Management of invasive invertebrates: lessons from the management of an invasive alien ant - Invasive Species Management - page 166

boundaries of supercolonies were unlikely to encounter bait because they rarely
emerge from their burrows in the dry season, when aerial baiting operations were
conducted. Robber or Coconut Crabs ( Birgus latro ) were of more concern because
they forage more widely and are more active under dry conditions. Moreover,
the robber crabs on Christmas Island are genetically distinctive from robber crabs
elsewhere (Lavery et al . 1996), and, until the YCA invasion were the largest and
least impacted population anywhere in the world (Schiller 1988). Although the
YCA extirpates robber crab populations in supercolonies, baiting operations posed
a signifi cant threat because Birgus have a well-developed olfactory system, move
large distances for food, are attracted to the fi shmeal matrix in Presto ® 01, and are
extremely susceptible to fi pronil.
Considerable eff ort was invested in minimizing non-target impacts on this cha-
rismatic species. Two kinds of attractive food lures were used to entice robber crabs
from baited areas during helicopter operations. First, 44 senescent trees of the
endemic palm Arenga listeri were felled several weeks prior to aerial operations, sev-
eral hundred meters apart and up to 300m away from baited supercolony bound-
aries: the pith of this monocarpic palm attracts robber crabs in large numbers
from hundreds of metres. Second, where possible, targeted supercolonies were also
ringed with depots of poultry food pellets mixed with shrimp paste. h is lure had
been used during initial ground-based control. Using the CIGIS, over 250 sites
were selected around supercolony perimeters. One or two days before areas were
aerially treated with ant bait, 15 kg of the lure was dropped from the helicopter at
intervals of 150-250 m around the perimeter and at distances between 50-200 m
outside supercolony boundaries.
h e eff ect of these lures on mitigating robber crab mortality was assessed after
the aerial baiting operation. Mortality was estimated at 5.3% for sites close to edges
of baited supercolonies, but because all supercolonies were ringed with food lures,
there were no supercolonies without lures to act as controls. However, estimates
made during initial ground baiting indicated mortality rates of 15% in the absence
of food lures (D. Slip, pers. comm.). In spite of this considerable eff ort, a com-
prehensive assessment of food lures to reduce Birgus mortality in association with
baiting has since shown them to be ineff ective (h omas 2005).
An analysis of non-target impacts on litter invertebrates during ground control
operations using the identical bait formulation (but at a higher rate of application)
found no evidence for off -target impacts (Marr et al . 2003). Despite several reports
of dead and dying introduced cockroaches in a hollow tree stump, and mortality
of the introduced ant Campanotus melichloros in the immediate aftermath of the
aerial operation, it is highly unlikely that the aerial operation caused broad and
substantive non-target impacts on litter invertebrates. h e sheer abundance of the
foraging YCA in supercolonies meant that these ants largely monopolized the bait
(Marr et al . 2003). Contractors engaged to assess impacts on canopy invertebrates,
forest reptiles, and land birds found no evidence of non-target impacts (Stork
et al . 2003).