Opposing effects of dopamine on agonistic behaviour in crayfish.

The effects of dopamine on the agonistic behaviour of crayfish were analysed. When dopamine concentrations of 1 µmol l-1 were injected into large crayfish, individuals were beaten by smaller opponents, despite their physical advantage. Injection of 10 µmol l-1 dopamine into small animals increased their rate of winning against larger opponents. Injection of a D1 receptor antagonist prohibited the onset of a 'loser' effect in subordinate animals, suggesting that the inhibitory effect of dopamine on larger animals is mediated by D1 receptors. Similarly, injection of a D2 receptor antagonist prohibited the onset of a 'winner' effect in dominant animals, suggesting that the facilitating effect of dopamine on small animals is mediated by D2 receptors. Since the inhibitory effect of 1 µmol l-1 dopamine was similar to that seen with 1 µmol l-1 octopamine and the facilitating effect of 10 µmol l-1 dopamine was similar to that of 1 µmol l-1 serotonin, functional interactions among dopamine, octopamine and serotonin were analyzed by co-injection of amines with their receptor antagonists in various combinations. The inhibitory effect of 1 µmol l-1 dopamine disappeared when administered with D1 receptor antagonist, but remained when combined with octopamine receptor antagonist. Octopamine effects disappeared when administered with either D1 receptor antagonist or octopamine receptor antagonist, suggesting that the dopamine system is downstream of octopamine. The facilitating effect of 10 µmol l-1 dopamine disappeared when combined with serotonin 5HT1 receptor antagonist or D2 receptor antagonist. Serotonin effects also disappeared when combined with D2 receptor antagonist, suggesting that dopamine and serotonin activate each other through parallel pathways.

Anxiety induces long-term memory forgetting in the crayfish.

When two male crayfish encounter, agonistic bouts are initiated and a winner-loser relationship is established. Larger animals are more likely to win with their physical advantage, but they are frequently beaten by small dominant animals with previous winning experience. This winner effect remains for several days. In mammals, anxiety impairs learning and induces memory forgetting. In this study, dominant crayfish were exposed to electrical shocks two days after their first win, after which they were paired with large or small naive opponents the following day. Our results showed that electrical shock-applied dominant animals were beaten by large naive opponents, but overcame small naive opponents, suggesting that electrical shocks cause animals to forget their previous winner effect. Electrical shocks appeared to elicit serotonin-mediated anxiety since electrical shocks had no effect on mianserin-injected dominant animals. A 0.5 µM serotonin injection induced a caused anxiety-like reaction, while a 1.0 µM serotonin injection-induced no changes in posture and walking activity. For pairings between dominant and naive animals 1 day after serotonin injection, 0.5 µM serotonin caused similar forgetting of the winner effect, but 1.0 µM serotonin had no effect. Serotonin of low concentrations mediated anxiety and stimulated forgetting of the winner's memory.

Enhancement of synaptic responses in ascending interneurones following acquisition of social dominance in crayfish.

When crayfish have attained dominant status after agonistic bouts, their avoidance reaction to mechanical stimulation of the tailfan changes from a dart to a turn response. Ascending interneurones originating in the terminal ganglion receive sensory inputs from the tailfan and they affect spike activity of both uropod and abdominal postural motor neurones, which coordinates the uropod and abdominal postural movements. Despite the varying output effects of ascending interneurones, the synaptic responses of all interneurones to sensory stimulation were enhanced when they acquired a dominant state. The number of spikes increased as did a sustained membrane depolarizations. Regardless of social status, the output effects on the uropod motor neurones of all interneurones except VE-1 remained unchanged. VE-1 mainly inhibited the uropod opener motor neurones in naive animals, but tended to excite them in dominant animals. Synaptic enhancement of the sensory response of ascending interneurones was also observed in naive animals treated with bath-applied serotonin. However, subordinate animals or naive animals treated with octopamine had no noticeable effect on the synaptic response of their ascending interneurones to sensory stimulation. Thus, enhancement of the synaptic response is a specific neural event that occurs when crayfish attain social dominance and it is mediated by serotonin.

Aversive operant conditioning alters the phototactic orientation of the marbled crayfish.

Aversive learning was applied to affect the phototactic behaviour of the marbled crayfish. Animals initially showed negative phototaxis to white light and positive taxis to blue light. Using an aversive learning paradigm, we investigated the plasticity of innate behaviour following operant conditioning. The initial rate of choosing a blue-lit exit was analysed by a dual choice experiment between blue-lit and white-lit exits in pre-test conditions. During training, electrical shocks were applied to the animals when they oriented to the blue-lit exit. Memory tests were given to analyse the orientation rate to the blue-lit exit in trials 1 and 24 h after training and these rates were compared with the pre-test. In general, animals avoided the blue-lit exit in the memory tests. When training was carried out three times, the long-term memory was retained for at least 48 h, although a single bout of training was also enough to form a long-term memory. Cooling animals at 4°C or injection of cycloheximide immediately after training altered the formation of long-term memory, but had no effect on short-term memory formation. Administration of the adenylate cyclase inhibitor SQ22536, the PKA inhibitor H89 or the CREB inhibitor KG-501 immediately after training also blocked the formation of long-term memory, but had no effect on short-term memory formation. Thus, our pharmacological behavioural analyses showed that new protein synthesis was necessary to form long-term memories and that the cAMP/PKA/CREB pathway is the main signal cascade for long-term memory formation in the marbled crayfish.

The effects of shelter quality and prior residence on marmorkrebs (marbled crayfish).

Many animals fight over a limited valuable resource. In marmorkrebs (marbled crayfish), large animals usually defeat small opponents but they are frequently beaten by small opponents that are shelter owners. A prior residence effect of marbled crayfish was analysed quantitatively in the present study. More than 2 h of residency in a shelter was sufficient for small owners to defeat large intruders. Small animals that stayed in a shelter for 24 h still tended to win following removal of the shelter 10 min before pairing with large intruders, but 2 h residents were occasionally beaten by large intruders without the support of shelters during pairings. The prior residence effect thus developed depending on the duration of residency. To clarify whether the strength of the prior residence effect was affected by the quality of a shelter, large and small owners with different combinations of high- and low-quality shelters were paired. When both large and small owners possessed a high-quality shelter, the frequency of agonistic bouts was reduced. Even if agonistic bouts occurred, the win frequency of small owners was almost equal to that of large owners. Thus, the residence effect on small owners was sufficiently strong to overcome the physical disadvantage of small animals to large opponents. By contrast, small owners of low--quality shelters were frequently beaten by large owners with the shelters of same or better quality. We conclude that the outcome of fights over the resource shelter is highly dependent on both the perception of shelter quality and body size differences.

Enhancement of habituation during escape swimming in starved crayfish.

Feeding is important to supply the immediate energy needs of animals and starved animals must expend energy in attempting to acquire foods irrespective of the danger of predation risk. Crayfish escape from attack of predators by tailflipping and in response to rostral stimuli crayfish show backward escape swimming following an initial rapid flexion of the abdomen. Since the tailflip is an energetically costly behaviour, the occurrence of a tailflip diminishes if a stimulus is repeatedly applied through habituation. In this study, we have compared the process of this habituation between fed and starved crayfish. We found that in starved animals habituation was enhanced compared to fed animals. The presence of food in the experimental tanks further enhanced habituation of starved animals. Starved crayfish thus showed trade-offs between energy saving and predation risk.

Nitric oxide-mediated intersegmental modulation of cycle frequency in the crayfish swimmeret system.

Crayfish swimmerets are paired appendages located on the ventral side of each abdominal segment that show rhythmic beating during forward swimming produced by central pattern generators in most abdominal segments. For animals with multiple body segments and limbs, intersegmental coordination of central pattern generators in each segment is crucial for the production of effective movements. Here we develop a novel pharmacological approach to analyse intersegmental modulation of swimmeret rhythm by selectively elevating nitric oxide levels and reducing them with pharmacological agents, in specific ganglia. Bath application of L-arginine, the substrate NO synthesis, increased the cyclical spike responses of the power-stroke motor neurons. By contrast the NOS inhibitor, L-NAME decreased them. To determine the role of the different local centres in producing and controlling the swimmeret rhythm, these two drugs were applied locally to two separate ganglia following bath application of carbachol. Results revealed that there was both ascending and descending intersegmental modulation of cycle frequency of the swimmeret rhythm in the abdominal ganglia and that synchrony of cyclical activity between segments of segments was maintained. We also found that there were gradients in the strength effectiveness in modulation, that ascending modulation of the swimmeret rhythm was stronger than descending modulation.

Tyraminergic modulation of agonistic outcomes in crayfish.

Octopamine, a biogenic amine, modulates various behaviors, ranging from locomotion and aggression to learning and memory in invertebrates. Several studies recently demonstrated that tyramine, the biological precursor of octopamine, also affects behaviors independent of octopamine. Here we investigated the involvement of tyramine in agonistic interaction of the male crayfish Procambarus clarkii. When male crayfish fight, larger animals (3-7% difference in body length) are more likely to win. By contrast, direct injection of tyramine or octopamine counteracted the physical advantage of larger animals. Tyramine or octopamine-injected naive large animals were mostly beaten by untreated smaller naive animals. This pharmacological effect was similar to the loser effect in which subordinate larger animals are frequently beaten by smaller animals. Furthermore, loser effects were partly eliminated by either injection of epinastine, an octopamine blocker, or yohimbine, a tyramine blocker, and significantly diminished by injection of a mixture of both blockers. We also observed that tyramine levels in the subesophageal ganglion were remarkably increased in subordinate crayfish after losing a fight. These results suggest that tyramine modulates aggressive levels of crayfish and contributes to the loser effect in parallel with octopamine.

Cyclic AMP-regulated opposing and parallel effects of serotonin and dopamine on phototaxis in the Marmorkrebs (marbled crayfish).

Phototactic behaviours are observed from prokaryotes to amphibians and are a basic form of orientation. We showed that the marbled crayfish displays phototaxis in which the behavioural response reversed from negative to positive depending on external light conditions. Animals reared in a 12-L/12-D light cycle showed negative phototaxis during daytime and positive phototaxis during night-time. Animals reared under constant light conditioning showed negative phototaxis during day- and night-time, while animals reared under constant dark conditioning showed positive phototaxis during day- and night-time. Injection of serotonin leads to a reversal of negative to positive phototaxis in both light/dark-reared and light/light-reared animals while injection of dopamine induced reversed negative phototaxis in dark/dark-reared animals. Four hours of dark adaptation were enough for light/dark-reared animals to reverse phototaxis from negative to positive. Injection of a serotonin 5HT1 receptor antagonist blocked the reverse phototaxis while serotonin 5HT2 receptor antagonists had no effects. Similarly, dark/dark-reared animals reversed to showing negative phototaxis after 4 h of light adaptation. Injection of a dopamine DA1 receptor antagonist blocked this reverse phototaxis, while dopamine DA2 receptor antagonists had no effects. Injection of a cAMP analogue into light/dark-reared animals blocked reverse phototaxis after dark adaptation, while adenylate cyclase inhibitor in dark/dark-reared animals blocked reverse phototaxis after light adaptation. These results strongly suggest that serotonin mediates positive phototaxis owing to decreased cAMP levels, while dopamine-mediated negative phototaxis occurs due to increased cAMP levels. Supporting this, the ratio of serotonin to dopamine in the brain was much higher in dark/dark-reared than light/dark-reared animals.

Role of cAMP signalling in winner and loser effects in crayfish agonistic encounters.

For territorial animals, establishment of status-dependent dominance order is essential to maintain social stability. In agonistic encounters of the crayfish Procambarus clarkii, a difference of body length of 3-7% is enough for larger animals to become dominant. Despite a physical disadvantage, small winners of the first pairings were more likely to win subsequent conflicts with larger inexperienced animals. In contrast, the losers of the first pairings rarely won subsequent conflicts with smaller naive animals. Such experiences of previous winning or losing affected agonistic outcomes for a long period. The winner effects lasted more than 2 weeks and the loser effect lasted about 10 days. Injection of 5HT1 receptor antagonist into the dominant animals 15-30 min after establishment of dominance order blocked the formation of the winner effects. In contrast, injection of adrenergic-like octopamine receptor antagonist into subordinate animals blocked the formation of the loser. 5HT1 receptors are negatively coupled to adenylyl cyclase and adrenergic-like octopamine receptors are positively coupled. Consistent with this, dominant animals failed to show the winner effect when injected with pCPT-cAMP, a cAMP analogue, and subordinate animals failed to show a loser effect when injected with adenylyl cyclase inhibitor SQ 22536. These results suggest that an increase and decrease of cAMP concentration is essential in mediating loser and winner effects, respectively. Furthermore, formation of the loser effect was blocked by injection of protein kinase A (PKA) inhibitor H89, suggesting long-term memory of the loser effect is dependent on the cAMP-PKA signalling pathway.

Habituation of Backward Escape Swimming in the Marbled Crayfish.

In the present study, we performed behavioral analyses of the habituation of backward escape swimming in the marbled crayfish, Procambarus fallax. Application of rapid mechanical stimulation to the rostrum elicited backward swimming following rapid abdominal flexion of crayfish. Response latency was very short-tens of msec-suggesting that backward swimming is mediated by MG neurons. When stimulation was repeated with 10 sec interstimulus intervals the MG-like tailflip did not occur, as the animals showed habituation. Retention of habituation was rather short, with most animals recovering from habituation within 10 min. Previous experience of habituation was remembered and animals habituated faster during a second series of experiments with similar repetitive stimuli. About half the number of stimulus trials was necessary to habituate in the second test compared to the first test. This promotion of habituation was observed in animals with delay periods of rest within 60 min following the first habituation. After 90 min of rest from the first habitation, animals showed a similar time course for the second habituation. With five stimuli at 15 min interval during 90 min of the rest, trained animals showed rapid habituation, indicating reinforcement of the memory of previous experiments. Crayfish also showed dishabituation when mechanical stimulation was applied to the tail following habituation.

Serotonergic modulation of social status-dependent behavioural plasticity of the crayfish avoidance reaction.

Small crayfish usually showed escape-like dart responses to mechanical stimulation of the tailfan. Following agonistic bouts with conspecifics, dominant crayfish showed a defensive-like turn response to the same sensory stimulus. During the dart response, both uropods closed and animals walked forwards with the abdomen extended, while during the turn response the uropod on the stimulated side opened and animals turned towards the stimulus source with the abdomen frequently flexed. Using an isolated nerve cord preparation, we found that the spike activities of both the abdominal postural and uropod motor neurones in response to sensory stimulation of the exopodite reversed in dominant animals. In naive and subordinate animals, extensor motor neurones were excited and antagonistic flexor motor neurones inhibited in response to sensory stimulation. Furthermore, the spike frequency of uropod closer motor neurones increased while that of the antagonistic opener motor neurones decreased. By contrast, in more than half of the dominant animals, flexor and opener motor neurones showed excitatory responses while extensor and closer motor neurones showed inhibitory responses to the same sensory stimulation. This reverse of activity of the abdominal postural and uropod motor neurones was also observed when serotonin of more than 12.5 µM in concentration was applied.

Excitatory connections of nonspiking interneurones in the terminal abdominal ganglion of the crayfish.

The output effects of the nonspiking interneurones in the crayfish terminal abdominal ganglion upon the uropod motor neurones were characterized using simultaneous intracellular recordings. Inhibitory interactions from nonspiking interneurones to the uropod motor neurones were one-way and chemically mediated. The depolarization of the motor neurones with current injection increased the amplitude of the nonspiking interneurone-mediated hyperpolarization, while hyperpolarization of the motor neurone decreased it. By contrast, excitatory interactions from the nonspiking interneurones to the motor neurones were not mediated via chemical synaptic transmissions. These excitatory connections with the slow motor neurones were one-way while connections with fast motor neurones were bidirectional. Nonspiking interneurone-mediated membrane depolarization of the motor neurones was not affected by the passage of hyperpolarizing current. Each motor neurone spike elicited a time-locked EPSP in the nonspiking interneurones with very short delay (0.2 ms) that suggested electrical coupling between nonspiking interneurones and motor neurones. Nonspiking interneurones directly control the organization of slow motor neurone activity, while they appear to regulate the background activity of the fast motor neurones. A single nonspiking interneurone is possible to inhibit some inter and/or motor neurones via direct chemical synapses and simultaneously excite other neurones via electrical synapses.

Habituation of LG-mediated tailflip in the crayfish.

Crayfish escape from threatening stimuli by tailflipping. If a stimulus is applied to the rear, crayfish escape up and forwards in a summersault maneuver that is mediated by the activation of lateral giant (LG) interneurons. The occurrence probability of LG-mediated tailflip, however, diminishes and habituates if a stimulus is repeatedly applied. Since crayfish have a relatively simple CNS with many identifiable neurons, crayfish represent a good animal to analyze the cellular basis of habituation. A reduction in the amplitude of the EPSP in the LGs, caused by direct chemical synaptic connection from sensory afferents by repetitive stimulations, is essential to bring about an inactivation of the LGs. The spike response of the LGs recovers within several minutes of habituation, but the LGs subsequently fail to spike when an additional stimulus is applied after specific periods following habituation. These results indicate that a decline in synaptic efficacy from the mechanosensory afferents recovers readily after a short delay, but then the excitability of the LGs themselves decreases. Furthermore, the processes underlying habituation are modulated depending on a social status. When two crayfish encounter each other, a winner-loser relationship is established. With a short interstimulus interval of 5 s, the rate of habituation of the LG in both socially dominant and subordinate crayfish becomes lower than in socially isolated animals. Serotonin and octopamine affect this social status-dependent modulation of habituation by means of activation of downstream second messenger system of cAMP and IP3 cascades, respectively.

Nitric oxide modulates a swimmeret beating rhythm in the crayfish.

The modulatory effects of nitric oxide (NO) and cAMP on the rhythmic beating activity of the swimmeret motor neurones in the crayfish were examined. Swimmerets are paired appendages located on the ventral side of each abdominal segment that show rhythmic beating activity during forward swimming, postural righting behaviour and egg ventilation in gravid females. In isolated abdominal nerve cord preparations, swimmeret motor neurones are usually silent or show a continuous low-frequency spiking activity. Application of carbachol, a cholinergic agonist, elicited rhythmic bursts of motor neurone spikes. The co-application of L-arginine, the substrate for NO synthesis with carbachol increased the burst frequency of the motor neurones. The co-application of the NO donor SNAP with carbachol also increased the burst frequency of the motor neurones. By contrast, co-application of a NOS inhibitor, L-NAME, with carbachol decreased beating frequency of the motor neurones. These results indicate that NO may act as a neuromodulator to facilitate swimmeret beating activity. The facilitatory effect of L-arginine was cancelled by co-application of the soluble guanylate cyclase (sGC) inhibitor ODQ suggesting that NO acts by activating sGC to promote the production of cGMP. Application of L-arginine alone or membrane-permeable cGMP analogue 8-Br-cGMP alone did not elicit rhythmic activity of motor neurones, but co-application of 8-Br-cGMP with carbachol increased bursting frequency of the motor neurones. Furthermore, application of the membrane-permeable cAMP analogue CPT-cAMP alone produced rhythmic bursting of swimmeret motor neurones, and the bursting frequency elicited by CPT-cAMP was increased by co-application with L-arginine. Co-application of the adenylate cyclase inhibitor SQ22536 ceased rhythmic bursts of motor neurone spikes elicited by carbachol. These results suggest that a cAMP system enables the rhythmic bursts of motor neurone spikes and that a NO-cGMP signaling pathway increases cAMP activity to facilitate swimmeret beating.

Aminergic control of social status in crayfish agonistic encounters.

Using pairings of male crayfish Procambarus clarkii with a 3-7% difference in size, we confirmed that physically larger crayfish were more likely to win encounters (winning probability of over 80%). Despite a physical disadvantage, small winners of the first pairings were more likely to win their subsequent conflicts with larger naive animals (winning probability was about 70%). By contrast, the losers of the first pairings rarely won their subsequent conflicts with smaller naive animals (winning probability of 6%). These winner and loser effects were mimicked by injection of serotonin and octopamine. Serotonin-injected naive small crayfish were more likely to win in pairings with untreated larger naive crayfish (winning probability of over 60%), while octopamine-injected naive large animals were beaten by untreated smaller naive animals (winning probability of 20%). Furthermore, the winner effects of dominant crayfish were cancelled by the injection of mianserin, an antagonist of serotonin receptors and were reinforced by the injection of fluoxetin, serotonin reuptake inhibitor, just after the establishment of social order of the first pairings. Injection of octopamine channel blockers, phentolamine and epinastine, by contrast, cancelled the loser effects. These results strongly suggested that serotonin and octopamine were responsible for winner and loser effects, respectively.

Social status-dependent modulation of LG-flip habituation in the crayfish.

Strong stimuli applied to the tailfan of the crayfish Procambarus clarkii evoked lateral giant interneurone (LG)-mediated tailflips. When the sensory stimulus was applied repeatedly, the response of the LG habituated until it failed to give rise to a spike. We found that this LG-flip habituation was dependent on social status. With a short interstimulus interval of 5 s, the rate of habituation of the LG in both socially dominant and subordinate crayfish was lower than that in socially isolated animals. By contrast, with a long interstimulus interval of 60 s, the rate of habituation of subordinate animals was lower than that of both socially isolated and dominant animals. The excitability of the LGs following habituation was also dependent on social status. Following habituation, the spike response of LGs recovered within several minutes; however, they showed significant depression with a decrease in excitability. With a 5 or 60 s interstimulus interval, subordinate animals showed longer delays of depression compared with dominant animals. A decrease in the rate of habituation and a delay of depression in subordinate crayfish would be advantageous for maintaining an active escape response to evade repeated attacks of dominant animals and a reduced learning ability to adapt to social status.

Interlocking of chelae is a key factor for dominance hierarchy formation in crayfish.

We characterized the role of chelae during agonistic encounters of the crayfish Procambarus clarkii. Physical asymmetries in body length, body mass and chelae size were directly related to dominance hierarchy formation. More than 80% of winning crayfish had longer body and chelae lengths, and winners were usually heavier in body mass, even if their differences were less than 3% compared with losing opponents. In mismatched pairings, large crayfish with short chelae were beaten by small crayfish with large chelae. Three physical elements of body length, body mass and chelae length equally affected the outcome of agonistic bouts. Chelae restriction, in which chelae were tightly closed using rubber bands, affected the outcome of agonistic bouts between large and small crayfish. Before chelae restriction, large crayfish won in all pairings. Following chelae restriction for at least 30 days from the first encounters, the winning rate of large crayfish that were previous winners decreased significantly in the second encounters against the same opponents that were previous losers. The handicap of chelae restriction significantly prolonged the time to formation of the winner-loser relationship. Individual fights escalated during agonistic bouts between large crayfish with one chela restricted and small crayfish with intact chelae, whereas the number of fights increased but the duration of individual fights did not increase in large crayfish with both chelae restricted. Furthermore, when the chelae of both large and small crayfish were disabled, the dominance order was frequently not formed during 30 min of agonistic bouts. Preventing chelae from interlocking prevented escalation of agonistic bouts. We show that interlocking of chelae acted as a key factor for the formation of dominance hierarchy.

IP3-mediated octopamine-induced synaptic enhancement of crayfish LG neurons.

The biogenic amines, octopamine and serotonin, modulate the synaptic activity of the lateral giant interneuron (LG) circuitry of the crayfish escape behavior. Bath application of both octopamine and serotonin enhances the synaptic responses of LG to sensory stimulation. We have shown previously (Araki et al. J Neurophysiol 94:2644-2652, 2005) that a serotonin-induced enhancement of the LG response was mediated by an increase in cAMP levels following activation of adenylate cyclase; however, octopamine acts independently. Here, we clarify how octopamine enhances the LG response during sensory stimulation using physiological and pharmacological analyses. When phospholipase C inhibitor U-73122 was directly injected into the LG before biogenic amine application, it abolished the enhancing effect of octopamine on direct sensory input to the LG, but did not block indirect input via sensory interneurons or the effect of serotonin. Direct injection of IP(3), and its analogue adenophostin A, into the LG increased the synaptic response of the LG to sensory stimulation. Thus, IP(3) mediates octopamine-induced synaptic enhancement of the LG, but serotonin acts independently. These results indicate that both octopamine and serotonin enhance the synaptic responses of the LG to sensory stimulation, but that they activate two different signaling cascades in the LG.

Development of agonistic encounters in dominance hierarchy formation in juvenile crayfish.

We have characterized the behavioural patterns of crayfish during agonistic bouts between groups of crayfish of four different body lengths (9-19, 20-32, 41-48 and 69-75 mm) to characterize changes in the patterns of agonistic encounter during development. The behaviour of both dominant and subordinate animals was analysed by single frame measurement of video recordings. Behavioural acts that occurred during agonistic bouts were categorized as one of seven types: capture, fight, contact, approach, retreat, tailflip and neutral. Dominant-subordinate relationships were formed between juvenile crayfish as early as the third stage of development. Patterns of agonistic bouts to determine social hierarchy became more aggressive during development. The dominant-subordinate relationship was usually determined after contact in crayfish of less than 20 mm and 20-32 mm in length, while several bouts of fights were necessary for crayfish of 41-48 and 69-75 mm in length. Furthermore, social hierarchy was formed more rapidly in small crayfish. In larger animals, the number of approaches by dominant animals that promoted retreat in subordinate animals increased after the establishment of the winner-loser relationship. In smaller crayfish, in contrast, no measurable changes in these behaviour patterns were observed before and after the establishment of the winner-loser relationship. With increasing body size, the probability of tailflips decreased while that of retreats increased as the submissive behavioural act of subordinate animals.