336 Diane W. Davidson and Steven C. Cook
nutrient flows as essential for understanding both
the functional biology of individual organisms
and higher-level processes in populations, com-
munities,andecosystems(SternerandElser2002,
Raubenheimer and Simpson 2004). Food stor-
age aside, consumers act to maintain elemen-
tal homeostasis in body composition, and they
respond ecologically and evolutionarily to dietary
resource ratios. All organisms regularly harvest
energy and matter from their environments, con-
vert a fraction of these resources to biomass, and
return energy and materials back to their sur-
roundings. Natural selection should have molded
each of these processes by producing regulatory
mechanisms that both enhance acquisition or
retention of limiting macronutrients and either
dispose of nutrient excesses or channel them to
useful functions (Davidson 1997, Raubenheimer
and Simpson 1998, 2004). Selection for such reg-
ulation should be strongest in organisms with
nutritional imbalances, and given the remarkably
broad range of dietary resource ratios utilized
across the ants, these insects represent a com-
pellingstudysystemforresearchintotheeffectsof
elemental and ecological stoichiometry on insect
ecology and evolution (Davidson 1997, Cook
and Davidson 2006). Importantly also, workers
can be assayed directly for evidence of resource
imbalances (Kay 2002, Davidson 2005).
Earliest ants were stinging huntresses, pro-
visioning prey to brood in terrestrial nests
(Hölldobler and Wilson 1990). For taxa adapted
to forage on highly connected planar terrestrial
surfaces, colonization of the three-dimensional,
poorly connected, and vertically structured arbo-
real zone must have been accompanied by marked
increases in foraging costs. Ancestors of contem-
porary arboreal taxa appear to have solved this
problembyexploitingenergy-richplantandinsect
secretions (EFNs, plant wound sap, and tropho-
biont honeydews) to fund energetically expen-
sive foraging. (Present-day examples of terrestrial
taxa transitioning to arboreal life include some
Myrmicaria,Pheidole, andAnoplolepis, as well as
Paraponeraclavata.) Recent isotopic studies suggest
that many exudate-feeders also obtain substantial
N from exudates (Blüthgenet al.2003, Davidson
et al.2003, but see below). If this is so, ants must
be ingesting and processing very large volumes
of CHO-rich exudates for the sparse N they con-
tain.Digestivesystemsspecializedtoprocessliquid
foods in volume (Eisner 1957, Davidsonet al.
2004), perhaps together with nutrient contribu-
tions from microsymbionts (see below), may have
enabled these taxa to persist and even thrive on
highly imbalanced, N-poor diets.
High CHO:N dietary ratios of exudate-feeders
should be correlated with availability of “excess
CHOs,” that is, CHOs exceeding those fund-
ing growth and reproduction through primary
metabolism (Davidson 1997). Natural selection is
postulated to have turned these energy sources
to good use in supporting large populations of
CHO-dependent workers (Tobin 1994) and sub-
sidizing pursuit of N-rich prey by one or more of
three mechanisms (Davidson 1997): (1) increased
activity “tempos” (velocities), correlated with
higher “dynamic densities” of foraging workers;
(2) defense of absolute spatial territories and
associated prey; and (3) N-free or N-poor defen-
sive/offensive chemical weaponry. A review of
existing information provided at least circum-
stantial evidence for each of these mechanisms,
all of which contribute to ecological dominance
(Davidson 1997, see also Yanoviak and Kaspari
2000).
If arboreal exudate-feeders have evolved to
invest excess CHOs in mechanisms ensuring
ecological dominance (and thus access to lim-
iting N), to what extent do these taxa remain
N-deprived? Answering this question could also
help to resolve the degree to which vari-
ous ant taxa convey anti-herbivore protection
to plants. Recently, Davidson (2005) evaluated
N-deprivation in a behavioral assay of 54 arbo-
real or terrestrial Amazonian ant taxa with
diets ranging from carnivorous to highly her-
bivorous, as evaluated by worker δ^15 N ratios
(Davidsonetal.2003). Relative N-deprivation was
quantified as an exchange ratio (ER), defined as
SUCmin/AAmin, or the minimum sucrose con-
centration divided by the minimum amino acid
concentration, accepted as food by ≥50% of
tested workers (see Kay 2002, pioneering this
approach). ER values averaged almost five-fold
higher (corresponding to greater N-deprivation)
for “N-omnivores” and “N-herbivores” (N-OH
taxa) than for “N-carnivores,” that is, taxa which