Aplysia feeding biomechanics
Studies of the biomechanics and neural control of feeding behavior in the marine mollusk Aplysia californica have clarified the mechanisms of motivated behavior, learning, neuromodulation, the neuromuscular transform, and multifunctionality. Consummatory feeding behaviors in Aplysia are mediated by a muscular structure, the buccal mass. Studies of the buccal mass have demonstrated that multifunctionality - the ability of the same periphery and neural controller to generate qualitatively different behaviors - emerges from the interaction of neural control and biomechanics. These studies demonstrate that biomechanical context can change the functions of motor neurons and interneurons during behavior. Moreover, feeding behavior in Aplysia uses neuromechanical equilibrium points for control. These studies demonstrate that adaptive behavior emerges from the interaction of the dynamics of the nervous system, the biomechanics of the body, and the dynamics of the environment.
Feeding in Aplysia as a Model System
The behavioral and neural mechanisms controlling feeding in Aplysia have been studied as a model system for complex behavior. In particular, Aplysia feeding has been studied to understand the mechanisms of:
- Motivation: arousal, satiation, and the neural correlates of motivated behavior
- Learning: associative learning that food is inedible; classical conditioning and operant conditioning
- Neuromodulation: extrinsic and intrinsic neuromodulation
- The neuromuscular transform: role of multiple transmitters and muscle biophysics in force generation
The mechanisms of multifunctionality are the focus of this article.
Buccal Mass Anatomy
The buccal mass of Aplysia is a structure consisting of muscle and cartilage that the animal uses for feeding and egg laying. A schematic illustration of the buccal mass is shown in Figure 1. The tongue-like central structure consists of a thin, cartilaginous sheet (referred to as the "radula" or rasper) covered with fine, backward facing teeth, whose opening and closing are controlled by a series of muscles associated with the odontophore (tooth carrier). The odontophore is a cartilaginous structure that provides support for the radula.
The muscles of the odontophore that mediate opening and closing are illustrated schematically in Figure 2. Each of the intrinsic muscles of the buccal mass is labeled with an "I", followed by a number. The horse-shoe shaped I4 muscle makes up the majority of the bulk of the odontophore, and contains within it cartilaginous structures known as "bolsters" that stiffen it and constrain its motion. The radula is continuous with the radular stalk. When the radular stalk is pulled towards the radular surface by the I7 muscles, the radular halves open; when the radular stalk is pushed away from the radular surface by activation of the I4 muscle, the radular surface is pulled downwards and closes together tightly. The I5 muscle (also known as the accessory radula closer or ARC muscle) can assist closing. The small muscles around the circumference of the collostylar cap are referred to as I8, I9 and I10. The muscle that forms the back surface of the odontophore (viewed from the jaws), which has been partially dissected away in the top panel of Figure 2, is known as I6.
Surrounding the radula and odontophore (the grasper) are muscles whose activation can move the grasper towards the jaws (this movement is referred to as protraction) or towards the esophagus (this movement is referred to as retraction). The I2 muscle is a thin, sheet-like muscle whose activation can protract the grasper towards the jaws ( Figure 1). The thin I1 muscle surrounds the thicker bands of the I3 muscle, which in turn overlay a circular lumen of cartilage that is referred to as the jaws ( Figure 1). When the I1/I3/jaw complex is activated, if the grasper is in its resting position, the grasper retracts towards the esophagus.
At the base of the grasper are a series of interdigitating fibers that connect the I4 and I6 muscles of the odontophore with the I3 and I2 muscles. These fibers are referred to collectively as the "hinge" (although the structure is not rigid) because the grasper can rotate about these fibers as it is protracted, and activation of the hinge can contribute to retraction.
The buccal mass is suspended within the head of Aplysia by thin extrinsic muscles ("E" for extrinsic, followed by a number; muscles E1 - E6) that can rotate the entire buccal mass dorsally and ventrally, and from side to side, and contribute to the efficiency of feeding behavior.
By combining protraction and retraction with opening and closing, an Aplysia can use its buccal mass to attempt to grasp edible material (biting), to ingest material that it has succeeded in grasping (swallowing), and to remove inedible material from the buccal cavity (rejection).
A variety of techniques have been used to relate the anatomy of the buccal mass to its function. Some of these techniques may be of general interest for the analysis of other soft tissue structures (for example, tongues, trunks or tentacles) which have been collectively termed (by William Kier and Kathleen Smith) muscular hydrostats. Studying these structures is especially challenging because there are fewer constraints on their degrees of freedom, but very interesting, since they are capable of complex dexterous movements.
Initial studies on the anatomy of the buccal mass were done as part of studies of the anatomy of Aplysia by the great naturalist Georges Cuvier (1803) and by Nellie B. Eales (1921). H. H. Howells published a more detailed description of the functional anatomy of the buccal mass of Aplysia punctata in 1942. Initial studies of the buccal ganglia, the collection of nerve cells controlling the buccal mass, were published by Daniel Gardner in 1971, and initial descriptions of the movements of the buccal mass during biting, swallowing and rejection were published by Irving Kupfermann in 1974 as part of his study of Aplysia feeding as a model for motivated behavior.
The functional anatomical studies of Howells led him to propose that contraction of I1 and I2, as well as the dorsal extrinsic muscle E1, pulled the entire buccal mass forwards. Forward rotation of the odontophore was due to the contraction of the muscles I6 and I4, which also induced opening of the radular halves. Retraction was due to contraction of the I3 muscles, as well as extrinsic muscle E2, aided by the I5 muscle. Subsequently, studies by Colin Evans and colleagues in 1996 demonstrated that activation of the I7 muscle induced the radular halves to open.
Videography and extracellular recording
Elizabeth Cropper and colleagues recorded extracellularly from the I5 (ARC) muscle while videotaping biting, swallowing and rejection behavior in 1990, and demonstrated that the muscle was differentially activated in the three behaviors.
Douglas W. Morton and Hillel J. Chiel recorded extracellularly from the nerve innervating the radula/odontophore (radula nerve, RN) and from one of the nerves innervating the I1/I3/jaw complex (buccal nerve 2, BN2) in 1993 while videotaping biting, swallowing and rejection behavior.
They observed that ingestive behaviors (biting and swallowing) are mediated by protracting the grasper with the radular halves open, and then retracting them with the radular halves closed. They showed that ingestive behaviors could be recognized by activation of the I1/I3/jaw complex (mediating retraction) at the same time as the I4 muscle is activated (mediating closure).
During egestive behavior (rejection), they observed that the radula protracted closed, pushing inedible material out of the mouth, and then retracted open. They demonstrated that egestive behavior could be recognized by activation of the I4 muscle (closing) prior to activation of the I1/I3/jaw complex (retraction). They showed that these shifts in the phase of opening/closing relative to protraction/retraction could be observed both in intact, behaving animals and in a semi-intact preparation in vitro. These criteria have been widely used by other investigators to distinguish ingestive and egestive behavior in reduced preparations and in isolated ganglia.
To visualize the entire buccal mass during feeding in an intact animal, Richard F. Drushel and colleagues obtained perpendicular lateral and dorsal views of transilluminated juvenile Aplysia during swallowing in 1997 ( Figure 3). This study, in addition to other neurophysiological and lesion studies done by Itay Hurwitz and colleagues in 1996, supported the hypothesis that contraction of the I2 muscle was crucial for protraction. The study also provided evidence for movement of the radular stalk into the I4 muscle during opening, and movement of the radular stalk out of the I4 muscle during closing. The internal rotations of the entire buccal mass were consistent with activation patterns of the extrinsic muscles that had been measured by Hillel J. Chiel and colleagues in 1986.
Irina Orekhova and colleagues used sonometric measurements of the surface of the radula to examine its movements in response to activation of the motor neurons (B15 and B16) that had previously been shown (by Joshua L. Cohen and colleagues in 1978) to control the I5 (ARC) muscle. They showed that the motor neuron for the I4 muscle (B8) caused clear closing movements. In contrast, the effects of activating the B15 and B16 motor neurons were more complex, producing relatively small closing movements and strong inward movements. Moreover, the strength of closing was dependent on the prior position of the radula.
David M. Neustadter and colleagues developed a novel apparatus that allowed them to record magnetic resonance images (MRI) of the buccal mass during feeding behavior in a whole body MRI machine ( Figure 4). The apparatus provided 1 mm spatial resolution for buccal masses that were approximately 20 mm in each linear dimension, and could record approximately three images per second, which provided sufficient temporal resolution for feeding behaviors, which range in duration from three to ten seconds. To ensure that images were obtained through the sagittal plane of the animal, interleaved orthogonal images (i.e., axial and coronal images) were obtained at lower spatial and temporal resolution. The feeding behavior of Aplysia appeared to be completely normal during the imaging process. MRI images served as the basis for detailed modeling of the buccal mass and analysis of its biomechanics during biting, swallowing and rejection (see below).
In addition to the experimental techniques described above, insight into the function of the buccal mass has been obtained by developing a series of increasingly sophisticated quantitative models. These computer simulated models have made experimentally testable predictions. Two main classes of models have been developed: kinematic models, i.e., models that represent the geometric relationship between elements of the buccal mass, and kinetic models, i.e., models that represent the forces among the muscles of the buccal mass.
A sequence of kinematic models of the buccal have been constructed. The initial model ( Figure 5) assumed that the grasper had a fixed spherical shape, the I3 was represented as a series of six isovolumetric tori, and the I2 was represented as a simple hemisphere. The overall shape of the model was compared with transilluminated shapes measured in vivo during swallowing. The model corresponded well to the shape observed during protraction, but was a poor fit otherwise, suggesting that the grasper changed shape during the retraction phase of swallowing. Models based on changing grasper shapes generated significantly better fits throughout the feeding cycles. Models based on the MRI data generated excellent fits (within a symmetric difference error of about 10%) to actual coronal and axial images obtained in vivo, which were not used to generate the parameters of the model (images from the most detailed kinematic model are shown below).
A sequence of kinetic models of the buccal mass have also been constructed. Sung-Nien Yu and colleagues analyzed the force/frequency, length/tension, and force/velocity properties of the I2 muscle, and constructed both Hill-type and Huxley-type (crossbridge based) models. The initial kinetic model, constructed by Gregory P. Sutton and colleagues, used the same mechanical components as the initial kinematic model. Its kinetic parameters were set by the biomechanical studies of I2, and by Sutton's analysis of the biomechanical properties of the hinge muscle. More recently, a kinetic model constructed by Valerie Novakovic and Sutton ( Figure 6) incorporates a grasper whose shape changes from spherical (corresponding to the shape of the grasper when it is open) to ellipsoidal (corresponding to the shape of the grasper when it is closed).
Analysis of the kinetic model, and experimental tests of some of its predictions, suggest that the mechanics of the buccal mass can be understood as a first order, quasi-static system. The mechanics are primarily first order, i.e, forces depend on velocities and positions but not on acceleration, because movements are slow and the mass is small, so that inertial forces (the product of mass and acceleration) are negligible compared to muscle forces. Even when an animal generates high forces (e.g., on resistant seaweed), it does so slowly. Because the muscle forces change slowly relative to the resonant frequency of the buccal mass, the system reaches its equilibrium state so quickly that a quasi-static analysis is possible. In a static system, the forces acting on the system must be in equilibrium. In a quasi-static system, slowly changing muscle forces move the equilibrium state of the system, and it is the motion of this equilibrium state that results in motion of the system.
Analysis of the kinetic model of the buccal mass has also shown that the balance of forces within the buccal mass creates neuromechanical equilibrium points that determine the current and future state of the system. As the balance of forces changes due to changes in neural activation, the changes in the shapes of the muscles themselves, or external loading (i.e., from seaweed), the location of an equilibrium point may shift. Moreover, as some muscles are activated and others relax, new equilibrium points may be created. Finally, during some behaviors, the control may shift the system to either side of an unstable neuromechanical equilibrium point, so that the system moves from the basin of attraction of one stable equilibrium point to that of another stable equilibrium point.
Neural and Mechanical Mechanisms of Multifunctionality
How can the same neural circuit and the same peripheral structures be used to generate multiple behaviors? A multifunctional engineered device (e.g., a Swiss Army knife) consists of several devices with single functions (e.g., knife, scissors, awl), which can only be used one at a time. In contrast, multifunctional biological devices (e.g., the human hand) can flexibly switch among multiple different uses of degrees of freedom (e.g., typing, pounding on a table, or unscrewing a jar), and can sometimes blend different functions (e.g., grasping and pulling). The detailed analysis of the biomechanics and neural control of the buccal mass of Aplysia has provided insights into the neuromechanics of multifunctionality.
Aplysia generate swallowing responses once they have grasped edible food. The grasper is protracted open to reposition it further forward on food. The grasper then closes, and very strongly retracts, pulling the food deeply into the buccal cavity (see Figure 7). Thus, swallowing is characterized by a relatively weak protraction, and a strong retraction (see Figure 8, top panel). Opening occurs during protraction, and closing occurs during retraction.
Hui Ye and colleagues demonstrated that swallowing of a polyethylene tube (or of a rigid tube-shaped seaweed stipe) occurs in two forms: a smaller amplitude form (Type A) and a larger amplitude form (Type B).
During Type A swallows, the functions of the muscles are straightforward: activation of the I2 muscle protracts the grasper, activation of the I4 muscle closes the grasper, and activation of the I3 muscle retracts the grasper ( Figure 8, middle panel).
During Type B swallows, the I2 muscle is activated more strongly, inducing a stronger protraction ( Figure 8, bottom panel). As a consequence, a cylindrical piece of stiff food is held by the tip of the radular cleft rather than resting within the cleft (as it does in a Type A swallow). When the I4 muscle is activated, it not only closes the radula, but induces the food to rotate ventrally and to translate into the buccal cavity, so that the I4 muscle now functions both as a closer and as a retractor. Because the odontophore is more strongly protracted, the hinge is stretched. Retraction is initiated by activating the hinge, after which activating the I3 muscle completes retraction. During Type A swallows, the hinge is not stretched; as a consequence, in the context of a Type A swallow, activating the motor neuron for the hinge has no effect. In contrast, during a Type B swallow, activating the hinge motor neuron plays a critical role during retraction. Thus, the function of motor neurons depends on their biomechanical context.
Using the kinetic model, it is possible to analyze the dynamics of swallowing. As I2 is activated, it protracts the grasper, stretching the hinge, which generates antagonistic forces. Furthermore, as I2 contracts, its mechanical advantage and the amount of force it can generate both fall. Thus, as I2 protracts the grasper, the forces it can generate decrease, whereas the forces from the hinge antagonizing its protractive force increase. Where the protractive I2 force and the antagonistic hinge force become equal defines a stable mechanical equilibrium point, and this correctly predicts the extent of protraction.
At the end of protraction, activation of I2 shuts off, and the I3 muscle is activated. The I3 muscle generates retractive force, which is balanced by the passive forces in I2. The closed shape of the grasper acts to stretch the I3 muscle, enhancing its ability to exert retractive forces. The point where the I3 muscle force balances the antagonistic passive I2 force defines a different stable equilibrium point, which correctly predicts the extent of retraction.
Aplysia generate rejection responses when they ingest inedible material, or when food that they are attempting to swallow is too large or too tough. Rejection responses can be reliably induced using a uniform polyethylene tube. At the onset of rejection, the grasper closes on the inedible material, and protracts, pushing the material out of the buccal cavity. The grasper then opens and retracts, leaving the inedible material behind. Thus, rejection is characterized by both a strong protraction and a strong retraction. In contrast to biting and swallowing, closing occurs during protraction, and opening occurs during retraction ( Figure 9, top).
The kinetic model predicted that closing at the onset of protraction would lead to mechanical reconfiguration, i.e., the change in shape of the grasper would lead to a change in the positions and length of I2. In particular, the closing of the grasper causes the overall grasper shape to elongate, which in turn stretches the thin I2 muscle that surrounds the grasper (this can be seen schematically by comparing the I2 in Figure 5 and Figure 6). As a consequence, I2's protraction equilibrium point moves more anteriorly. In vitro tests of this prediction demonstrated that closing the halves of the radula significantly enhanced the ability of the I2 muscle to protract the grasper.
Rejection of a polyethylene tube also occurs in two forms: a smaller amplitude form (Type A) and a larger amplitude form (Type B). The major difference between the two types of rejection is that the larger amplitude protraction of a Type B rejection is associated with rotation about the hinge ( Figure 9, middle and bottom).
In both forms of rejection, there must be a delay before the I3 muscle is activated, because otherwise contraction of the jaw lumen will force the halves of the radula closed, and inedible material would be pulled back into the buccal cavity. Because the protraction phase has a larger amplitude in Type B rejections, the delay until the onset of activation of the I3 muscle must be longer. Thus, for both forms of rejection, the hinge initiates the retraction phase, and then after a delay, activation of the I3 muscle completes retraction.
Aplysia generate biting responses when they sense the presence of food (due to chemical and tactile sensations). A bite is an attempt to grasp food. At the onset of a bite, the grasper opens and protracts strongly; before the peak of protraction, the grasper closes, and then weakly retracts. Thus, biting is characterized by a strong protraction and a weak retraction. Opening occurs during protraction, and closing occurs during retraction ( Figure 10, top).
In vivo MRI measurements show that near the peak of protraction, the length of I2 becomes very short. As a consequence, I2's ability to exert force declines, both because of its length/tension property, and because it loses mechanical advantage. At the same time, as the grasper protracts, the hinge is stretched, generating forces that antagonize I2. The kinetic model shows that the mechanical equilibrium point defined by the I2 and hinge forces falls short of the peak protraction of biting. Because the grasper must protract open during a bite, it has a spherical shape, and so does not act to lengthen I2.
What generates the strong protraction observed during biting? The original kinematic model ( Figure 5) suggested that the posterior part of I3, which at the peak of protraction is posterior to the midpoint of the grasper, could act to further protract the grasper. The kinetic model demonstrated that, in this configuration, the forces in I3 and the antagonistic forces in the hinge defined a stable mechanical equilibrium point positioned at the peak protraction of biting. In other words, neural control could exploit the unstable equilibrium point created when the center of I3 is above the center of the grasper, and by pushing the system into the basin of attraction of the anterior equilibrium point, create a strong protraction.
What generates the weak retraction observed during biting? Prior to the peak of protraction, the grasper closes, changing its shape, reducing the mechanical advantage of the I3 muscle. The reduction in the I3 forces and the large antagonistic hinge forces combine to generate a mechanical equilibrium point whose position corresponds to the position of the grasper during the retraction phase of biting ( Figure 10, bottom).
Multifunctionality and Behavioral Transitions
Part of the flexibility of multifunctional systems comes from their ability to rapidly switch among different behaviors. Thus, when Aplysia have strongly protracted the radula in attempt to grasp food (bite), and succeed in grasping it, the immediately succeeding retraction is much stronger, similar to that seen during swallowing. This response is referred to as a bite/swallow.
Similarly, if Aplysia encounter inedible material, such as a polyethylene tube, they will initially swallow it, and then reject it. Before rejection behavior is observed, intermediate behaviors, in which the tube moves in and then out, are often observed, along with corresponding intermediate patterns of neural activity, in which the grasper is closed during both protraction and retraction.
Itay Hurwitz and Abraham Susswein showed that Aplysia oculifera respond to increased mechanical loading by generating slower, larger amplitude swallows. If the load increased very rapidly, animals would cut the food, suggesting a mode in which they clamped the food between the jaws and sharply pulled back with the grasper while tearing the food.
If Aplysia encounter algae growing on a flat substrate, they will switch to a grazing behavior, similar to that of the pond snail Lymnaea, in which they will press the surface of the radula against the substrate and rasp material off of it.
The feeding behavior of Aplysia is not stereotyped, but highly variable, as was shown by Charles C. Horn and colleagues in 2004 using chronic extracellular recording in intact, feeding animals. Vladimir Brezina and colleagues have suggested that cycle-to-cycle variability may enhance feeding efficiency in the face of uncertain and changing mechanical loads.
The ability to rapidly switch among different motor patterns and to generate motor "blends" is similar to that observed in the motor systems of vertebrates. For example, the turtle can use its hindlimbs for walking, swimming, and scratching, and Paul S. G. Stein and colleagues have observed patterns that blend components of these different motor responses.
Implications for Neural Control
The studies of multifunctionality in Aplysia have several implication for neural control.
- First, it is clear that muscles do not have single functions. Rather, groups of muscles work together to form muscular coalitions that generate different behavioral components in different contexts. It is likely that this observation applies beyond Aplysia, and even beyond soft-bodied animals, to all organisms that have excess degrees of freedom in their periphery that can be exploited in multiple ways to generate complex movements.
- Second, the behavioral significance of motor neuron activity depends on biomechanical context. For example, the motor neuron for I4, B8, can act as a closer during Type A swallows, and as a closer and retractor during Type B swallows. Activity in the motor neuron for the hinge, B7, has no effect during Type A swallows, because the hinge is not stretched. Activity in neuron B7 contributes to retraction during Type B swallows, as well as in both variants of rejection, because the hinge is stretched. But B7 also contributes to the protraction and ventral rotation of inedible food during a Type B rejection, because protraction and rotation of a closed grasper affects the movements of the inedible material.
- Third, mechanical reconfiguration leads to another form of interaction between motor neurons: neuromechanical modulation, the ability of one set of motor neurons to alter the effectiveness of another set of motor neurons. Unlike central neuromodulation, in which release of a neurotransmitter changes the responsiveness of a neuron to other inputs, neuromechanical modulation works through the biomechanics of the periphery. Activity in the B8 motor neurons induces the grasper to close; this in turn stretches the I2 muscle, changing the effectiveness of the motor neurons for the I2 muscle (B31, B32, B61, B62).
- Fourth, neuromechanical equilibrium points are crucial for control of feeding movements in Aplysia. The equilibrium points can be clearly defined by a quasi-static analysis. A similar analysis is likely to apply to slow movements in a wide variety of organisms, clarifying the control variables for the nervous system.
- Fifth, the functions of interneurons within the pattern generator for feeding must be understood within a biomechanical context. For example, Hurwitz and colleagues showed that the axons of the key interneurons that initiate all feeding responses, B31 and B32, project to the I2 muscle. Although the somas of B31 and B32 are inexcitable, whenever sufficient synaptic input induces them to generate plateau potentials, the axons are excited. As a consequence of this linkage, protraction is the obligatory first phase of all feeding responses. Another example is provided by the studies of Jian Jing and Klaudisusz R. Weiss. They showed that cerebral buccal interneurons play a crucial role in shifting the timing of closing relative to protraction/retraction, which underlies the shift from ingestive to egestive behaviors. Interneuron B20 activates the B8 motor neuron, inducing I4 to close during protraction, and thus plays a crucial role in generating egestion. Interneuron B40 activates B8 during retraction, and thus plays a crucial role in generating ingestive behaviors.
- Sixth, the timing, duration, and intensity of interneuronal activity may allow a neuron to contribute to different biomechanical functions in different behavioral contexts. The timing of the end of activity in the I2 protractor motor neuron and the onset of activity in the I3 muscle play crucial and distinct roles in biting, swallowing, and rejection ( Figure 11). During biting, I2 is on for long enough that the grasper moves far anterior relative to the posterior of I3, so that activation of I3 prior to the relaxation of the I2 muscle could contribute to protraction. Thus, in biting, I3 is activated before I2 has turned off. During swallowing, protraction is weak, I2 is on for a much shorter period, and I3 can be activated almost immediately after I2 turns off to induce a retraction. During rejection, at the onset of retraction the open grasper is within the lumen of the I1/I3/jaw muscle complex, and so activation of I3 must be delayed. Neurons B4/B5 make widespread inhibitory connections to the motor neurons of I3, so if they are active, they can delay the onset of activity in I3. Eduardo Warman and Chiel showed that, in intact, behaving animals, the B4/B5 neurons are almost inactive during biting (allowing overlap of activity in I2 and I3), moderately active during swallowing (inducing I3 to become active right after I2 shuts off), and intensely active during rejection (delaying I3 activity for some time after I2 has shut off). Hui Ye and colleagues showed that activating or inactivating B4/B5 could increase or decrease the delay to the onset of activity in the I3 motor neurons. Understanding the role of B4/B5 thus requires an understanding of its activity within the biomechanics of the buccal mass.
A Robotic Application
Another way of testing the hypothesis that the function of the I3 muscle could depend on context, inducing retraction when the grasper is posterior to its midpoint, and inducing protraction when the grasper is anterior to its midpoint ( Figure 12), is to construct a physical model. Elizabeth Mangan and colleagues constructed a soft robot gripping device using pneumatic actuators, and showed that if the grasper was at the center of the device, activating the rings surrounding the central grasper from rear to front could protract the grasper, whereas activating the rings from front to rear could retract the grasper. The novel gripper may be useful for robotics applications that require handling soft, irregular materials.
Links to additional information
- Brezina, V., Proekt, A. and Weiss, K. R. 2006. Cycle-to-cycle variability as an optimal behavioral strategy. Neurocomputing 69:1120 - 1124.
- Chiel, H. J., Weiss, K. R., and Kupfermann, I. 1986. An identified histaminergic neuron modulates feeding motor circuitry in Aplysia. J. Neurosci. 6: 2427-2450.
- Cropper, E. C., Kupfermann, I., and Weiss, K. R. 1990. Differential firing patterns of the peptide-containing cholinergic motor neurons B15 and B16 during feeding behavior in Aplysia. Br. Res. 522:176 - 179.
- Drushel, R. F., Neustadter, D. M., Shallenberger, L.L., Crago, P.E., and Chiel, H.J. 1997. The kinematics of swallowing in the buccal mass of Aplysia californica. J. Exp. Biol. 200: 735 - 752.
- Drushel, R. F., Neustadter, D. M., Hurwitz, I., Crago, P. E. and Chiel, H. J. 1998. Kinematic models of the buccal mass of Aplysia. J. Exp.Biol. 201:1563-1583.
- Drushel, R. F., Sutton, G. P., Neustadter, D. M., Mangan, E. V., Adams, B. W., Crago, P. E. and Chiel, H. J. 2002. Radula-centric and odontophore-centric kinematic models of swallowing in Aplysia. J. Exp. Biol. 205:2029-2051.
- Evans, C. G., Rosen, S., Kupermann, I., Weiss, K.R. and Cropper, E.C. 1996. Characterization of a radula opener neuromuscular system in Aplysia. J. Neurophysiol. 76:1267 - 1281.
- Horn, C.C., Zhurov, Y., Orekhova, I. V., Proekt, A., Kupfermann, I., Weiss, K. R. and Brezina, V. 2004. Cycle-to-cycle variability of neuromuscular activity in Aplysia feeding behavior. J. Neurophysiol. 92: 157 - 180.
- Howells, H. H. 1942.The structure and function of the alimentary canal of Aplysia punctata. Q. J. Microsc. Sci. 83: 357- 397.
- Hurwitz, I, and Susswein, A. J. 1992. Adaptation of feeding sequences in Aplysia oculifera to changes in the load and width of food. J. Exp. Biol 166:215 - 235.
- Hurwitz, I, Neustadter, D., Morton, D. W., Chiel, H. J., and Susswein, A.J. 1996. Activity patterns of the B31/B32 pattern initiators innervating the I2 muscle of the buccal mass during normal feeding movements in Aplysia californica. J. Neurophysiol. 75:1309-1326.
- Jing, J. and Weiss, K. R. 2002. Interneuronal basis of the generation of related but distinct motor programs in Aplysia: Implications for current neuronal models of vertebrate intralimb coordination. J. Neurosci. 22:6228 - 6238.
- Kier, W. M. and Smith, K. K. 1985. Tongues, tentacles and trunks: the biomechanics of movement in muscular-hydrostats. Zool. J. Linn. Soc. 83: 307 - 324, 1985.
- Kupfermann, I. 1974. Feeding behavior in Aplysia: A simple system for the study of motivation. Behav. Biol. 10: 1 - 26.
- Mangan, EV, Kingsley, DA, Quinn, RD, Sutton, GP, Mansour, JM and Chiel, HJ. 2005. A biologically inspired gripping device. Industrial Robot 32:49-54.
- Morton, D. W. and Chiel, H. J. 1993. The timing of activity in motor neurons that produce radula movements distinguishes ingestion-like and rejection-like motor patterns in a reduced preparation of Aplysia. J. Comp. Physiol. A. 173:519 - 536.
- Morton, D. W. and Chiel, H. J. 1993. In vivo buccal nerve activity that distinguishes ingestion from rejection can be used to predict behavioral transitions in Aplysia. J. Comp. Physiol. A. 172:17-32.
- Morton, D. W. and Chiel, H. J. 1994. Neural architectures for adaptive behavior. Trends Neurosci. 17:413-420.
- Murphy, A. D. The neuronal basis of feeding in the snail, Helisoma, with comparisons to selected gastropods. 2001. Prog. Neurobiol. 63: 383 - 408.
- Neustadter, D. M., Drushel, R. F., Crago, P. E., Adams, B. W., and Chiel, H. J. 2002. A kinematic model of swallowing in Aplysia californica based on radula/odontophore kinematics and in vivo MRI. J. Exp. Biol. 205:3177-3206.
- Neustadter, D. M., Herman, R. L., Drushel, R. F., Chestek, D. W. and Chiel, H. J. 2007. The kinematics of multifunctionality: Comparisons of biting and swallowing in Aplysia californica. J. Exp. Biol. 210: 238 - 260.
- Novakovic, V.A., Sutton, G.P., Neustadter, D.M., Beer, R.D. and Chiel, H.J. 2006. Mechanical reconfiguration mediates swallowing and rejection in Aplysia californica. J. Comp. Physiol. – A. 192: 857 – 870.
- Orekhova, I.V., Jing, J., Brezina, V., DiCaprio, R. A., Weiss, K. R. and Cropper, E. C. Sonometric measurements of motor-neuron-evoked movements of an internal feeding structure (the radula) in Aplysia. J. Neurophysiol. 86:1057 - 61, 2001.
- Stein, P.S.G. Neuronal control of turtle hindlimb motor rhythms. 2005. J. Comp. Physiol. A 191:213 - 229.
- Elliott, C. J. H. and Susswein, A. J. 2002. Comparative neuroethology of feeding control in molluscs. J. Exp. Biol. 205: 877 - 896.
- Sutton, G. P., Mangan, E. V., Neustadter, D. M., Beer, R. D., Crago, P. E., Chiel, H. J. 2004. Neural control exploits changing mechanical advantage and context dependence to generate different feeding responses in Aplysia. Biol. Cybern. 91: 333-345.
- Warman, E. N. and Chiel, H.J. 1995. A new technique for chronic single extracellular recording in freely behaving animals using pipette electrodes, J. Neurosci. Methods 57:161-169.
- Ye, H., Morton, D. W., and Chiel, H. J. 2006. Neuromechanics of multifunctionality during rejection in Aplysia californica. J. Neurosci. 26: 10743 - 10755.
- Ye, H., Morton, D. W. and Chiel, H. J. 2006. Neuromechanics of coordination during swallowing in Aplysia californica. J. Neurosci 26:1470-1485.
- Yu, S.-N., Crago, P. E., and Chiel, H. J. 1999. Biomechanical properties and a kinetic simulation model of the smooth muscle I2 in the buccal mass of Aplysia. Biol. Cyber. 81:505 - 513.
- Edward Ott (2006) Basin of attraction. Scholarpedia, 1(8):1701.
- James Meiss (2007) Dynamical systems. Scholarpedia, 2(2):1629.
- Eugene M. Izhikevich (2007) Equilibrium. Scholarpedia, 2(10):2014.
- Philip Holmes and Eric T. Shea-Brown (2006) Stability. Scholarpedia, 1(10):1838.