allogrooming than P. dives. It therefore appears that disease
resistance in A. echinatior places a stronger emphasis on
group-level defence than P. dives which places more
emphasis on individual defence.
The gland blockage experiment provides evidence that
the secretion from the venom gland is a further mechanism
by which weaver ants resist parasites. Blocking the
metapleural gland of leaf-cutting ants with nail varnish
has previously been shown to substantially increase their
susceptibility to Metarhizium (Poulsen et al. 2002), and we
also found this to be the case in our study. Blocking the
acidopore of weaver ants had a similar effect by reducing
their survival when they were exposed to Metarhizium, but
not under control conditions. In both species, gland
blockage did not increase the mortality of ants exposed to
the control solution, so did not appear to result in any
general physiological stress. Weaver ants with a functioning
venom gland therefore are more resistant to Metarhizium,
suggesting that the secretion has antimycotic properties.
This was confirmed in vitro with the venom inhibiting the
germination of 100% of both Metarhizium and Aspergillus
spores to the same extent as >10% formic acid. The
secretion of formic acid by the venom gland is characteristic
of the Formicinae ant subfamily (Hölldobler and
Wilson 1990; Blum 1992), so the antimicrobial action of
venom would seem likely to be a common feature of
formicine ants that may make the metapleural gland of less
importance in this subfamily. It is notable in this regard that
the species in Experiment 1 with the greatest resistance to
Metarhizium was Formica fusca, a formicine ant which
both produces formic acid in its venom and which has a
metapleural gland. However, blocking the venom gland of
A. echinatior workers also reduced their resistance to
Metarhizium and, surprisingly, did so to a similar extent to
blockage of the metapleural gland. Venom of ants from
other subfamilies has been shown to have antimicrobial
properties (Storey et al. 1991; Orivel et al. 2001; Zelezetsky
et al. 2005), and it may therefore be the case that the venom
of ants in general is important in disease resistance. The
silken nests of P. dives are contaminated with many viable
spores of fungi (Fountain and Hughes 2011), so while the
venom may be used in directly defending the ants, it would
not appear to be used in sterilising the nest.
The results suggest that the while the metapleural gland
may be important in the disease resistance of some ant
species, it is not necessarily essential, at least for the
resistance of mature, adult ants. Weaver ants which lack the
metapleural gland are individually more resistant than some
ants which have the gland and this appears to be at least in
part due to higher frequencies of self-grooming behaviour.
Furthermore, the formic acid-containing venom of weaver
ants, and also the venom of leaf-cutting ants, appears to
have antimicrobial properties making it too important in
resistance to parasites. The results therefore emphasise the
multiple, complementary behavioural and chemical defence
mechanisms that ants use to resist disease. It will be very
interesting to see whether the disease resistance of other
species of weaver ants is similar to that of P. dives, and how
the evolutionary regaining of the metapleural gland by
some species of Camponotus has impacted their disease
resistance.