The five segmented gaster (in yellow, arabic numeral) in a Cerapachys nitidulus worker. Roman numerals refer to abdominal segments (Scanning Electron Micrograph, Roberto Keller/AMNH)

Gaster is a morphological term that is very useful and yet imprecise for the purpose of comparative anatomy as it is currently used in ants. It comes from the Greek for “belly” and it refers to the collection of segments in the metasoma that remain after the pedicel of ants and wasps. It is the bulbous part of the body that hosts the insect viscera.

The problem when applying this term comes from the fact that ants can have a pedicel or waist composed of one or two metasomal segments (either abdominal segment II or II-III), and so, depending on the group, the ant gaster is a belly made up of either four or five external segments.

Four segmented gaster (in yellow, arabic numerals) in a Manica rubida worker. Roman numerals refers to abdominal segments (Scanning Electron Micrograph, Roberto Keller/AMNH)

It may seem harmless to say that formicine ants, for example, have a five segmented gaster while myrmicine ants (see image above) have a four segmented one. The imprecision arises, however, whenever there is a reference to an individual gastral segment. The first segment of the gaster in a carpenter ant is not homologous with the first segment of the gaster in a fire ant. The former correspond to the third abdominal segment (III) while the latter refers to the forth (IV).  This peculiarity is not only the result of defining the tagma by what segments it excludes (that is, metasoma minus the segments of the pedicel), but by the fact that the excluded segments are anterior to it, and hence there is a shift in anatomical reading frame.

Note that the shift is purely nominal, there is no insertion/deletion of body segments along the insect axis between the different group of ants.

Luckily, authors are consistent when referring to gastral segment on a given publication, even if they don’t define gaster beforehand, so there is no internal ambiguity. The headache problem comes when one is compiling information from different publications, by different authors across all ant groups. Without context, reading that something is or occurs in gastral segment number X is ambiguous until you scan the rest of the publication for an explanation. Moreover, it may seem obvious that in groups in which the pedicel is clearly two segmented (e.g., Myrmicinae, Pseudomyrmicinae) the second gastral segment will correspond to the Vth abdominal for example. But the author may have decided to stick to a conservative terminology and consider that the gaster starts after the petiole (abdominal II) regardless of how modified the third abdominal segment is. And I won’t even go into groups that have genera with one and two segmented pedicels (the New World army ants, Ecitoninae) or groups in which it is difficult to decided if the pedicel has one or two segments (Cerapachinae, see the image of Cerapachys above).

This brings us to the topic of cybertaxonomy. We humans can scan a publication and resolve an ambiguity in anatomical terminology. Machines can’t (or at least not easily). This is the sort of challenges faced by initiatives like Plazi that aim to translate legacy data into digital format.

Moral of the story? Gaster is fine when referring to the whole section of the ant’s body. For individual segments, in comparative anatomy, it is always better to stick with the abdominal segmentation number.