in session (5) pairs were again isolated. At the middle of each session,
except for session (3), both members of the pair were simultaneously
presented with a live a crab and the latency of attack to the crab was
measured. Crabs were attached to a thin string that was quickly pulled
to remove them before being seized. The timeline of the behavioural
experiment has been chosen basing on results found in a previous
study on E. cirrhosa by Malham et al. [18]. Indeed, these authors reported
a rapid clearance of live bacteria fromthe circulationwithin 4 h from
injection.We thus decided to concentrate our observation immediately
after injection andwithin 2 h frominjection to catch the period of maximum
immune activation.
The behaviour of each pair was video-recorded and subsequently
analysed to avoid any effect of disturbance due to the presence of an observer.
For each session we recorded the following behaviours: ì) time
spent inside the den over the total observation time, ìì) time spent outside
the den in activity over total observation time. Individuals were
considered in activity when exploring the surroundings by crawling
behaviour, i.e. moving relatively slowly in contact with the ground
and with the tank walls or performing brief swimming sequences
[29], and ììì) latency to attack the crab offered. In the sight-allowed condition
we also recorded ìv) time spent interacting with the other individual
of the pair over total observation time. Since physical contact
between individuals was prevented by a transparent partition an individual
was considered to visually interact with the other member of
the pair when moving onto the partition and showing active interest
towards the conspecific. This interest was demonstrated through the
typical gradual raising and lowering of the head and sometimes splaying
of the arms towards the conspecific [29].