DYSITHAMNUS AND THAMNOMANES (AVES, FORMICARIIDAE) AS ARMY ANT FOLLOWERS

Antvireos (Dysithamnus mentalis and D. puncticeps, Formicariidae), which glean in moderately open understory foliage, and derived antcatchers (Thamno-manes ardesiacus and T. caesius), which sally to open understory foliage, follow ground-living army ants little; probably the ants flush little prey in understory foliage, and irregularly move through zones too dense for the antvireos and antcatchers. Antbirds (Formicariidae) often follow army ant swarms for flushed arthropods in neotropical forests (Willis and Oniki, 1978). Here, in the seventeenth of a series of reports on occasional ant followers, I report on "antvireos" of the genus Dysithamnus and related "antcatchers" of the genus Thamnomanes. Taxonomically, species of these genera have been confused because they are similar morphologically but differ in behavior; Dysithamnus hop in moderately open understory foliage while Thamnomanes wait and sally like flycatchers in even more open understory foliage. Thamnomanes are probably derived from Dysithamnus, for T. ardesiacus is morphologically intermediate. On discovering in 1961 that ardesiacus sallies for flying insects like Thamnomanes caesius, I suggested to R. Meyer de Schauensee that ardesiacus be put in Thamnomanes. E. Eisenmann (pers. comm.) then recommended that several other species (saturninus, schistogynus, occidentalis, and plumbeus) also be transferred to Thamnomanes, and Meyer de Schauensee (1966) concurred. While saturninus and schistogynus sally for insects like flycatchers, and hence belong in Thamnomanes, P. Schwartz (pers

Antbirds (Formicariidae) often follow army ant swarms for flushed arthropods in neotropical forests (Willis and Oniki, 1978).Here, in the seventeenth of a series of reports on occasional ant followers, I report on "antvireos" of the genus Dysithamnus and related "antcatchers" of the genus Thamnomanes.
Taxonomically, species of these genera have been confused because they are similar morphologically but differ in behavior; Dysithamnus hop in moderately open understory foliage while Thamnomanes wait and sally like flycatchers in even more open understory foliage.Thamnomanes are probably derived from Dysithamnus, for T. ardesiacus is morphologically intermediate.On discovering in 1961 that ardesiacus sallies for flying insects like Thamnomanes caesius, I suggested to R. Meyer de Schauensee that ardesiacus be put in Thamnomanes.E. Eisenmann (pers.comm.) then recommended that several other species (saturninus, schistogynus, occidentalis, and plumbeus) also be transferred to Thamnomanes, and Meyer de Schauensee (1966) concurred.While saturninus and schistogynus sally for insects like flycatchers, and hence belong in Thamnomanes, P. Schwartz (pers.comm.) and I soon discovered that plumbeus gleans in foliage and should have been left in the genus Dysithamnus.The behavior of occidentalis is unknown, but morphologically it seems like plumbeus and hence is likely to be a Dysithamnus.
The behaviorally unknown species "Xenornis" setifrons also seems like plumbeus and perhaps should be placed in Dysithamnus.The large-billed Thamnistes anabatinus of fairly dense foliage in forest midlevels links slender--billed Dysithamnus of open understory foliage and large-billed Thamnophilus "antshrikes" of dense foliage, and perhaps should be placed in one or the Department of Biology.University of Miami.Coral Gables, Florida 33124, USA.Present address: Departamento de Zoologia.UNESP.13.500, Rio Claro, SP. other of the two last genera.In color, it is apparently a mimic of similarly foraging ovenbirds (Furnariidae) and vireos (Vireonidae).Unusual coloration may justify retention of the genus Thamnistes for it.

Results
1. Dysithamnus mentalis (Plain Antvireo) briefly followed army ants on Bush-bush Island, Nariva Swamp, Trinidad (I), at Paraiso, 1100 m, Caqueta, Colombia (1 pair), Zatzayacu, 800 m, E. Ecuador (1), Bacabal, Maranhao, Brazil (1 pair), Boraceia (1), and Fazenda Barreiro Rico, Anhembi, Sao Paulo (5 raids, 6 birds).All raids were of the ant Eciton burchelli except at Bacabal and 2 raids at Barreiro Rico, which were of Labidus praedator.The antvireos wandered past raids rapidly, 2 m up on horizontal slender twigs in moderately open foliage, and rarely descended to 0.4-0.8m up to stare at the ants.One that descended at Barreiro Rico was immediately supplanted by an ant-following tanager (Trichothraupis melanops) and abandoned the ants.At Bacabal, the pair gave loud kimp calls at an ant-following raptor, Micrastur ruficollis, and then deserted the ants.The "song" there was an accelerating weep, weep, weep, we-we-e-e-e-e-i-i-i-oo-oo. Young call chee-che-che as they follow adults, with records 14 Aug (1976, Barreiro Rico) to 18 May (1975, Campinas) in southern Brazil, mostly Dec-Feb (5 families).D. mentalis is a species of the fairly open understory of montane, swamp, semiarid, island, and dry woodlands; it avoids dark and species-rich terra fir me Amazonian lowland forests, where Thamnomanes species occur.A color-banded male (14.1 g) near Campinas, Brazil stayed on a territory at least 19 mo.(1976)(1977)(1978).It readily joins mixed forest flocks, although it also occurs away from them as long as there are dense tangles to hide in.Rarity of ant following may be due to its requiring moderately open foliage for rapid gleaning, while the ants move mainly on the ground through foliage of varying densities; also, competing birds and raptors around ants may discourage it.Gochfeld and Tudor (1978) also record mentalis following ants.
2. Dysithamnus puncticeps (Spot-crowned Antvireo) briefly followed 7 swarms of burchelli (1 with 3 birds, 3 with 2) on Barro Colorado Island, Panama.It wandered around raids in fairly open foliage 1-6 m up, rarely descending to 0.5 m up to look at ants, and normally arrived and departed with the mixed-species forest-interior flocks of antwrens.One puncticeps sallied to catch a small grasshopper, another pecked a small spider off a twig.Twice a Slaty Antshrike (Thamnophilus atrinuchus) supplanted one.At times two male puncticeps, probably of adjoining territories, chase and pivot to display ruffed-up heads, sleeked bodies, and white eyes at each other, giving chat-at and ee kee KWEEP notes and ending with downscale trilling "songs".Pairs usually forage near each other, giving faint weet or weet-it "chirps", in open foliage and relatively open undergrowth.N. Brokaw and D. Schemske (pers.comm.)netted it "in" treefalls, but I suspect that it found the open lanes they cleared to put their nets in the treefalls perfect new sites for its short-distance examination of fairly open foliage.Birds of this genus, unlike Thamnophilus antshrikes, sally and glean in open foliage around rather than in treefalls or other dense foliage.Ants rarely flush insects to open foliage, so that antvireos find little food over ants.
Normally a flycatching species of the understory in mixed-species flocks, it sometimes leaves the flocks to stay near ants for 2-3 hours.At Reserva Ducke, it perched 0-6 m up (records 27-29-12-8-3-2 by 1-m intervals) on slender (records 30-9-2 from 0-3 cm diameter, by 1-cm intervals; 1 record on a perch just under 8 cm diameter) twigs that were as often horizontal as vertical (records were 17-5-4-5-16 by 20° intervals from 0° to 100° from the horizontal).Frequently it perches on palm leaflets, a behavior pattern noted for no other species.Its pose is less upright and flycatcherlike than in T. caesius, but ardesiacus seldom hops like an antvireo.Usually it sallies short distances (up to 2 m) for prey: records from all regions were of 19 prey on foliage.8 in the air. 1 on debris, 2 on trunks.3 on the ground, and / on above-ground sites (mostly foliage or air near foliage).One prey was pecked off a leaf, one off a trunk.At Reserva Ducke, ardesiacus sometimes joined various species of woodcreepers sallying after winged ants fleeing from understory nests on the approach of burchelli.One ardesiacus flew in when an antbird (Gymnopithys rufigula) tried sallying to a leaf 2 m up, and one supplanted a rufigula that was eating prey on the ground.Another rufigula supplanted an ardesiacus, while Myrmeciza fortis supplanted one at Benjamin Constant, and Rhegmatorhina hoffmannsi supplanted it at Coata.Usually ardesiacus wandered from one perch to another in fairly open undergrowth, avoiding the dense ground saplings used by other ant-following species.At times it sallied around a recent treefall, and zones of overhanging palm fronds were favorite sites, but it did not enter lianas or other dense foliage that would hinder sallying.
Normally caesius sally for arthropods in mixed-species flocks in the open forest midlevels, 5-15 m up, and follow ants only as long as mixed flocks wander nearby.Only ants in open areas of the understory attract caesius, which never works in dense zones.Over ants in various regions, it sallied to the air (15 records), foliage (8), stems (6), trunk (1), ground (2), or undetermined sites near or on foliage (6); it usually flew short distances for captures, except when diving after falling prey.It rarely hops along a bough to glean prey.Perch heights over ants were mostly 1-7 m up, although a few individuals at Palhao and Itubera moved down to 0.3-0.5 m up.Most perches are horizontal (15-5-3-1-4 records from 0 o -100 o with the horizontal, by 20° intervals) and slender (15-6-6 records from 0-3 cm, by 1-cm intervals), and the species sits upright like a small flycatcher.Other species rarely supplant it; I recorded one supplanting by an antbird (Hypocnemis hypoxantha) near its nest at Mitu.
At Mitu, two young caesius -one changing from brown to the gray adult male plumage -followed a pair around ants 3-4 May 1966.The pair fed the noisily "squeaking" young at times, and young "grunted" ch-ch-ch often, but they were getting much of their own food.Nov 26, 1973, a male dissecting prey on the ground at Manaus fed a female that "chirped" cheu and squeaked, Nov 15-19, 1974, a male fed an adult female at Itubera; she gave chirps followed by rough chrahhh sounds.
At Maloquinha, an adult male fed another (probably young) 18 Sep 1974.Nesting from Manaus and Belem south was in the southern-hemisphere spring and summer, that at Mitu in the northern spring.At Maloquinha, a female weighed 18.1 g and had a cloacal temperature of 43.0 o C.
The noisy alarm of this species at a human, hawk, or even alarm calls of distant fruit-crows (Querula purpurata, at Belem), is spectacular: several loud whert notes (or where chep, at Belem) are followed by a "chipping" chi-chi-chi series as the bird flicks its wings outward, its spread tail upward, and pivots one way and then the other or flees behind cover.The "song", in territorial disputes at Umbria (Colombia) and Maloquinha, was an accelerating downscale, wheezy wheesp, wheesp, wheesp whe-whe-we-we-we-e-e-i-i-i-o-o-oo-oor, introduced at Maloquinha by a few upscale notes (whoor whee whee).
Like drongos (Dicrurus spp.) in African forest flocks, caesius keeps other species alarmed by its loud calls, and probably captures insects flushed by nervous other birds.However, it is relatively quiet once it is habituated to an observer, and then captures insects flushed by the normal movements of other species in the forest flocks.It follows ants infrequently even compared to ardesiacus, probably because caesius exploits the very open and high understory where the ground ants flush little prey.

Discussion
Dysithamnus antvireos glean rapidly, moving in relatively open foliage in the forest understory and occasionally sallying for prey.They thus contrast with most Thamnophilus antshrikes, which glean in dense lianas and foliage, but both groups are equally hindered in following army ants -the ants move irregularly from zones of dense foliage to zones of open foliage, so that a bird specializing on a given foliage density can follow the ants only irregularly.In the case of antvireos and antcatchers, the ants also flush little prey on open sprays of foliage, preferring instead tree trunks and the ground.
Antvireos show preadaptations for open foliage and for flycatching, extended in the probably derived Thamnomanes antcatchers to flycatcher-like behavior in very open forest understory.Like antvireos, the antcatchers find prey density in their foraging sites little increased by ants, and find that the ants often move to sites too dense for effective flycatching.Low-level flycatching to foliage (T.ardesiacus) is evidently more productive over ants than is high aerial sallying (T.caesius), for ardesiacus follows ants more often and persistently; but neither Thamnomanes follows ants for more than a small fraction of its food.