BR. .1. LINT. NA'I. HIST., (,: 1993 



Two different aspects of the attack behaviour of the anis when they encountered 

 a ladybird were recognized and recorded; (i) one or more ants managed to grab hold 

 of the ladybird; and (ii) one or more ants squirted formic acid at the ladybird, 

 recognizable when an ant curls its abdomen under its body towards the ladybird. 

 This was often difficult to see and may not have been recorded in all cases. A variety 

 of ladybird behaviours were recorded (see Table 3). 



Each encounter, defined as contact between an ant and a ladybird which had not 

 been in contact with any ants for the previous 10 seconds, was recorded. This was 

 designed to eliminate cases when ants came in to help or to take over from others 

 which had already attacked. 



Each encounter was scored as either 'investigate' or 'escalate'. Investigate indicates 

 that the ant touched the ladybird, generally with its antennae, but did not fight. 

 Escalate indicates that the ant did attack. Attacks generally took the form of biting 

 at the elytra or legs, squirting formic acid, or pushing to try to dislodge the ladybird. 

 It should be noted that our analysis does not take into account the duration or ferocity 

 of the attack — it only analyses the initial decision to attack. 



Hypothetically, this decision will depend on the ants making some assessment of 

 the threat to their food supply, in this case the colony of aphids, which in turn could 

 depend on a variety of factors: how valuable the food source is, whether the food 

 is scarce or abundant, how far the food source is from the ants' nest, and how close 

 the predator is to the colony. To assess this final factor, encounters were divided 

 into those before the ladybird reached the colony and those while the ladybird was 

 actually on the colony. 



Once a ladybird had been found on the colony the ants seemed to become more active 

 and attack. To assess this the encounters between ants and ladybirds on the colony 

 were split into those within the first 2 minutes after contact with the colony and those 

 more than 2 minutes after initial contact. Two minutes were chosen arbitrarily. 



As a large number of repeat experiments were conducted on the same colony there 

 is a possibility that the ants' basic level of hostility changed during the day — this did 

 not appear to be the case, but it is a possible criticism of the method used. 



Results 

 Summary of observations 



A summary of the results of encounters between ants and ladybirds is given in 

 Table 2. Typically, when introduced onto a stem, the ladybird ran up and reached 

 the colony of aphids fairly quickly. In most cases the encounter with an ant occurred 

 after the ladybird had reached the colony. The ant first palpated the ladybird with 

 its antennae. In many cases the ants then escalated to attack behaviour. Other ants 

 often joined in the attack, up to five being seen attacking at one time. Each ant only 

 persisted with an attack for about 30-60 seconds, although continual recruitment of 

 new ants meant that ladybirds were often under sustained attack for several minutes. In 

 cases where an ant managed to secure a hold on the ladybird with its mandibles, usually 

 on a leg, the ant's attack was often sustained for longer, in one case for 14 minutes. 



It appeared that the elytra of adult ladybirds are defensively very effective against 

 ants and, in our experiments, ants only gained an effective hold on six of the 67 

 ladybirds used. 



In cases where an ant encountered a ladybird with an aphid in its mandibles, the 

 ant tended to concentrate on retrieving the aphid rather than attacking the ladybird. 

 This was despite that fact that the aphid was often already fatally injured. 



