RESUMO
Resumen La microcuenca del río Anchique se ubica en el municipio de Natagaima (Tolima) y drena directamente al río Magdalena, la distribución y abundancia de los peces está influenciada por múltiples factores a escalas temporales y espaciales, el objetivo del estudio es describir la estructura y composición de la comunidad íctica a lo largo del río y probar la hipótesis de que estas comunidades varían a nivel temporal y espacial. Se realizaron dos muestreos en altas y bajas lluvias, y se establecieron cuatro estaciones, colectándose 625 ejemplares y 36 especies. No se encontraron diferencias estadísticamente significativas entre algunas variables ambientales del río a nivel temporal, según la prueba Kruskal-wallis. Los valores de los índices de complejidad estructural y composición del sustrato, muestran que la estación E4 presentó los valores más altos, además fue la estación con mayor riqueza y diversidad según los números de Hill. El análisis de correspondencia canónica evidenció que el sustrato, complejidad y ancho del cauce influyeron de forma significativa en la composición y distribución de los peces. La comunidad íctica del río no presentó variación de la diversidad y riqueza temporalmente, pero si espacialmente, lo cual podría deberse a que cada estación presentó características estructurales diferentes.
Abstract The distribution and abundance of fish is influenced by multiple factors at temporal and spatial scales, and their community composition represent good river integrity indicators. The main objective of this study was to describe the composition and structure of fish assemblages along the Anchique river sub-basin (that drains to Magdalena river), and to test the hypothesis that these communities vary temporally and spatially. For this, four sites (E1, E2, E3 and E4) were sampled along the river (from river birthplace to Magdalena river), during the periods of high and low waters, in 2014. In each site, longitudinal transects (100 m long x 2 m wide, 60 min) were established and fishing was made with both electrofishing and trawl (10 m x 1 m, 0.05 m) nets. Fish samples were fixed and transported to the laboratory for posterior taxonomic identification with the use of keys; fish diversity and structure composition were calculated with the use of standard models. Besides, substrate composition and environmental variables were also considered. A total of 652 fish belonging to 36 species were recorded; high water showed higher values (325 fish with 32 species) than low water period (300 fish with 31 species). Siluriformes was dominant (51 %), followed by Characiformes (32 %), Gymnotiformes (5 %), Cyprinidontiformes and Cichliformes (6 %). Most abundant families were Characidae (28.3 %; 9 spp.), Loricariidae (23.4 %; 8 spp.) and Trichomycteridae (15.0 %; 2 spp.). Kruskal-Wallis test indicated that there is no significant difference among some environmental variables at the temporal level. The values of structural complexity and substrate composition indices showed that E4 had the highest values (0.8), followed by E1 (0.6), and E2 and E3 (0.4). Moreover, E4 had the highest richness and diversity, according to Hill's number. The canonical correspondence analysis showed that the substrate, and the complexity and width of the river channel, had a significant influence on fish composition and distribution. The fish community richness and diversity did not change temporarily, but did spatially, which could be related to the different structural features of sampling sites. Rev. Biol. Trop. 66(1): 336-351. Epub 2018 March 01.