ABSTRACT
The temporal characteristics of statocyst and leg proprioceptive inputs to the uropod motor system were investigated in crayfish using behavioural and electromyographic analyses to elucidate their functional roles in the control of the uropod steering response under natural conditions. When the animal, which was suspended in the air without a footboard, was actively extending its abdomen, prolonged stimulation of the statocysts by body rolling elicited a maintained asymmetrical configuration of the bilateral uropods. Prolonged stimulation of the walking legs by footboard tilting with the animal body held in the upright position elicited a transient uropod response. When the treadmill was tilted while the animal was walking on it in the upright position, the uropods showed the same transient response. However, when the animal body was rolled, together with the treadmill, while the animal was walking on it, the uropods showed a transient response which was reversed in direction compared with that observed during body rolling without a footboard. This transient response was abolished by the removal of the statoliths. The results show that the statocysts and leg proprioceptors exert sustained and transient control effects, respectively, on the uropod motor system during walking. It is also suggested that the uropod response to body rolling during walking is controlled primarily by leg proprioceptor signals which result from statocyst-induced changes in the leg position.
Introduction
The upright body posture of crustaceans is maintained by equilibrium responses which are guided by sensory information from the statocysts, leg proprioceptors and eyes (Kühn, 1914; Alverdes, 1926). The contributions of each of these sense organs to the control of the equilibrium responses displayed by various appendages, including the eyestalks, walking legs, swimmerets and uropods, have been extensively studied (Schöne, 1954, 1971; Davis, 1968, 1971; Neil, 1975a,b,c, 1982; Hisada and Neil, 1985; Newland, 1989). Different sense organs are activated in combination under natural conditions, and how their nerve signals interact with each other to elicit and control the equilibrium responses has also been well studied (Schöne and Neil, 1977; Schöne et al. 1983; Neil, 1985, 1993). In previous studies, however, the functional roles of sense organs were mainly investigated in tethered animals by analyzing their motor output in response to controlled stimulation of a particular sense organ or of various combinations of sense organs, even though under natural conditions the control of body posture by equilibrium responses becomes more important during locomotor behaviour, such as during landing behaviour following tail-flips in the water (Newland and Neil, 1987), than at rest.
The uropod steering behaviour of the crayfish Procambarus clarkii is one of the equilibrium responses elicited by a change in body posture (Yoshino et al. 1980, 1982). This behaviour is controlled by the statocyst input and is expressed as bilaterally asymmetrical configurations of the uropods provided that the roll stimulus is given when the animal, which is fixed by its cephalothorax in the air without a footboard, is actively extending its abdomen (Takahata et al. 1981). When the animal is provided with a footboard and rolled together with it, the response is reversed in direction (Takahata et al. 1984). It has also been reported that the uropods show bilaterally asymmetrical configurations when the walking legs are stimulated by footboard tilting while the animal body is kept upright (Newland, 1989).
Behavioural and electromyographic analyses of crayfish walking freely on a tilted substratum with the abdomen actively extended have revealed, however, that the bilateral uropods showed not only the reversed configuration, which was dominant as expected, but also the normal configuration as well as the bilaterally symmetrical configuration in response to the same stimulus conditions (Takahata et al. 1984). This variability in the uropod response cannot be explained by the results obtained using tethered animals. Since both statocysts and leg proprioceptors are stimulated continuously during walking on a tilted substratum, the temporal characteristics of the sensory inputs from these organs during prolonged stimulation should be analyzed in detail under controlled experimental conditions.
In the present study, we used behavioural observations and electromyogram recordings to investigate the temporal nature of the sensory inputs from the statocysts and leg proprioceptors to the uropod motor system in the tethered condition and during walking driven by a treadmill. The results showed that the statocysts control the uropod response in a sustained way, whereas the leg proprioceptors affect it transiently in tethered animals. When the animal body was tilted together with the treadmill during walking, the uropods showed the reversed response, which soon disappeared as the animal continued to walk in the same stimulus condition. The results suggest that sustained signals from the statocysts to the uropod motor system are overridden by transient signals from the leg proprioceptors to reverse the uropod response, but only during the early phase of prolonged stimulation.
Materials and methods
Experimental animals and apparatus
Experiments were performed on crayfish, Procambarus clarkii Girard, of both sexes ranging from 8 to 12 cm in body length. No significant differences were observed between the results for both sexes. They were obtained commercially and kept in laboratory tanks. They were fed weekly on a diet of raw potato and liver pieces. Animals that were to be used in experiments were given extra food on the day before the experiment to make them as responsive as possible to external stimuli. Results of behavioural observation were obtained from 20 animals for each experimental protocol, except one case in which 11 animals were used. Electromyographic recordings were obtained from 44 animals.
A steel nut was glued to the dorso-anterior region of the cephalothorax of an experimental animal, which was then fixed by the nut to the tiltable holder of the experimental apparatus. Depending on the experiment, the animal was either suspended from the holder without a footboard or placed on a treadmill which was also tiltable. The velocity and direction of the treadmill were controlled by a d.c. motor (Mabuchi Motor, RE-280) with a gearbox. The velocity was adjusted to 1.7–2.1 cm s−1 to be comparable with that of the walking animal. When the sensory input to the uropod motor system during walking was to be studied, the treadmill was tilted, either by itself or together with the animal body, while the motor was running. When the sensory input during the stance phase was to be investigated, the treadmill was tilted with the motor turned off. In this case, the stationary treadmill simply acted as a footboard on which the animal actively performed abdominal extension. The tiltable animal holder and the treadmill were driven by a stepping motor (Nippon Pulse Motor, PJ55-B1) controlled by a personal computer (NEC PC-8801). The rotation axis of the whole apparatus was rigidly coupled to a variable potentiometer, through which the position of the animal’s body was monitored during experiments. Tilting of the animal body and the treadmill could therefore be controlled separately. In order to stimulate the walking legs alone, only the treadmill was tilted while the animal body was kept in the upright (0 ˚) position. When the statocysts were to be stimulated selectively, the treadmill was uncoupled from the apparatus. All experiments were conducted in air. The animal was kept moist by occasionally spraying it with water.
Behavioural observation
In the behavioural study, uropod movements were observed directly or recorded using a video camera and replayed later for analysis. Although we attempted to keep the condition of the experimental animals as uniform as possible by controlling feeding and the environment, there remained a certain amount of variability in the behavioural data. Therefore, the behavioural response of an individual animal is given in this paper as the most dominant response from ten stimulation trials with at least 1 min between trials. The results were analyzed statistically using the G-test and Model II contingency tables (Sokal and Rohlf, 1995). Since the uropod steering movement is elicited only while the animal is performing abdominal extension (Takahata et al. 1981), the nature of the sensory input to the uropod motor system was examined only when the animal was actively extending its abdomen or was engaged in driven walking with the abdomen actively extended. Although abdominal extension could occur spontaneously or in response to various stimuli, such as intense light or pricking of the abdomen, we used leg stimulation (Larimer and Eggleston, 1971) to evoke consistent abdominal extension in this study. The uropods typically showed a symmetrical configuration during abdominal posture movements.
When the treadmill motor was turned on with the experimental animal fixed to the holder and resting on the treadmill, the animal usually began to show walking behaviour characterized by alternating rhythmical leg movements. In rare cases, when the animal was simply inactive or became arrested following the fixing procedure, all legs were passively retracted in response to the treadmill movement. In such cases, the experiment was interrupted and postponed until the animal began to walk actively on the treadmill.
Electromyography
Electromyogram (EMG) recordings were obtained from the slow opener muscles of the uropod (abductor exopodite ventral muscle; Larimer and Kennedy, 1969) on both sides, the abdominal slow extensor muscle (Kennedy et al. 1966; Page, 1982) and the depressor muscle of the third walking leg (Ayers and Davis, 1977a) on either side. Recordings were made using pairs of Teflon-coated stainless-steel wires (127 µ,m in diameter) insulated except at their tips. Signals were amplified (Nihon-Kohden MEG-1100, MEG-2100) and stored on magnetic tapes using a Biologic DTR-1801 DAT data recorder. After experiments, the recorded EMG data were replayed with the body/treadmill position monitor signal and printed using a Gould TA240S EasyGraf recorder.
Results
Sustained control by statocysts
When the statocyst receptors were selectively stimulated by body rolling without a footboard while the animal was actively engaged in abdominal extension, the uropod on the lifted side was spread open, whereas that on the lowered side was closed, as reported previously (Yoshino et al. 1980). This response was extremely reliable in these experimental conditions: all 20 animals showed this response immediately following the body roll stimulus. However, when the animal was first rolled by 60 ˚ and kept in that position for approximately 1 min, the animal became quiescent and the uropods no longer showed any specific configuration. When the animal was again activated to elicit abdominal extension, the uropods showed the same, bilaterally asymmetrical configuration as before (black bar in Fig. 1A).
In the experiment illustrated in Fig. 2, an animal with EMG electrodes inserted was suspended without a footboard. When the animal was rolled while it was engaged in abdominal extension (Fig. 2A), abductor muscle activity increased on the lifted side and decreased on the opposite side. This reciprocal activation of bilaterally homologous muscles has been reported previously (Takahata et al. 1981), but how the prolonged stimulation of statocysts affected the uropod response was not understood. When the animal was kept in the tilted position and became quiescent, the muscle activity level became low on both sides. When abdominal extension was again induced in this tilted position, the reciprocal activity pattern reappeared (Fig. 2C). This pattern was identical to that observed upon tilting from the upright position (Fig. 2A).
Transient control by leg proprioceptors
When the footboard was tilted sideways by 15 ˚ in either direction along the longitudinal axis of the animal body, while the animal was kept in the upright position, 19 out of 20 animals tested responded by opening the uropod on the lifted side and closing the opposite uropod, as reported previously (Newland, 1989). One animal did not show any distinct asymmetrical uropod configuration. When the footboard was tilted by 15 ˚ and kept in the tilted position for about 1 min while the animal’s body was maintained upright (quiescent animals activated to elicit abdominal extension), seven out of 20 animals were found to show the same asymmetrical configuration of bilateral uropods as before (black bar in Fig. 1B), whereas the remaining 13 animals showed no distinct asymmetrical configuration (hatched bar in Fig. 1B).
EMG recordings revealed that the bilaterally reciprocal muscle activities elicited by the footboard tilt became unclear when the animal was kept in the tilted position for 1 min, although the abdominal posture system remained active (Fig. 2B). When the crayfish was returned to the original horizontal position, the reversed reciprocal pattern of bilateral uropod muscle activities was always observed. When the footboard was kept in a tilted position and the abdominal posture system was then activated, no reciprocal activation pattern of bilateral uropod muscles was observed, even when the animal was activated to elicit abdominal extension (Fig. 2D). These results indicate that sensory signals from the leg proprioceptors do not exert sustained control over the uropod motor system, but activate it dynamically following displacement of the leg position.
Effects of leg proprioceptive input during walking
When the animal was kept upright and engaged in forward walking driven by the treadmill, it responded to treadmill tilting by opening the uropod on the lifted side and closing the opposite uropod in 17 out of 20 cases (black bar in Fig. 1C). However, when the animal was forced to walk on the tilted treadmill for 1 min, the uropod configuration showed more variability than that observed immediately after the tilt stimulus (P<0.05): eight out of 20 animals tended to open the uropod on the lifted side and close the opposite uropod (black bar in Fig. 1D). Another five animals showed the reversed configuration: the uropod on the lifted side tended to be closed, whereas the opposite uropod tended to be opened (Fig. 1D, white bar). The remaining seven animals did not show any distinct tendency (hatched bar).
In response to treadmill tilting during driven walking, the activity of the abductor muscle on the lifted side increased while that on the opposite side decreased (Fig. 3). This is the same as the pattern observed when the footboard is tilted while the animal is in the stance phase (Newland, 1989). The activity of the leg depressor muscle was also affected by treadmill tilting: it was reduced on the lowered side and enhanced on the raised side. Fig. 3 also shows that the reciprocal pattern of bilateral uropod muscle activities in response to prolonged tilting of the treadmill during walking was more persistent than that observed in response to prolonged tilting of the footboard (Fig. 2B). This difference could be accounted for, at least partly, by the fact that the right and left leg proprioceptor systems are activated alternately by the rhythmical movement of the legs during walking so that they can exert a relatively sustained effect on the uropod motor system. However, the result of the behavioural experiment (Fig. 1D) indicates that such alternate activation of leg proprioceptors during walking cannot completely override the phasic nature of leg inputs to the uropod motor system.
Interactions between the statocyst and leg inputs during walking
In response to rolling of the animal body together with the footboard while the animal was actively extending its abdomen, the uropod on the lifted side was closed whereas that on the lowered side was spread open in 10 out of 11 animals tested, as reported previously (Takahata et al. 1984). One animal did not show any distinct asymmetrical configuration. Since the animal was in the stance phase, no systematic stimulation of leg proprioceptors was involved in this experimental condition. When the crayfish and the treadmill were tilted simultaneously while the animal was engaged in driven walking, the uropod on the lifted side was closed whereas that on the opposite side was spread open in 19 out of 20 animals (white bar in Fig. 1E). One crayfish did not show any distinct asymmetrical configuration (hatched bar in Fig. 1E). When both the animal body and the treadmill were kept in a tilted position before the treadmill was turned on, to induce forward walking for 1 min, the uropod configuration showed significantly more variability than immediately after the tilt stimulus (P<0.025): nine out of 20 animals tended to open the uropod on the lifted side and to close the opposite uropod (black bar in Fig. 1F). Another three animals showed the reversed configuration (white bar in Fig. 1F) and the remaining eight animals did not show any distinct tendency (hatched bar in Fig. 1F).
EMG recordings from a crayfish tilted together with the treadmill during walking (Fig. 4) confirmed that the abductor muscles on both sides were reciprocally activated to produce the uropod steering movement observed. This was in the opposite direction to that observed in response to body rolling without the footboard (Fig. 2). This reversed reciprocal pattern of bilateral uropod muscle activities was no longer elicited by body rolling after the statoliths had been removed on both sides (hatched bar in Fig. 1G). The activity of the leg depressor muscle on the lowered side was also enhanced, whereas that on the opposite side was reduced. This pattern of leg muscle activity continued as long as the body roll was maintained (Fig. 4), but it remains to be determined whether the walking system also receives sustained control signals from the statocysts, as does the uropod system.
Discussion
In the present study, we have investigated the sensory control mechanisms of equilibrium responses during free walking by examining the sensory effects on the activity of the uropod motor system during driven walking on a treadmill. It has been shown that driven walking on a treadmill differs from free walking in several respects, including the temporal organization of the stepping cycle (MacMillan, 1975; Ayers and Davis, 1977a; Clarac and Chasserat, 1983) and the forces developed by the legs in a series of consecutive steps (Clarac and Cruse, 1982). However, detailed analyses of these differences have revealed that driven walking is controlled by a neuronal mechanism that is fundamentally identical to that underlying free walking (Ayers and Davis, 1977b; Ayers and Clarac, 1978; Chasserat and Clarac, 1980; Cruse et al. 1983; Cruse, 1983, 1990).
It is also noteworthy that crayfish showed a strong tendency to extend the abdomen when driven walking was initiated by the treadmill, as reported for free walking (Ward, 1879; Takahata et al. 1984). This tendency of the animal was advantageous for the present study since the behavioural output of the uropod motor system in response to a positional change could be reliably observed only when the tilt stimulus was accompanied by extension of the abdomen (Takahata et al. 1981, 1984).
Statocyst and leg proprioceptive inputs to the uropod motor system
It has been reported that uropod steering behaviour is under the strict control of sensory signals from the statocysts (Yoshino et al. 1980) and leg proprioceptors (Newland, 1989) as well as central signals from the abdominal posture and walking systems (Takahata et al. 1984). However, the temporal characteristics of the sensory signals responsible for controlling uropod motor output have not been investigated and are important to the understanding of control mechanisms during locomotor behaviour under natural conditions. In the present study, we have addressed this problem and showed that the statocyst sensory input exerts tonic control over the uropod motor system, whereas the leg proprioceptive input affects it in a transient way (Figs 1, 2). The finding that the uropods resumed the appropriate steering configuration when the abdominal system was reactivated after the animal had become quiescent during maintained body roll (Fig. 2A,C) is consistent with previous results suggesting that a maintained body posture is detected by tonic-type statocyst receptors and is represented in the spiking activity of a set of interneurones descending to the thoracic and abdominal ganglia (Takahata and Hisada, 1979, 1982), although the synaptic input to the uropod motor system remains subthreshold when the animal is in the quiescent state (Takahata and Murayama, 1992). The results of this study thus confirm our previous conclusion that sensory control signals from the statocyst descend continuously to the uropod motor system irrespective of whether the animal is actively engaged in abdominal posture movements, but that they are effectively transmitted to the motor neurones only when the animal is active (Takahata and Hisada, 1986).
The major leg proprioceptive input to the uropod motor system for the control of steering behaviour comes from the coxo-basipodite joint (Newland, 1989). At present, the type of information that is transmitted from the joint proprioceptor to the central nervous system is unknown. However, the presence of position-and movement-sensitive receptors in the pro-dactylopodite, coxo-basipodite, mero-carpopodite and carpo-propodite joints in the crab Carcinus maenas (Burke, 1954; Wiersma and Boettiger, 1959; Bush, 1965a,b; Mill, 1976) suggests that a similar receptor system is also present in the coxo-basipodite joint of the crayfish. The present finding, that the bilateral uropods no longer showed the steering configuration when the abdominal system was reactivated after the animal had become quiescent during maintained footboard tilt (Fig. 2D), suggests the possibility that only dynamic information on leg joint displacement is effectively transmitted to the uropod motor system.
The results obtained in this study therefore suggest that the different control effects of the statocyst and the leg proprioceptive inputs on the uropod steering response are due not to a lack of either tonic or phasic sensory signals in each receptor system but to different sensori-motor functional connections, although the anatomical bases of these selective connections remain to be investigated. The observations that the leg proprioceptors can exert both a long-lasting tonic effect on the oculomotor system and a phasic effect on the antennal motor system of the spiny lobster (Schöne et al. 1976; Neil and Schöne, 1979) also support this possibility.
Control mechanisms of the uropod steering response during walking
When the animal body is rolled together with the treadmill while the animal is walking, the uropods show a bilaterally asymmetrical configuration which is reversed in direction compared with that observed during body rolling without a footboard (Figs 1A,E, 4). It is not known how the presence of a footboard reverses the direction of uropod movement. However, the results of this study seem to provide a clue for further analysis of this mechanism. The leg proprioceptive input during walking is phasic in nature (Fig. 1D), as it is during the stance phase (Fig. 1B), although the phasic nature is less obvious (Fig. 3). When the walking animal is rolled together with the treadmill, the uropods adopt a steering configuration (Fig. 4) which tends to disappear during maintained stimulation (Fig. 1F). Since tonic sensory signals should still descend to the uropod motor system during this maintained stimulation, as seen in Fig. 2C, the absence of the reversed uropod configuration after movements in response to phasic input from leg proprioceptors (Fig. 1F) strongly suggests that the sensory signals from the statocysts and leg proprioceptors summate in a manner that places much greater weight on signals from the leg proprioceptors. An alternative possibility would be that statocyst signals are suppressed by those from leg proprioceptors (Fig. 5).
The site of interaction of the sensory signals from the statocysts and leg proprioceptors also remains unknown. Newland (1989) suggested that the signals enter the terminal abdominal ganglion via separate pathways, although the descending interneurones which selectively carry leg proprioceptive information have not been identified. The presence of independent parallel pathways for the leg and statocyst inputs would indicate that their interaction takes place in the terminal abdominal ganglion. However, the presence of multiple gates in the descending statocyst pathway, including one in the terminal ganglion (Takahata and Hisada, 1986; Takahata and Murayama, 1992), suggests that this interaction may also occur outside the terminal ganglion. Further study is needed to identify the site of interaction and to investigate its synaptic mechanisms at the cellular level.
In an earlier study, behavioural observations of crayfish walking freely on a tilted substratum revealed that the uropods sometimes showed not only the reversed configuration, which was dominant, but also the normal configuration as well as the bilaterally symmetrical configuration (Takahata et al. 1984). This variability in the uropod response to the same tilt stimulus during walking can be accounted for, at least partly, by the phasic nature of the leg input: when the animal walks continuously on a tilted substratum, the leg input becomes weaker, making the uropod response more ambiguous. The present study demonstrates experimentally that the uropod configuration shows significant variability when the animal continues to walk on a tilted substratum (Fig. 1D), reflecting the phasic nature of the leg proprioceptive input (Figs 1B, 2D).
ACKNOWLEDGEMENTS
This work was supported in part by Grants-in-Aid (05640758, 07640894) from MESC Japan to M.T.