Sumário As aves das matas do rio Sopotuba, Mato Grosso, foram estudadas durante uma "friagem" no mês de julho de 1975. As aves insetívoras foram menos conspícuas e as aves frugívoras mais conspícuas durante a "friagem". É possível que em "desastres" ocasionais como a "friagem", nas margens de florestas tropicais, haja seleção contra aves insetívoras e seleção favorecendo as aves "onívoras", isto é, que comem insetos e frutas. A tendência para se ter menos Passeriformes "suboscines" nas margens de florestas neotropicais parece ser devido à falta de suboscines do grupo de "furnarioides", que são insetívoros. As teorias de que os suboscines são ancestrais de "oscines", e que os oscines entraram na América do Sul no Plioceno, parecem improváveis, pois se assim fosse, teria faltado oscines comendo sementes por várias épocas geológicas.

Abstract The avifauna of the forested upper reaches of the Sopotuba River, in the upper Paraguay drainage just south of the Serra dos Parecis, proved a mixture of central-Amazonian birds from just east of the Madeira River and of birds of dry forests (Xingú-Tapajóz) or lower Andean forests peripheral to the southern Amazonian region. A cold wave that passed during the study period caused a major drop in insect activity and in the activity of insect-eating birds and hummingbirds. Fruit-eating birds and "omnivores" (insect and fruit eaters) became relatively conspicuous. It is suggested that occasional "disasters" of the type of a cold wave inhibit specialized insectivores at this and other margins of Amazonian forests, and encourage higher percentages of omnivores. The previously noted tendency for oscines to replace suboscines at various types of neotropical forest margins is reinterpreted as primarily due to declines of furnarioid suboscines (mostly forest-living insectivores) and their replacement by omnivorous tyrannoid suboscines and by omnivorous oscines. The conventional theory that furnarioid suboscines are primitive and that they and other neotropical suboscines are being pushed into forests by recently immigrating oscines leads to two ecological anomalies: virtual absence of omnivorous passerines in the early Cenozoic and absence of seed-eating passerines for much of the late Cenozoic. As alternate theory, that of separation of forest-edge suboscines by continental drill followed by evolution of forest-dwelling furnarioid suboscines in the New World, with open-country oscines either separated by the drift or entering South America early in the Cenozoic, removes both anomalies.