The drivers of intraspecific niche variation and its effects on species interactions are still unclear, especially in species-rich Neotropical environments. Here, we investigated how ecological opportunity and interspecific competition affect the degree of individual trophic specialization and the population niche breadth in tetra fish. We studied the four ecologically similar species (Psalidodon aff. gymnodontus, P. aff. paranae, P. bifasciatus, and Bryconamericus ikaa) in subtropical headwater streams (three sites with two co-occurring species and three sites with only one species). We sampled fish in two contrasting seasons (winter/dry and summer/wet), and quantified their trophic niches using gut content analysis. Psalidodon bifasciatus was the only species distributed over all the sampled streams. We observed seasonal differences in population trophic niche breadth of P. bifasciatus just when this species co-occurred with P. aff. gymnodontus. These findings confirm the complex nature of the effects of interspecific competition, depending, for instance, on the identity of the competitor. The degree of individual specialization of P. bifasciatus was higher in the winter, and it was not influenced by the presence of another species. Conversely, the other two Psalidodon species studied presented greater individual specialization in the summer, when fish consumed a higher proportion of allochthonous items (terrestrial insects and seeds), and there were no effects only for B. ikaa. Herein, our results suggest that seasonality in food-resource availability is a major driver of niche variation and it has the potential to play an important role in how these similar tetra species interact and coexist.

We sampled six headwater streams (2^nd and 3^th orders – Figure S1, Table S1) and collected all small characids species found: Psalidodon aff. paranae (Eigenmann, 1914), Psalidodon bifasciatus (Garavello & Sampaio, 2010), Psalidodon aff. gymnodontus (Eigenmann, 1911) and Bryconamericus ikaa Casciotta, Almirón & Azpelicueta, 2004. We sampled fish in July (winter) and December (summer) 2017. According to the historical climate series of the region, 2017 represents a typical year with cold and dry winter, and hot and rainy summer . In each season, we realized one sampling event per stream. We conducted the fish sampling in reaches of 50 m using electrofishing with three passes of 40 min. After capture, specimens were anaesthetized, fixed in 10% formalin, and preserved in 70% alcohol. We collected fish with authorization from the Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio) (license number 25039-1) and approved by the Ethics Committee on Animal Use of the Universidade Federal do Rio Grande do Sul (CEUA – 32734) in accordance with protocols in their ethical and methodological aspects for the use of fish.

We identified specimens in the laboratory according to specific identification keys (Baumgartner et al., 2012; Ota, Deprá, Graça & Pavanelli, 2018). To avoid bias caused by ontogenetic variations, we selected 403 adult individuals (Table S2). Voucher specimens were deposited in the fish collection of the Departamento de Zoologia, Universidade Federal do Rio Grande do Sul. We analyzed the stomach contents under stereoscopic and optical microscope and feeding items were identified to the lowest possible taxonomic level, using specific literature for the algae (Bicudo & Bicudo, 1970) and invertebrates (Mugnai, Nessimian & Baptista, 2010). We quantified the proportion of feeding items in each individual’s diet following the volumetric method (Hyslop, 1980) using graduated test tubes and a glass counting plate (Hellawell & Abel, 1971).