UK. J. INI . NAI . HIST., (< 1993 



The response of ladybirds to ant attacks varied between species. These differences 

 are summarized later. In general ladybirds either clamped down as soon as attacked, 

 presumably to prevent ants gaining an effective hold on them, or they moved away 

 from the colony. In some cases the ladybirds only moved a short distance from the 

 colony, returning to it once the ant attacks had subsided. In other cases, the ladybirds 

 moved further from the colony and did not return. 



After a sustained attack, and especially if squirted with formic acid, the ladybird 

 often escaped up a leaf and clamped down for several minutes before starting to clean 

 itself. This self-cleaning took several minutes on and off, the duration of cleaning 

 being longer when they had been sprayed with formic acid. 



One ladybird defence mechanism is 'reflex bleeding' (Majerus & Kearns, 1989). 

 When attacked, a pungent yellow fluid is exuded from pores in the ladybird's legs, 

 from where it runs along channels to the edge of the pronotum or elyta, where it 

 forms small droplets. This defence was not used by the adult ladybirds against the 

 ants in any of our experiments. On the other hand, the ladybird larvae did reflex 

 bleed in response to sustained ant attacks. The reflex blood was secreted mainly from 

 the 1st, 2nd, 8th and 9th abdominal segments, although other points of secretion 

 were seen and it is possible that a secretion can be at any attack point on the abdomen, 

 or at the base of the legs. It appeared to be used as a last-ditch defence against 

 prolonged attacks. This may be because it reduced mobility as it tended to foul-up 

 the larva's legs. Individual larvae only produced large amounts of fluid once, implying 

 that the reservoir of fluid available at any time is limited. Indeed, it has recently been 

 shown that adult A. bipunctata and C. 7-punctata only have a limited supply of reflex 

 blood (de Jong et al., 1991; Holloway et al., 1991). 



The failure of adults to reflex bleed in response to ant attacks probably indicates 

 first, that reflex blood is a valuable resource which is costly to replenish, and secondly, 

 that ants are not a serious threat to adult ladybirds, at most depriving them of a meal 

 and costing them cleaning time. On the other hand, larvae, with their softer 

 exoskeletons, are far more vulnerable to ant attacks, and are more likely to be killed 

 (Majerus, 1989). Their use of reflex bleeding against ants is presumably a reflection 

 of this greater vulnerability. 



In terms of ladybird success, that is a ladybird actually managing to eat an 

 aphid, a summary of the results is given in Table 2. For carnivorous ladybirds, the 

 success rate on ant-tended aphid colonies was significantly lower than on the untended 

 colony (x?= 19.84, P< 0.001 with Yates' correction). (This test uses totals released 

 onto plants rather than just the number which reached the colony because it was 

 noted that presence of ants on a plant could prevent the ladybird reaching the 

 colony.) This result shows that ants do significantly reduce the effect of predation 

 by ladybirds on aphid colonies. That five aphids were eaten by ladybirds on the plant 

 supporting ant-tended colonies, but away from the main colony, suggests that first, 

 it is ordinarily the main colony that the ants defend, and secondly, there is a 

 considerable advantage to aphids in remaining within the main colony, in terms of 

 reduced risk of predation. 



The results of the ants' responses to ladybirds in terms of investigation and escalation 

 are given in Table 4. The level of ant response before the ladybird reaches the colony 

 can be used as a basic level of hostility before the ladybird has posed a direct threat 

 to the aphid colony. Statistical comparisons of this base level with the level of response 

 in the other classes show that the proportion of ants which escalate attacks is 

 significantly increased while the ladybird is on the aphid colony (xf = 13.18, 

 P< 0.001), and during the first 2 minutes after the ladybird has left the aphid colony 

 (x? = 7.29, P<0.01). However, there is no significant difference between the base 



