However, the S. invicta venom solution alone and
S. invicta venom solution followed by Dufour’s
solution groups died at similar, very high rates
(Wilcoxon: c2 = 2.1, df = 1, P = 0.15) (Fig. 3A).
Dufour’s gland products do not detoxify S. invicta
When N. fulva venom was tested, survival times
differed between groups (Wilcoxon: c2 = 44.1,
df = 2, P < 0.0001), with very low mortality in the
N. fulva venom–treated control group (10%). The
fate of workers in the S. invicta venom followed
by carrier solution group differed dramatically
from those in the S. invicta venom solution followed
by N. fulva venom solution group (Wilcoxon: c2 =
25.7, df = 1, P < 0.0001). Only 20% of workers
treated with S. invicta venom alone survived, as
compared to the survival of 100% of workers
treated with both venoms sequentially (Fig. 3B).
Therefore, the venom gland of N. fulva contains
the detoxifying agent.
As is clear from the odor and respiratory trauma
resulting from collecting N. fulva with an aspirator, N. fulva venom consists primarily of concentrated formic acid (15). Using the same Argentine
ant–based bioassay, we tested whether formic
acid is the detoxifying agent. We used a formic
acid solution with a pH equivalent to that of the
N. fulva venom solution used in the previous assay. Controls were application of formic acid solution followed by carrier solution. Treatments were
application of S. invicta venom solution followed
by carrier solution, and application of S. invicta
venom solution followed by formic acid solution (14).
Treatments and controls differed in survivorship (Wilcoxon: c2 = 51.1, df = 2, P < 0.0001), with
no mortality among formic acid–treated control
workers. Again, workers treated with S. invicta
venom followed by carrier solution suffered high
Fig. 3. Results of bioassays. Survivorship curves for topically treated
L. humile workers. Legends at bottom
left identify the sequence of solutions applied. Solutions consisted
of the test agent diluted in a carrier
solution of water plus surfactant.
Si denotes S. invicta; Nf denotes
N. fulva. (A) Test of N. fulva Dufour’s
gland as the source of the detoxifying agent. (B) Test of N. fulva venom
gland as the source of the detoxifying agent. (C) Test of formic acid as
the detoxifying agent. Triangular symbols denote controls, and circular symbols denote experimental treatments.
Error bars present standard errors of
the Kaplan-Meier survivorship estimator. The inset illustrations detail
internal gland anatomy and chemical structures. The source gland is
Si venom carrier
Si venom Nf Dufours
Nf Dufours carrier
Time since treatment (hrs)
Si venom carrier
Si venom formic acid
Formic acid carrier
Si venom carrier
Si venom Nf venom
Nf venom carrier
8 765 4 321
Fig. 4. Evaluation of the specificity of detoxification behavior. The
dependent variable is a 2-min count of N. fulva detoxification behaviors
after chemical conflict with a variety of test species (x axis). Symbols indicate the median number of detoxification behaviors expressed. Error bars
present interquartile ranges. The horizontal dashed line separates no conflict from post-interaction data. Interactions terminated when the test
species deployed its defensive compounds against the N. fulva worker.
Horizontal lines and text at bottom indicate species progressively less
related to S. invicta. The inset image shows an N. fulva worker performing
the detoxification behavior.
Solenopsis invictaSolenopsis richteriSolenopsis geminata Monomorium inimum Crematogaster laeviusculaForelius pruinosus Linepithema humile No conflict