Discussion Lessons from Reconstructing Neurobehavioral Pedigrees

Tracing the evolutionary history of an individual neurobehavioral circuit backward through time uncovers sequential convergences with other behavioral lineages (these are more commonly described in the reverse order, i.e., as major branch points) in its lineage, and should ultimately lead to its origin. This approach is akin to the one used by Dawkins to arrive at his series of concestors in his pilgrimage to the dawn of life (Dawkins, 2004). Some discoveries made along the way, at confluences with variants of the circuit whose lineage is being traced, may seem counterintuitive and offer wonderfully assorted jumping off points for new lines of neurobehavioral investigation. Because nervous systems are heter-archically rather than hierarchically organized (Cohen, 1992), each change, whether peripheral or central, in the ancestry of sand crabs, or any other species, would have presented a new challenge to inherited central circuitry.

Ontogenetic variation within populations arises by chance and then evolves to make species anatomically and behaviorally distinct (True and Haag, 2001; Shubin and Dahn, 2004). The initial mutations may affect peripheral or central structure or function. The neurobehavioral lineage of sand crabs presents examples of both (Figures 2 and 10). The pedigrees of individual species' neurobehavioral mechanisms offer more than descriptions of nervous system evolution. The comparative data generate functional hypotheses for individual species that are testable when the species have identifiable neurons that can be studied physiologically.

1.07.6.1 Peripheral Is Important

Patterns of efferent activity produced by the central nervous system are transformed into sequences of movements by the neuromusculature that operates the skeleton (whether internal, external, or hydrostatic). The association of peripheral morphological changes with major evolutionary events, such as vertebrates' movement onto land and the repeated evolution of flight (in insects, reptiles, and mammals), is well known. However, ignoring the contributions that peripheral differences in musculature make to species-specific behaviors risks misinterpreting the roles of central mechanisms in producing those behaviors and may obscure the very features sought in comparative neuroethological studies (Friel and Wainwright, 1997, 1999; Antonsen and Paul, 2000; Hooper and Weaver, 2000). In addition to illustrating well the interplay of peripheral and central factors during the evolution of adaptive behavior in new species, reconstructing the neural pedigrees of sand crab swimming and digging behaviors has opened new avenues for investigating function and evolution of nervous systems in other species that otherwise would have gone unnoticed.

1.07.6.2 Spin-Offs from Reconstructing the Neurobehavioral Pedigrees of Hippid Sand Crabs

The species encountered at convergence points along the backward evolutionary chronology of neurobehavioral circuitry raise new questions, provide new experimental animals, and suggest mechanisms for neurobehavioral evolution that could not be evident from research on a few model systems or in vitro experimentation, useful as the latter are for elucidating potential morphogenetic mechanisms that could lead to similarity by either common descent or convergence (Gerhart and Kirschner, 1997; Hall, 1999; True and Haag, 2001; Antonsen and Paul, 2001, 2002; Eisthen and Nishikawa, 2002; Wainwright, 2002; Wray, 2002).

1.07.6.2.1 Testable hypotheses from comparative studies Because neuronal positions and basic morphologies are conserved between species, comparison of the central positions and morphologies of motoneurons innervating individual heads of complex musculature both within and between species assists in identifying homologous muscles. The power of comparative morphological analysis is beautifully exemplified by the results of its application to the innervation of electric organs in fish (electrocytes are modified muscle fibers), which illustrate well the greater evolutionary plasticity of the physiology than of the morphology of neurons (Labas et al., 2000).

Similar central morphologies of motoneurons in derived and ancestral species suggest that similarly organized central motor networks control them, the root of my argument for homology of swimming by uropod beating and tailflipping. Conserved network structure does not, of course, imply identical function or behavior, as neurons and synapses may evolve individually and be correlated with behavioral differences among species (Shaw and

Figure 10 Partial phylogeny of the class Malacostraca with traits associated with the tailfan and tailflipping behaviors mapped. The tree includes the orders (except Euphasiacea, kril, for which neurobiological date are lacking) that retain the elongate abdomen and tailfan that distinguishes the class. Most Decapoda (e.g., shrimp, lobsters, crayfish; see Figure 2) share with the ancient orders Stomatopoda (mantis shrimp) and Syncarida (represented by Anaspides tasmaniae) all muscular components of the tailfan. A. tasmaniae share with Decapoda the characters of fast abdominal flexor and extensor muscles (FF/FE), the MG and LG interneurons, and the ability to perform rapid tailfips. Non-G (the CPG for tailflipping, character 7 in Figure 2) is shown as appearing with the Decapoda, although it may have evolved later (see Figure 2). The association of abdominal musculature for performing tailflips (not present in stomatopods) and the two sets of giant command interneurons for rapid ventral flexion of the body appeared together in syncarids, which retain many primitive features (Silvey and Wilson, 1979). Note the evolution of the machinery for tailflipping (FF/FE in the abdomen, giant interneurons (MG, LG)) after the divergence of Stomatopoda and the greater antiquity of the tailfan muscles (their number is the same and their organizations are very similar in these three orders, Paul and Macmillan, 1997). Clearly, the enlargement of the tailfan in Decapoda and Stomatopoda, which do not tailfip, evolved independently. N.B. the loss of the AT muscle at the base of the Anomala (absent also in Brachyura, whose relationship with the Anomala are poorly understood). The subsequent derivation of the uropod PS musculature in Hippidae from heterogeneous components of ancient malacostracan tailfan muscles that flex the telson or help cup the tailfan is summarized in Figure 2 and illustrated in Figures 5 and 6.

Malacostraca

Figure 10 Partial phylogeny of the class Malacostraca with traits associated with the tailfan and tailflipping behaviors mapped. The tree includes the orders (except Euphasiacea, kril, for which neurobiological date are lacking) that retain the elongate abdomen and tailfan that distinguishes the class. Most Decapoda (e.g., shrimp, lobsters, crayfish; see Figure 2) share with the ancient orders Stomatopoda (mantis shrimp) and Syncarida (represented by Anaspides tasmaniae) all muscular components of the tailfan. A. tasmaniae share with Decapoda the characters of fast abdominal flexor and extensor muscles (FF/FE), the MG and LG interneurons, and the ability to perform rapid tailfips. Non-G (the CPG for tailflipping, character 7 in Figure 2) is shown as appearing with the Decapoda, although it may have evolved later (see Figure 2). The association of abdominal musculature for performing tailflips (not present in stomatopods) and the two sets of giant command interneurons for rapid ventral flexion of the body appeared together in syncarids, which retain many primitive features (Silvey and Wilson, 1979). Note the evolution of the machinery for tailflipping (FF/FE in the abdomen, giant interneurons (MG, LG)) after the divergence of Stomatopoda and the greater antiquity of the tailfan muscles (their number is the same and their organizations are very similar in these three orders, Paul and Macmillan, 1997). Clearly, the enlargement of the tailfan in Decapoda and Stomatopoda, which do not tailfip, evolved independently. N.B. the loss of the AT muscle at the base of the Anomala (absent also in Brachyura, whose relationship with the Anomala are poorly understood). The subsequent derivation of the uropod PS musculature in Hippidae from heterogeneous components of ancient malacostracan tailfan muscles that flex the telson or help cup the tailfan is summarized in Figure 2 and illustrated in Figures 5 and 6.

Moore, 1989; Fetcho, 1992; Katz and HarrisWarrick, 1999; Katz et al., 2001; Remmers et al.,

2001). Differences between homologous motoneur-ons in squat lobsters and crayfish, for example whether neurites cross or do not cross the midline, may indicate differences in central circuitry correlated with the degree of bilateral interactions between motor centers or reflex pathways (Antonsen and Paul, 2000). Neural differences without apparent behavioral differences occur between two species of squat lobsters (Wilson and Paul, 1987). Future comparative neuroethologists may identify, with hindsight, this difference as incipient behavioral evolution! Behavioral evolution without altering central circuitry may occur by altering neu-romodulatory control, as when key neurons loose, gain, or change their ability to synthesize or release a neurotransmitter or modulator (Harris-Warrick et al., 1992; Katz et al., 2001). An example of altered neuromodulatory control of a complex behavior is the ability of artificially elevated blood serotonin levels to transform normally gregarious squat lobsters (Munida quadrispina) into aggressive and vicious combatants (Antonsen and Paul, 1997,

2002). The details of this uncharacteristic behavior, which is displayed only under the influence of serotonin, strongly resemble the aggressive behavioral repertoire displayed by dominant crayfish, and some other crustaceans, to maintain their position in the social hierarchy (Edwards et al., 1999). We think that fighting is the ancestral condition and its expression was lost in M. quadrispina's lineage while the fight circuitry was retained, at least in part (Antonsen and Paul, 1997). Detailed comparisons of immunocytochemical maps of the nerve cords of crayfish and M. quadrispina could not provide suggestions for which neurons might be involved with fight initiation, but did provide suggestions about possible loci of some other functional differences between these species, along with insights into the evolution of aminergic systems (Antonsen and Paul, 2001).

1.07.6.2.2 Mosaic neurobehavioral evolution from modularity of neural mechanisms and ontogeny

Some themes recur frequently, and similarities by homoplasy are possible. Grillner and Dickinson (2002) caution against associating common themes in motor mechanisms and homology between motor elements, especially when the phylogenetic related-ness of the species is not close. Recognition of convergence requires well-understood functional demands. These are more easily specified for sensory than motor systems, which explains in part the larger number of examples from sensory systems

(Eisthen and Nishiikawa, 2002). The requirement for controlled movements for survival appears to constrain rapid evolutionary change in core elements of motor systems (e.g., CPGs) but allows changes in their coupling and control by other neural centers (Figure 4; Cohen, 1992; Fetcho, 1992; Nishikawa et al., 1992; Remmers et al., 2001; Paul et al., 2002; Vasilakos et al., 2003). Recent advances in our understanding of the onto-genetic processes that can give rise to convergent similarities provide insight into how morphological novelty and selection bring about evolution (Gerhart and Kirschner, 1997; Hall, 1999; True and Haag, 2001; Wray, 2002).

1.07.6.3 The Emerging Picture

1.07.6.3.1 Understanding the present by reconstructing the past A brief look further back into the ancestry of hippid sand crabs' mode of swimming with the uropods, stepping back into the phylogeny of Malacostraca (Figure 10), provides greater depth to the timescale and gives a sense of relatively rapid and recent upheavals in an ancient neurobehavioral network in anomalen decapods. The decapods share with some other, including the most basal, malacostracan orders: (1) organization (but not size) of tailfan musculature, (2) the MG and LG interneurons, and (3) the ability to tailflip. Most interesting is the appearance of these features in the syncarid Anaspides tasmaniae (Silvey and Wilson, 1979; Paul and Macmillan, 1997), which is one of the two most ancient orders of Malacostraca (Figure 10). From the distribution of neurobehavioral traits in Figure 10, a picture of sets of nested changes seems to emerge. The number and organization of tailfan muscles are essentially the same in stomatopods, syncarids, and decapods, although the tailfans differ substantially in size, shape, and behavioral use (Paul and Macmillan, 1997). Not until the loss of the only axial muscle without serial homologues in anterior abdominal segments (the AT muscle) in reptantian decapods was there any change in this ancestral plan.

The well-developed abdominal flexor and extensor muscles and the MG and LG interneurons preceded evolution of the stiff decapod tailfan, with its recessed uropod joint, which increased the power of tailflips. The tailfan (telson and uropods) of crayfish is often referred to as an appendage so tightly integrated are its parts. It is hard to escape the impression that this complex structure of tightly integrated parts with the MG and LG neurons as centerpiece restrained evolution of new morphologies and uses of the malacostracan tail until one or both of these interneurons were lost, as happened apparently several times (Figures 3 and 10). The more ancient AT muscle was apparently an even greater restraint on tinkering with the premiere malacostracan invention, the tailfan, because only following its demise did evolution produce a delightful assortment of anomalan tails, some of which are described in this article. Evolutionary neuroethology, by its comparative analytical approach to understanding motor systems, and other neural networks in general, is becoming increasingly fruitful. This is because the rapid progress in understanding the mechanisms that work at different levels of organization to foster and limit morphological and functional evolution provides plausible explanations for how the neural evolutionary events, charted by comparative anatomy and physiology of extant species, may have come about (Gerhart and Kirschner, 1997; Hall, 1999; True and Haag, 2001; Shubin and Dahn, 2004). These hypotheses are testable if the species in question are amenable to genetic and developmental analysis. Stated in reverse: evolutionary neuroethology tests the plausibility and sufficiency of mechanisms described in research on in vitro systems and model organisms (which model only themselves) to explain how nervous systems really evolve.

1.07.6.3.2 Predicting the future The ability to predict the course of future neurobehavioral evolution in any lineage would be proof that neurobiologists have achieved their goal of understanding the rationale behind the organizations of neurobehavioral networks (Section 1.07.1.1). The degree of understanding that will be required may not be achievable, given the dynamic interactions among the numerous systems at play, many of which are themselves already quite complex (Barrow, 1998). What can be predicted, however, is the course evolutionary neurobiological research will take in the near future to bring us closer to this goal. The combined use of diverse new research tools and methods of computational analysis now available for research at single levels of organization, from synapses and single neurons to large networks and whole nervous systems, will be applied comparatively to carefully chosen species, and the results should greatly deepen our understanding of how nervous systems evolve as conditions change.

Was this article helpful?

0 0

Post a comment