Tropical Forest Community Ecology

Tropical Arboreal Ants 343

species dominate EFNs on plants where they
tend trophobionts (authors’ observations), and
a phalanx of Do. quadridenticulatus workers
can wrest control of EFN plants from aggres-
siveCa. sericeiventrisworkers (Davidson personal
observation). Nevertheless, lacking wide-ranging
foraging, trophobiont-tenders may encounter
fewer potential prey (apart from tended tropho-
bionts) than do leaf-foraging formicines. Consis-
tent with this conjecture, larval formicines have
the anatomical structures to process solid food,
whereasdolichoderinelarvaedonot(Wheelerand
Wheeler 1976), though they may consume insect
haemolymph, particularly that of trophobionts
(Dillet al.2002).
Small-bodied taxa may be more likely than
their larger counterparts to accept low quality
resources, because absolute locomotory costs are
greater for large-bodied workers (although offset
somewhat by longer stride lengths; Fewellet al.
1996). Moreover, if they forage regularly over
shorter distances, this could also allow them to
be less selective (Davidson 1978). Consistent with
these hypotheses, SUCmin declines marginally,
and AAmin significantly, with log of ant body
size (P=0.06 andP=0.03, respectively). Addi-
tionally, small-bodied taxa tend to have popu-
lous colonies that potentially both increase the
colony’s capacity for intensive and extensive
search (e.g., Swain 1980 forCrematogaster) and
confer numerical advantage in battles of attri-
tion (McGlynn 1999, Palmer 2004). The afore-
mentioned attributes should contribute to both
exploitative and interference competitive ability,
making these taxa difficult to dislodge from con-
trolled resources. However, except where nesting
polydomously, and due to their limited foraging
ranges, smaller taxa may prevail more often at
localized resources than over widely distributed
ones, on a scale, for example, of whole trees.
Guild structure, or the structure of inter-
actions within communities, is a key determi-
nant of local species diversity (May 1972), and
once foraging functional groups are more finely
elucidated for arboreal taxa, it will be impor-
tant to determine whether interspecific inter-
actions are stronger within or between those
groups. It is obvious, however, that we currently
know little about the interaction structure within

communities of opportunistic arboreal rainforest

exudate-feeders, and about how that structure

maps to phylogenetic structure. Such questions

must be resolved if we are ever to fully understand

one of the earliest and most commonly noted pat-

terns in such communities, that is, associations

between particular territorial dominants and the

specific non-dominant ants capable of coexisting

within their territories (e.g., Leston 1978, Majer

1993, Dejean and Corbara 2003, Blüthgenet al.

2004b, but see Davidsonet al.2007).

Foraging functional groups and plant

defense

We return briefly to the matter of which

free-living ant taxa might be associated with

reduced damage to plants on which they for-

age. Table 20.1 summarizes attributes of the two

broad functional groups as they bear on that sub-

ject. By 7 of 10 criteria, leaf-foragers appear more

likely than trophobiont-tenders to exert a net pos-

itive effect on plants where they forage. Of these

seven, criteria not previously discussed include

alteration of plant metabolism and development,

frequenttransmissionof plantpathogensthrough

stylets of tended trophobionts (e.g., Buckley

1987), and plant resource losses to ants and tro-

phobionts.Althoughtwocriteriaapparentlyshow

greater benefits from trophobiont-tenders, they

could actually be associated with higher resource

lossesif theprincipaleffectof trophobiont-tenders

per unit time and worker number are negative,

as the first seven criteria suggest. By the final

criterion, effects of leaf-foragers and trophobiont-

tenders may be equivalent on average, with spa-

tially territorial, carnivorous, and small-bodied

taxa within each group exhibiting the greatest

N-deprivation (Davidson 2005). We do not mean

to suggest that trophobiont-tenders are always

less desirable associates of plants than are leaf-

foragers, but rather argue from first principles

why that should be so on average for free-

living ant taxa. Against natural enemies deterred

only by large numbers of foraging workers,

trophobiont-tenderscouldbethe“preferred”asso-

ciates, and plant protection has been demon-

strated for at least some trophobiont-tenders,