Although both interspecific competition and coexistence mechanisms are central to ecological and evolutionary theory, past empirical studies have generally focused on simple (two-species) communities over short time periods. Experimental tests of these species interactions are challenging in complex study systems. Moreover, a number of studies of ‘imperfect generalists’, consistent with Liem’s Paradox, raise questions about the ability of evolved species differences to effectively partition niche space when resources vary considerably across the annual cycle. Here we used a recently developed theoretical framework to combine past research on population-level processes with observational data on resource use to test for ongoing interspecific competition and understand the nature of resource overlap. We compared species diet overlaps and differences in several distinctive communities centred on a focal species, the American Redstart Setophaga ruticilla replicated both spatially and seasonally, in combination with documentation of population regulation to assess the ability of similar species to partition dietary niche space and limit interspecific competition. Our results document high dietary overlap in most of the communities studied, with only subtle differentiation consistent with known species-differences in foraging behaviour and morphology. These findings are largely consistent with species foraging as imperfect generalists. However, in contrast to past studies, the high diet overlaps observed here during times of inferred resource scarcity were driven by low value prey taxa (e.g. small ants) and did not involve truly ‘private’ resources. All of these factors increase the potential negative impacts of interspecific competition and limit the ability of these birds to avoid competition if conditions further deteriorate, either seasonally or at longer intervals.

We sampled insectivorous bird species in two diffusely competitive communities. In summer, redstarts and six other species were sampled in seasonally flooded, bottomland hardwood forest sites at Bogue Chitto National Wildlife Refuge and Honey Island Swamp State Park in southern Louisiana, 22 May - 6 July 2006. Both locations are protected, relatively natural bottomland hardwood forests, containing small, sluggish streams, cypress-tupelo swamp, and less inundation-tolerant hardwood species at slightly higher elevations. The other abundant insectivorous bird species studied in this site were Acadian Flycatcher Empidonax virescens, Hooded Warbler Setophaga citrina, Kentucky Warbler Geothlypis formosa, Prothonotary Warbler Protonotaria citrea, Swainson’s Warbler Limnothlypis swainsonii, and White-eyed Vireo Vireo griseus. These species comprise all of the most common small insectivorous birds at this site, and for which we could obtain at least seven stomach samples.

We sampled wintering redstarts along with four other species of arboreal, insectivorous warblers in Jamaican wet-limestone forest: Black-throated Blue Warbler Setophaga caerulescens, Black-and-white Warbler Mniotilta varia, Northern Parula Setophaga americana, and Worm-eating Warbler Helmitheros vermivorum. These five species were the only migrant, arboreal insectivores common in the habitat. This habitat is native, historically widespread, floristically diverse, and structurally complex, with diverse tree types including canopy emergents, dense ground cover and shrubs, vines and vine tangles, and epiphytes including bromeliads and orchids. The first of these sites was located on a cattle ranch near Betheltown in Westmoreland Parish (15 February - 15 March 2017). The second site was located near the Windsor Research Station in Trelawney Parish (16 - 31 March 2017), within Cockpit Country.

We included two systems more similar to classical competition studies, in which a pair of species exhibits interspecific aggression. In both communities, redstarts were subordinate to the other competitor. Breeding redstart and Least Flycatcher Empidonax minimus diets were sampled at a northern hardwood forest site in the Hubbard Brook Experimental Forest, NH, primarily in spring 1982, but additional data for redstarts were available from 18 May - 1 July 1981-95. Hubbard Brook is in the White Mountain National Forest, and is dominated by northern hardwoods—predominantly American beech Fagus grandifolia, sugar maple Acer saccharum, and yellow birch Betula alleghaniensis (Sherry & Holmes 1985).

Lastly, we sampled wintering redstarts and non-migratory, socially dominant, Yellow Warblers Setophaga petechia petechia (Sherry et al. 2020, Powell et al. 2021), the two most abundant parulids in mangrove habitat, at the Font Hill Reserve, Luana Point, St. Elizabeth Parish, Jamaica. The site where the stomach samples were obtained is adjacent to the coast and comprised overwhelmingly of black mangroves Avicennia germinans. Sampling took place from 13 January - 14 March 2014.

Field and Lab Methods—The majority of individuals’ diets were sampled during morning hours to help ensure full stomachs resulting from active feeding. Northern hardwood forest and Jamaican black mangrove birds were located by sight and sound and were obtained using shotgun or mist nets. The stomach was removed immediately in the field, perforated, and preserved with its contents in 70% ethanol. We obtained gut samples in the bottomland hardwood forest and Jamaican wet-limestone sites, as in shaded coffee habitat (Sherry et al. 2016), using the emetic antimony potassium tartrate to induce regurgitation (Poulin et al. 1994). Birds were then placed in a waxed paper-lined box for 15 minutes, released, and gut samples retrieved for preservation in 70% ethanol.

Stomach samples were examined in 70% ethanol-filled petri dishes using an Olympus SZH binocular dissecting microscope at 75-650X magnification. Samples were thoroughly searched for diagnostic arthropod fragments (such as heads, elytra, mandibles, legs, and wings). Sclerotized body parts of most arthropods make identification, at least to a certain taxonomic level, straightforward.

Any data on non-animal prey (e.g., fruit/seeds) is not included.