Magosternarchus is a genus of weakly electric knifefish in the family Apteronotidae, containing two species. They are endemic to Brazil, occurring in large river channels in the Amazon River basin. Both species are unusual benthic predators that specialize in biting off the tails of other knifefishes, and are characterized by their greatly enlarged jaws and teeth. Recent systematic studies indicate that both species should be included in Sternarchella instead of being placed in their own genus.
GENUS
via GBIF
Magosternarchus is a genus of weakly electric knifefish in the family Apteronotidae, containing two species. They are endemic to Brazil, occurring in large river channels in the Amazon River basin. Both species are unusual benthic predators that specialize in biting off the tails of other knifefishes, and are characterized by their greatly enlarged jaws and teeth. Recent systematic studies indicate that both species should be included in Sternarchella instead of being placed in their own genus.
==Taxonomy and etymology== The name Magosternarchus honors Dr. Francisco Mago Leccia, who has made many contributions to the study of gymnotiform knifefishes; the latter part of the name sternarchus is from the Greek sternon ("chest") and archos ("rectum"), referring to the forward placement of the urogenital opening in this group of fishes. The species name of M. raptor is from the Latin for "plunderer", referring to its tail-eating habits; the species name of M. duccis refers to the Duke University Center for Creative Ichthyology (DUCCIS), an ichthyology club. Based on several shared skeletal traits, the closest relative of Magnosternarchus is believed to be the genus Sternarchella. Several studies using both morphological and genetic data, published in 2013 and later, have strongly suggested that the genus should be merged into Sternarchella (making Magosternarchus a junior synonym) and this has been followed by the Catalog of Fishes. Nevertheless, FishBase continues to recognize the two genera as separate.
Discovered by embedding cosine similarity (sentence-transformers MiniLM, 384-dim).