The Southern rock lobster, Jasus edwardsii, is a temperate species of spiny lobster with established well managed fisheries in Australia and New Zealand. It has also been under consideration as a species with aquaculture potential. Agonistic behaviour has important consequences under aquaculture conditions that encompass direct effects, such as damage or death of protagonists, and indirect effects on growth that relate to resource access, principally food and refuge. This study aimed to identify and characterize behaviours and to make a preliminary investigation of their occurrence under tank culture. Juvenile Jasus edwardsii were examined in a flow-through seawater system using a remote video camera system. Twenty-nine behaviours were divided into three sub-groups: aggressive (11), avoidance (6) and others (12). Aggressive behaviours included attacks, pushing, lifting, clasping and carrying an opponent. Avoidance behaviours included moving away in a backwards-, forwards- or side-stepping motion as well as with more vigorous tail flips. These behaviours were components of twelve behavioural groups that described contact, attack and displacement between individuals. Activity was crepuscular with two clear peaks, one in the morning and the other in the evening. The occurrence of behavioural groups was not different between the morning and evening. The frequency of aggressive behaviours was not affected by changes made to stocking density or access to food. The implications of agonistic behaviours are discussed further in relation to developing aquaculture.

Spiny lobster, aquaculture, con-specific, behaviour, inter-individual interaction, growth depensation

Spiny lobster species are found around the world from tropical to temperate seas, they are important in relation to conservation, fisheries and aquaculture (Green et al. 2010). The Southern rock lobster, Jasus edwardsii, is a temperate species with established and well managed fisheries in Australia and New Zealand (Holland et al. 2005, Linnane et al. 2012, Linnane et al. 2013). It too has also been under consideration as a species with aquaculture potential (McKoy 1979, Crear et al. 2000, Simon and James 2007). Some advantages relate to reported behavioural characteristics that make it amenable to group culture and arguably more suitable for aquaculture than other lobster species, particularly clawed lobsters (McKoy 1979, Huber and Kravitz 1995, Thomas et al. 2003).

Clawed lobsters exhibit an array of aggressive behaviours, they readily fight to establish and maintain dominance hierarchies and therefore exhibit behaviours that are not ideal for high productivity under typical intensive communal aquaculture systems (Karavanich and Atema 1998, Kravitz 2000, Gherardi et al. 2010). Agonistic behaviour has important consequences under confined aquaculture conditions that relate to access to resources, principally food and refuge, and that encompass direct effects, such as damage or death of protagonists, and indirect effects on growth and health (Thorpe and Huntingford 1992, Drengstig and Bergheim 2013). Hierarchical social structures often lead to feeding hierarchies where there is unequal distribution of food between individuals, differences in food intake result in growth depensation (McCarthy et al. 1992, Winberg et al. 1993). Growth depensation is an increase in variance of a size distribution with time due to individuals growing at different rates (Magnuson 1962, Carter et al. 1996), and its management is critical for aquaculture production (Thorpe and Huntingford 1992, Drengstig and Bergheim 2013).

Compared to clawed lobsters (Huber and Kravitz 1995, Drengstig and Bergheim 2013) less is known about the agonistic behaviour of spiny lobsters, especially in relation to aquaculture (Thomas et al. 2003, Moyle et al. 2009). Spiny lobsters exhibit an extensive range of behaviours including agnostic behaviours which are particularly prevalent around obtaining and retaining shelter (Fielder 1965, Berrill 1976, Cobb 1981, Segura-García et al. 2004). Spiny lobsters aggregate in and around shelters in the wild, yet they also compete for shelters and show aggression around them (Childress 2007). The more solitary species of spiny lobsters are considered more aggressive (Childress 2007) and Jasus edwardsii, the subject of the current study, appears more gregarious and less aggressive than some other spiny lobsters (Fielder 1965, Berrill 1976, James et al. 2001, Moyle et al. 2009). Chemicals in the urine and physical contact are the important mechanisms mediating social interactions including aggregation, mating, and agonistic behaviours in spiny lobsters (Briones-Fourzan et al. 2008, Horner et al. 2008, Shabani et al. 2009). Agonistic interactions often involve overt aggressive and submissive behaviours, published information on the nature of these interactions in captivity is spread across spiny lobster taxa Jasus (Fielder 1965, Thomas et al. 2003) and Panulirus (Berrill 1976, Cobb 1981, Segura-García et al. 2004, Moyle et al. 2009). Consequently, the current study aimed to identify and characterize behaviours of Jasus edwardsii and to make a preliminary investigation of their occurrence under tank culture conditions and in relation to some key factors that can be manipulated in order to improve lobster growth and productivity.

Juvenile Jasus edwardsii stock were caught as puerulus from the East coast of Tasmania and maintained in 4 m³ rathbun tanks. Stock and experimental animals were held at 18 °C, and fed on a mixed diet of whole blue mussel, frozen squid and a commercial prawn feed. Lobsters held under standard conditions were used in experiments (Crear et al. 2000). Only the prawn feed was used during the experiments and fed at 1% body weight per day (Thomas et al. 2003). All animals survived and were returned to stock, they were only used once.

Behaviours were examined in two rectangular 25L (width 300 × length 500 × depth 250 mm) tanks that were part of a flow-through seawater system. The flow rate was 7.4 L min^-1 so that water was replaced approximately 18 times per hour. The system was enclosed by black plastic sheeting to control photoperiod and minimise visual disturbance from external sources. Light was provided from fluorescent tubes and a 12:12 photoperiod used with lights-on at 06:00 and lights-off at 18:00. An infrared light provided illumination for the camera during the dark phase. Each tank system was equipped with a video camera and the tanks were stocked with 5 animals (equivalent to 33.5 individuals per square metre, ind. m^-2). The carapace was marked to identify each individual. The floor of each tank was marked with a 100 × 100 mm grid so distance and direction of movements could be assessed more easily.

An initial assessment of activity was made on two groups of five animals that were monitored continuously for seven days. Focal sampling (Altmann 1974, Shelverton and Carter 1998) was used to record activity for the first ten minutes of every hour. The unit of activity was defined as movement equivalent to half the length of the lobster. Peaks of activity occurred in the morning between 07:00 and 10:00 and in the evening between 18:00 and 21:00. Twenty-nine component behaviours were identified and used to define twelve behavioural groups (see Results) including when no behaviour was observed (none).

Four new groups of five animals were used to investigate the occurrence of the behavioural groups. Occurrence was the number of times a behavioural group was observed as the first behaviour exhibited during each period of observation: each occurrence of behavioural unit was given a score of 1, observations were made over five days and data adjusted to one hour of observation (Moyle et al. 2009). The effect of time of day was then analysed using these four groups to compare occurrence of behavioural groups between the morning (AM) and evening (PM) periods. The effects of stocking density and feed availability on the occurrence of behavioural groups were measured in separate experiments. The experiments on stocking density and feeding availability were of a preliminary nature and each used 2 groups (10 animals). Stocking density compared standard with a higher density: density was doubled from the standard density of 33.5 to 67.0 ind. m^-2 by dividing the tanks in half (one half remained empty during the experiment). The effect of feed availability was investigated by comparing one with four feeding stations: the standard single feeding station in the centre of the tank or using four feeding stations simultaneously, one in each corner of the tank (the total ration was the same).

For each experiment the same sampling regime was used: eight 15 min time blocks, four from the morning (05:00–06:45) and four from the evening (18:00–19:45) periods of peak activity, were analysed and behavioural group recorded against individuals (occ. h^-1). Mean group occurrence was calculated for each behavioural group as total occurrence divided by the number of individuals (five) and standardised to an hour of observation, it was expressed as occurrence per individual per hour (occ. Ind^-1 h^-1). Non-parametric statistical analysis using the Kruskal-Wallis test followed by multiple comparison using all pairwise comparisons was conducted due to preliminary nature of the data and the lack of homogeneous variance according to Levene’s test. Statistical analysis was conducted using SPSS Statistics version 22.

Agonistic behaviour has been divided into approach- and avoidance-orientated behaviour (Huber and Kravitz 1995). Clawed lobsters exhibit extensive agonistic behaviours, they are aggressive, solitary and cannibalistic when cultured together (Drengstig and Bergheim 2013). The American lobster, Homarus americanus, has an extensive repertoire of aggressive behaviours that include use of the large claws to signal, touch, manipulate and damage opponents. An important aspect of aggressive bouts is to displace opponents, this includes attempting to lift and turn them on their backs as well as pushing or pulling them away from the substrate (Huber and Kravitz 1995). In contrast, many spiny lobsters are reported to be gregarious and only vulnerable to cannibalism following moulting and might, therefore, be predicted to exhibit behaviours that facilitate co-habitation (Childress 2007). However, in wild Jasus lalandei this was a function of the capacity of the shelter and smaller individuals were actively displaced by larger individuals (Fielder 1965). Furthermore, in aquaculture the local tank environment is a critical factor that influences behaviour with consequences on growth and production (Thomas et al. 2003, Moyle et al. 2009). The current research aimed to make a detailed study of Jasus edwardsii behaviours and behavioural groups in order to provide a basis for understanding the effects of the aquaculture environment.

Twenty-nine behaviours were detected and characterised into three sub-groups in this study: aggressive (11), avoidance (6) and others (12). Among the behaviours recorded, the occurrence of medium and high intensity attacks and displacement events were minimal. This suggested J. edwardsii juveniles are highly gregarious in culture, even at relatively high stocking densities (67 individuals per m²). Further research into characterising the agonistic behaviours of various spiny lobster species with potential for commercial aquaculture is warranted. Better characterisation of agonistic behaviours in culture will assist in testing and improving husbandry management strategies which may include the use of chemical cues, better feed management and shelter designs.

Particular thanks to Yvette Watt for the line drawings of lobsters. Professional staff at Taroona for assistance in maintenance of lobsters.