Distinct antennal lobe phenotypes in the leaf-cutting ant (Atta vollenweideri).

Leaf-cutting ants (Atta vollenweideri) express a remarkable size polymorphism across the two sexual castes (queens and males) but in particular within the worker caste. Worker size is related to behavior (alloethism), separating workers into behavioral subcastes. The neuronal mechanisms underlying differences in behavior within the worker caste are still unknown. In this study, we first compared selected neuropils, in particular, the antennal lobes (AL) in males, queens, and workers. The males' ALs contain three extremely large, sex-specific glomeruli (macroglomeruli; MGs) and in total comprise fewer glomeruli (242) than the ALs of queens (about 346 glomeruli). In contrast to males, the queen ALs contain only one large glomerulus at a lateral position. The largest number of glomeruli was found in workers (396-442). In a previous paper, we described an MG in the workers' AL, and, in the second part of this study, we show that within workers two distinct, size-related AL phenotypes exist: the MG phenotype (containing a macroglomerulus) and the RG phenotype, with all glomeruli of regular size. This neuroanatomical polyphenism is established during pupal development and separates the worker caste into two neuroanatomical subcastes. Third, we investigate the functional significance of the MG in workers. By using calcium imaging to monitor activity of AL projection neurons, we show that the releaser component of the trail pheromone is represented in the same region as the MG. We propose that phenotypic trait variation in the organization of the ALs leads to differences in odor information processing that finally result in size-related differences in trail-following behavior.

Perceptual differences in trail-following leaf-cutting ants relate to body size.

Leaf-cutting ants of the genus Atta have highly size-polymorphic workers, and size is related to division of labor. We studied trail-following behavior of different-sized workers in a laboratory colony of Atta vollenweideri. For small and large workers, we measured responsiveness and preference to artificial conspecific and heterospecific pheromone trails made from poison gland extracts of A. vollenweideri and A. sexdens. Responsiveness was measured as the probability of trail-following, and preference was measured by testing the discrimination between one conspecific and one heterospecific trail. Minute amounts of the releaser component methyl-4-methylpyrrole-2-carboxylate (0.4pg/1m), present in both, conspecific and heterospecific trails, suffice to elicit trail-following behavior. Workers followed heterospecific trails, and these trails (after normalizing their concentration) were as effective as conspecific trails. Small workers were less likely to follow a trail of a given concentration than large workers. In the discrimination test, small workers preferred the conspecific trail over the heterospecific trail, whereas large workers showed no significant preference. It is suggested that large workers primarily respond to the releaser component present in both trails, whereas small workers focus more on the conspecific traits provided by the blend of components contained in the trail pheromone.

Thermal radiation as a learned orientation cue in leaf-cutting ants (Atta vollenweideri).

We explored the ability of leaf-cutting ants (Atta vollenweideri) to learn the location of a food reward by using thermal information as an orientation cue. During training of single workers, the conditioned stimulus was a distant thermal source placed frontally, 15 mm away from a platform having a leaf fragment as reward. After training, single workers were confronted with the choice between two sides, one being coupled, in a pseudo-randomized design, with a thermal stimulus heated 5 degrees C above environmental temperature. After 10 learning trials, workers significantly chose the side with the thermal stimulus. This showed that workers can use thermal information for spatial orientation in the context of foraging, which may help them to locate, for instance, highly attractive sun-exposed leaves. Thermal radiation alone as orientation cue was sufficient to allow learning, since preclusion of thermal convection during training and test did not impair workers' response. Shielding of both thorax and gaster from the thermal source did not weaken learning, suggesting the sole participation of head and antennae in temperature reception. A thermal stimulus heated 1 degrees C above environmental temperature could not be used as a learned orientation cue, even when foragers were allowed to directly contact the thermal source.

A macroglomerulus in the antennal lobe of leaf-cutting ant workers and its possible functional significance.

Ants have a well-developed olfactory system, and pheromone communication is essential for regulating social life within their colonies. We compared the organization of primary olfactory centers (antennal lobes, ALs) in the brain of two closely related species of leaf-cutting ants (Atta vollenweideri, Atta sexdens). Both species express a striking size polymorphism associated with polyethism. We discovered that the ALs of large workers contain a substantially enlarged glomerulus (macroglomerulus, MG) at the entrance of the antennal nerve. This is the first description of an MG in non-sexual individuals of an insect. The location of the MG is laterally reversed in the two species, and workers of different size express a disproportional allometry of glomerular volumes. While ALs of large workers contain an MG, glomeruli in small workers are all similar in size. We further compared electroantennogram (EAG) responses to two common trail pheromone components of leaf-cutting ants: 4-methylpyrrol-2-carboxylate and 2-ethyl-3,6-dimethylpyrazine. At high concentrations the ratio of the EAG signals to 2-ethyl-3,6-dimethylpyrazine versus 4-methylpyrrol-2-carboxylate was significantly smaller in A. vollenweideri compared with the ratio of EAG signals to the same two components in A. sexdens. The differences in EAG signals and the species specific MG location in large workers provide correlative evidence that the MG may be involved in the detection of the trail pheromone.