Resumen del producto

¹ Instituto Politécnico Nacional, Centro Interdisciplinario de Ciencias Marinas

Coexistence among sympatric species requires a certain degree of resource partitioning. In the Mexican South Pacific, information regarding the coexistence of Stenella attenuata (SA) and Stenella longirostris (SL) is lacking. Stable isotope analyses (d15N and d13C) were conducted to assess the differences in feeding habits to infer trophic position and amplitude as well as habitat use, based on Bayesian inference and a random forest (RF) classifier. Potential trophic relationships with other species were assessed by Bayesian mixing models. Feeding segregation between species was mainly based on carbon sources (d13C: p(SA > SL) =?100%, RF Gini Impurity = 80%). Moreover, SA (n =?22) presented a broader isotopic niche than that of SL (n =?25; SIBER Bayesian Standard Ellipse areas = 0.91‰2 vs. 0.77‰2) with a 33% overlap, suggesting that SA uses more coastal habitats than SL. The most relevant prey species were the mesopelagic fish Benthosema panamense and the epipelagic fish Hyporhamphus naos (~50%), although B. panamense was more related to SL than SA: p(%SL > %SA) =?76.6%. The trophic positions were 4.0 (SA) and 3.8 (SL). Our results provide evidence of resource partitioning within a potential foraging ground for both dolphin species.

Palabras clave: Bayesian inference, dolphin, foraging ground, machine learning,; Mexican South Pacific, niche partitioning, prey, trophic ecology

Para obtener una copia del documento contacta la personal de la biblioteca a través del correo bibliocicimar{a}ipn.mx