These positive, population-level consequences of the
association between A. gracilipes, scale insects, and sooty
moulds help explain the lower proportion of growing shoots
on canopy trees at ant-invaded sites (Table 1h) and the
higher frequency of tree dieback. Tree dieback depended on
ant invasion [at invaded sites, 51% of trees (146/286)
showed evidence of dieback; at uninvaded sites just 18% of
trees were affected (53/294), v2 ¼ 77.7, P