RESUMO
BACKGROUND: Analyses of spatial and temporal patterns and interactions are important for determining the abiotic factors limiting populations and the impact from other species and different anthropogenic stressors that promote the extirpation of species. The fish Hubbsina turneri de Buen (1940) was studied as a model species in a historical context at varying locations. Originally distributed only in the Lerma-Chapala basin, the main lake complex in Mexico, this species has not been collected from Lake Cuitzeo (LC) and now is restricted to Lake Zacapu (LZ). At present, the Highland splitfin is classified as critically endangered. METHODS: Historical information of LC and historical and current information from LZ were explored by applying cluster analysis and generalized additive mixed model (GAMM) to describe the biotic interactions among fish species and the relationship between density and environmental variables, respectively. The two lakes' contrasting abiotic/biotic characteristics provided elements to describe some species distribution limits in chemical ion gradients. Extirpation calendar dates were estimated using an optimal linear estimation method. Finally, a bibliographic review was conducted on the causes that promoted the extirpation and restriction of H. turneri and the prognosis for its reestablishment and conservation. RESULTS: Clusters showed the fishes relationship according to their distribution along the lakes. GAMM indicated that high H. turneri density is related to low hardness/fecal coliforms, medium depth/suspended solids, and high oxygen concentration. Estimated extirpation dates were between the years 2013 and 2018. The extirpation was linked to an abrupt drop in the LC volume, water quality degradation, increased biotic interactions within macrophytes habitats with native and introduced species, and fisheries bycatches. The current restricted range of H. turneri resulted from the draining of the larger lake, forcing the remaining populations to small spring-fed remnants. Recent samplings in LZ resulted in a low number of individuals. CONCLUSIONS: The integration of ecological interactions derived from statistical models, extirpation dates from nonparametric tests, and the exhaustive analysis of historical information can be applied to define the current situation of imperiled, ecologically relevant species, in different aquatic ecosystems. We are confident that this general framework is important for determining (1) the requirements and limitations of populations regarding abiotic variables, (2) biotic interactions (trophic and spatial) with native and introduced species, and (3) different anthropogenic stressors within and around the ecosystem. This knowledge will also allow understanding those aspects that contribute to the extirpation of populations and could help the restoration of the habitat and the reintroduction of extirpated species.