The Formicariidae, formicariids, or ground antbirds are a family of smallish passerine birds of subtropical and tropical Central and South America. They are between 10 and 20 cm (4-8 in) in length, and are related to the antbirds, Thamnophilidae, and gnateaters,
Conopophagidae. This family contains probably (see below) some 100–120 species in 1 or 2 large and a number of fairly small genera.
These are forest birds that tend to feed on insects at or near the ground since many are specialist ant eaters. Most are drab in appearance with shades of (rusty) brown, black, and white being their dominant tones. Compared to other birds that specialize in following ants, this family is the most tied to the ground. The long, powerful legs (which lend the birds a distinctive upright posture) and an essentially vestigial tail aid this lifestyle.
They lay two or three eggs in a nest in a tree, both sexes incubating.
The Formicariidae are traditionally divided into two groups, which are supported by a wide range of morphological, molecular and other evidence. The antthrushes in the genera Formicarius and Chamaeza are similar in appearance to small rails. Their sexes are alike in plumage, and they walk like starlings. The thrush part of the name refers only to the similarity in size (and in Chamaeza also coloration) to true thrushes, not to an evolutionary relationship.
The antpittas in the genera Grallaria, Hylopezus, Myrmothera and Grallaricula are also sexually monomorphic; they resemble the true pittas in that they are virtually tailess; they hop like some thrushes, and are much easier to hear than see – although their vocalizations may be rather atypical for perching birds.
Recent research (Irestedt et al. 2002, Rice 2005a,b) indicates that the Formicariidae as previously delimited are highly paraphyletic, judging from comparison of several mt and nDNA sequences. The aberrant bar-bellied "antpittas" of the genus Pittasoma, which were formerly placed here, belong to the gnateater family (which initially was also considered part of the Formicariidae); as the gnateaters proper, they are sexually dichromatic. On the other hand, at least a large proportion of the Rhinocryptidae (tapaculos), including the type genus Rhinocrypta, seem to be closer to the antthrushes than these are to the antpittas.
Very little molecular data is currently available for rhinocryptids, which are moreover rather non-diagnostic morphologically. As relationships with related groups such as ovenbirds have not been fully resolved, it is certain that the systematics of the group will be revised soon, but it is not clear how exactly the new systematic lineup will be. On the other hand, morphological, behavioral and molecular analyses of the true antpittas agree astoundingly well as regards their internal systematics, considering the high amount of lifestyle-related homoplasies in the birds discussed here.
Barring a closer relationship of any of the three groups – antthrushes, tapaculos, and antpittas – to other suboscines, three basic approaches are possible:
Some or all Rhinocryptidae would be merged into the Formicariidae. "Rhinocryptidae" would have to be abolished as a family-level taxon, and if that group is not entirely merged into the Formicariidae, the remaining genera would receive a new family name. The Formicariidae would then contain an antthrush-tapaculo subfamily, possibly with two separate tribes, and one or two antpitta tribes as per option 2 in another subfamily.
It is alternatively possible to split the group into two families, of which one would contain the antthrushes and some or all tapaculos and their allies in two subfamilies and bear the name Formicariidae. The other family would contain the antpittas and receive a new name; it would also contain two subfamilies, one consisting of the typical Grallaria antpittas and the other of the usually smaller species of the remaining genera.
Finally, all three major lineages – antthrushes, tapaculos and allies, and antpittas – could be considered separate families. Phylogenetically, it might even be deemed appropriate to split the antpittas into two families (one containing only Grallaria and the other the remaining genera) though this would probably be considered oversplitting.
Altogether, a layout with two subfamilies and four families (option 1) or an analogous one with two tribes and four subfamilies (option 2) seems to represent the phylogeny of these birds best; a three-lineage approach (option 3) might be warranted if some other families are also part of the Formicariidae-Rhinocryptidae clade. Whether the typical antthrushes and tapaculos should be separated from the antpittas at the family level is not yet clear. It depends on whether or not the gnateaters and true antbirds are closer to the true antpittas than to the antthrush-tapaculo clade; there is some – albeit very tentative – indication that splitting off the true antpittas as a distinct family (option 2) is indeed more warranted (Rice 2005a).