as extraordinarily complex as the neuronal networks underlying behavior (neurobehavioral networks) tolerate change without compromising function? The nervous systems, including the brains, of species within large taxonomic groups look remarkably similar while often producing strikingly different behaviors. Biologists, at least since
Darwin, have puzzled over this conundrum of conserving neural architecture yet diversifying behavior during evolution. Current research using molecular techniques to analyze the development of nervous systems has made remarkable progress in exposing how the conserved genetic tool kit allows the generation of novel features. But are the principles derived from comparative genetic-developmental analyses sufficient to explain how behavior evolves under natural selection? ''It is one thing to know the laws of Nature, and quite another to know the outcomes of those laws'' (Barrow, 1998, p. 66). Evolutionary neuroetholo-gists ask the question: How do nervous systems really evolve? In this article, I reconstruct the pedigrees of the two unusual and surprising modes of locomotion displayed by all species in one decapod crustacean family. Some aspects of the gross anatomy of these animals have diverged so much from the corresponding parts in their close relatives that at first glance they appear unrelated. Why is it important to find out how individual nervous systems have evolved? An ultimate goal of basic neuroscience research is to understand the rationale for the functional organization of neural circuits, including those in the brain. These extraordinarily complex and species-specific networks are not what an engineer would design to execute the appropriately adaptive movements or to register only relevant physical characteristics of each species' environment because they are compromises between inheritance and natural selection. Thus, neither the nervous system nor the behavior of any species is fully understandable without knowing both the neural mechanisms producing its behaviors in its natural environment (neuroethology) and the ancestry of the neurobehavioral circuitry. Motor mechanisms are fundamental to all aspects of behavior, because movements present sensory structures to changing conditions, which leads to active as well as passive sensation. Furthermore, mutations are likely to be more serious (detrimental to survival) in motor control networks, particularly for locomotion, than in sensory systems, because sensory modalities are generally used in parallel to acquire information about the outside world. From this, it is expected that neural evolution would be more conservative in motor control than in sensory systems and, therefore, result in greater conservation of ancestral organization in motor than in sensory centers in the central nervous system of any group of animals (see Relevance of Understanding Brain Evolution, A Tale of Two CPGs: Phylogenetically Polymorphic Networks).
1.07.1.1.1 Prerequisites for tracing the evolutionary history of neurobehavioral circuitry To reconstruct the evolution of a species' functional neural networks requires that four criteria be met:
1. adequately known phylogeny to allow neural differences among species to be mapped in the sequence of their appearance during speciation;
2. experimentally accessible nervous systems (presence of individually identifiable neurons in species exhibiting distinctively different behaviors that are amenable to detailed analysis by kinematic and physiological methods);
3. surrogates for ancestral systems of interest in modern species; and
4. insight into the selective forces that shaped the evolutionary changes in the nervous system under study.
1.07.1.1.2 Hippid sand crabs satisfy the prerequisites for evolutionary neurobehavioral analysis Hippid sand crabs meet the criteria outlined above. Although there are controversies about some relationships within the Decapoda, they are peripheral to the discussion of the lineage of sand crabs. Sand crabs and their close relatives are amenable to investigation by the full spectrum of experimental, anatomical, and physiological techniques (Paul, 1991, 2003). Their one drawback as subjects for evolutionary neurobiological research is the extended time of their embryonic and larval development (Johnson and Lewis, 1942; Harvey, 1993), which severely limits their potential as subjects in genetic and developmental analyses. On the other hand, a fortuitous fact in the history of neuroscience is that crayfish have long been favorite experimental subjects (Huxley, 1880; Edwards et al., 1999) and have retained key traits not only of ancestral decapod skeletomusculature, but also neuroanatomical and behavioral traits that can be traced back to the earliest Malacostraca (Figures 1c and 2; Sections 1.07.4 and 1.07.5). The motor systems in crayfish are thus suitable surrogates for the ancestral motor systems from which sand crab neu-robehavioral networks would have evolved.
Novelties in evolution - such as invasion of new niches or evolution of new locomotor behaviors (which may or may not appear conjointly with the former) - are often associated with loss, reduction, or alteration of ancestral structures rather than evolution of new structures. Structural changes in a neural network, i.e., additions or deletions of neurons and changes in synaptic wiring, correlated with behavioral evolution are difficult to identify. Evolutionary losses of identified neurons have been documented in taxa whose behaviors clearly differ from those of their ancestors (Wilson and Paul, 1987; Antonsen and Paul, 2001; Paul, 2003; Faulkes, 2004), but whether they caused, resulted from, or were incidental to the behavioral change in the lineages in question is not easily recognized. The appearance of new types of neurons in a neurobehavioral network is, on the other hand, likely to be directly linked to behavioral evolution and, when identified, offers unique opportunities to question how nervous systems evolve. The rarity of examples of new neurons may reflect the fact that they are recognizable only by their absence from the homologous (similar to presumed antecedent) neurobehavioral circuit, i.e., appropriate comparative data must be available. The larger the nervous system, the more difficult it is to identify unique, functional contributions of particular neurons in individual species, let alone in related species with divergent behavior, and success in this regard has been greatest in the analysis of neurobehavioral networks in arthropods and mollusks (Katz and HarrisWarrick, 1999; Katz et al., 2001).
1.07.1.3 The Basic Plan of the Decapod Central Nervous System
The basic plan of the decapod body and central nervous system was established with the appearance of Malacostraca in the Paleozoic. Of particular interest here is the architecture of the last, sixth, abdominal ganglion (A6), because it innervates the tailfan, a defining character of the class Malacostraca (Hessler, 1983). This ganglion is more complex than the anterior segmental ganglia, because it is evolutio-narily and ontogenetically the fusion of as many as four neuromeres (Dumont and Wine, 1987; Scholtz, 1995; Harzsch, 2003). Nevertheless, homologous elements in A6 of different species, as well as in A6 and ganglia of anterior segments, can be inferred by their positions relative to recognizable landmarks (Dumont and Wine, 1987; Strausfeld, 1998; Harzsch and Waloszek, 2000; Mulloney et al., 2003). Condensation of neural centers has been a recurrent theme in the evolution of both invertebrates and vertebrates (see Aggression in Invertebrates: The Emergence and Nature of Agonistic Behavioral Patterns). Because the complexity of the malacostra-can A6 is intermediate between that of single neuromeres and that of the larger subesophageal ganglion or the brain, it is potentially informative for comparative research into the ways neural architecture is affected by condensation of neural centers as their functions adapt during morphological and c t a c
Figure 1 Comparison of swimming with the uropods and tailflipping. Direction of movement shown by long arrow. Tracings of single movie frames of (a1) a hippid crab, Emerita analoga, swimming by beating the uropods while keeping the telson flexed under the body and (a2) a tailflipping crayfish. Frames are selected to show form of movements and are not consecutive. Starting from rest position (right frame), hippids extend the uropods rearward (middle two frames) during the return stroke, then rapidly sweep them forward during the power stroke (left frame). When swimming hard (high frequency, large-amplitude uropod strokes), a slight extension of the anterior abdominal segments accompanies the return stroke and is reversed during the power stroke (short arrows over abdomen). Tailflip swimming (a2) begins with an extension of the abdomen (right frame; always evident in electromyograms, although its amplitude depends on the starting posture of the abdomen) prior to the powerful ventral flexion (the power stroke) of the abdomen, which carries the tailfan with it; followed by re-extension (left frame). b, The near immobility of the abdomen and telson during the large ark of the uropod stroke in hippids (b1) contrasts with the large excursion of the abdominal segments (extension-flexion) in crayfish (b2). In both behaviors, the blades of the uropod flare open just prior to the onset of the power stroke and increase surface area and hence thrust is generated. c, The uropods of sand crabs flank the abdomen (right, dorsal view of telson and uropods of E. analoga) rather than the telson as in crayfish, left. d, The external form, particularly of the abdomen and tailfan, of this decapod fossil from the Devonian resembles that of crayfish and other long-bodied, extant decapods. Scale bar (in d): a1, —1 cm; a2, 3-4 cm; d, 1 cm. a1, Adapted from Paul, D. H. 1971. Swimming behaviour of the sand crab, Emerita analoga (Crustacea, Anomura). I: Analysis of the uropod stroke. Z. Vergl. Physiol. 75,233-258. a2, Adapted from Krasne, F. B. and Wine, J. J. 1987. Evasion responses of the crayfish. In: Aims and Methods in Neuroethology (ed. D. M. Guthrie), pp. 10-45. Manchester University Press. d, Adapted from Schram, F. R., Feldman, R., Copeland, M. J. 1978. The late Devonian Palaeopalaemonidae and the earliest decapod crustaceans. J. Paleontol. 52,1375-1387.
behavioral evolution (Paul and Macmillan, 1997; Strausfeld, 1998; Loesel et al, 2002; Mulloney et al., 2003).
Comparing motor patterns, muscles, and neurons in members of selected decapods has revealed that the evolution of sand crabs' novel behaviors occurred in stages and involved deleting from, adding to, and messing around with ancient neurobehavioral circuitry. Loss of certain highly specialized neurons appears to have removed constraints against modifying neuromusculature and behavior and precipitated the evolution of the Anomala by allowing body form and locomotion to change (see Section 1.07.3).
1.07.2.1 The New and Ancestral Modes of Swimming
The swimming and digging behaviors of sand crabs differ between the two families (Hippidae and Albuneidae), and both forms of locomotion are
2 FF/FE muscles
6 Large tailfan
7 CPG for tailflipping
9 Walking (forward + backward) > digging
10 Coordination of legs and tai
11 RS muscle and motoneuron
12 Gait switch
13 PTF ^ uropod insertion
14 CPG: tailflipping ^ uropod beating
16 Fourth legs ~ uropods: phase coupling
Figure 2 Neurobehavioral traits discussed in this article mapped on a partial phylogeny of decapod crustaceans; most prominent exclusion are the Brachyura, true crabs (based primarily on Scholtz and Richter, 1995; see also Martin and Davis, 2001). The organization of the tailfan musculature is the same in shrimp and prawns and members of the first three divisions of the Reptantia (except unknown for slipper lobster), but has been modified in different ways within the Anomala (also referred to as Anomura, which originally included the Thallasinidea). Character 7 (the existence of a central pattern generator (CPG) for tailflipping = nongiant tailflipping) has been demonstrated only in crayfish (Reichert etal., 1981), but is generally assumed to be present in other Astacidea, all Achylata, and shrimp and prawns, and perhaps Thallasinidea. Since slipper lobsters tailflip without the giant interneurons (3,4), either 7 appeared prior to the Achylata (and perhaps is a decapod synapomorphy) or nongiant tailflipping evolved independently in slipper lobsters. Data from Paul etal. (1985), Wilson and Paul (1987); Faulkesand Paul (1997a, 1997b, 1998); Paul (1981a, 1981b, 2004), Paul and Wilson (1994), and Faulkes (2004).
unlike those of any other crustaceans. Most remarkable is hippid crabs' use of the uropods to rapidly propel themselves backward (Figures 1 and 3). These animals keep their abdomen and elongate telson flexed beneath the body at all times and rapidly beat their relatively long uropods through a large arc to swim and, in conjunction with the legs, to dig into sand (Paul et al., 2002). The large-amplitude movements of the uropods, independent of the abdomen and telson, contrast with the tight coupling of the uropods with axial (abdomen-telson) movements in crayfish and other species with tailfans, including albuneid sand crabs (Figure 1b).
Swimming by uropod beating almost certainly evolved from swimming by repetitive tailflipping (Paul, 1991). Three partially separate neural mechanisms are known to mediate the tailflip behaviors of crayfish. Two involve paired giant interneurons (the medial giants, MGs, and lateral giants, LGs), which mediate different forms of powerful ventral flexions of the abdomen and tailfan. The third, known as nongiant tailflipping because neither MG nor LG are involved, mediates repetitive extensions and flexions of the abdomen for rapid backward swimming (Reichert et al., 1981). The giant interneurons were lost sequentially in the lineage leading to sand crabs (Figure 2), and it was the nongiant circuitry for tail-flipping from which hippids' mode of swimming with the uropods is thought to have evolved. The digging behaviors of sand crabs are also unique and not identical in the two families (Hippidae and Albuneidae). They are summarized in the next section.
1.07.2.3 Digging Preceded Uropod Beating: The Mosaic Ancestry of Sand Crab Digging Behaviors
Three pairs of legs (legs 2-4, innervated by thoracic neuromeres 5-7) that other decapods use for walking have been so modified in sand crabs that they are unable to walk across the surface (Faulkes and Paul, 1997a, 1997b, 1998; Paul etal., 2002). Analyses of electromyograms (muscle activity) in individual muscles for each leg joint while sand crabs were digging provided descriptions of the temporal patterns of the motor output to each leg, as well as the coordination between limbs of different body segments. From this, it became apparent that digging is a mosaic of disparate behaviors that are mutually incompatible in sand crabs' walking relatives: walking forward, walking backward, and tailflipping (Figure 4; Paul et al., 2002). The fundamental similarities between the ways all sand crabs dig indicate that digging evolved in their common ancestor and later diverged in certain details following separation
Was this article helpful?