Contamination and chimerism are perpetuating the legend of the snake-eating cow with twisted horns ( Pseudonovibos spiralis). A case study of the pitfalls of ancient DNA

Abstract

In 1994, several unusual horns collected from markets in Vietnam and Cambodia were described as a new genus and species of wild bovid, Pseudonovibos spiralis (Peter and Feiler, 1994a,b). The horn sheaths possessed a unique and distinct lyriform twist, as well as annulations throughout their entire length. This animal, named ‘‘linh duong’’ in Vietnam or ‘‘kting voar’’ in Cambodia (but see Brandt et al., 2001), was known only from detached horn sheaths or frontlets (partial frontal bones with bony horn cores) with horn sheaths. This lack of anatomical information was largely responsible for the confusion regarding the taxonomic status of this species, with morphologists debating whether it was a close relative of the tribe Antilopini (gazelles) (Peter and Feiler, 1994a), Caprini sensu lato (goats, sheep, and allies) (Nadler, 1997), or Bovini (oxen, bison, and buffaloes) (Dioli, 1997; Timm and Brandt, 2001). Given both the paucity and variable taxonomic interpretation of the morphological data, hopes were initially high that DNA sequence would resolve the issue of Pseudonovibos taxonomic status and phylogenetic position. However, three ensuing DNA sequencing studies using different putative P. spiralis specimens gave rise to mutually incompatible hypotheses: (1) Hammer et al. (1999), using a 415-bp DNA fragment of the mitochondrial cytochrome b gene, proposed affinities with the Caprini sensu lato. However, Hassanin and Douzery (2000) challenged the authenticity of the Hammer et al. (1999) sequence and interpreted it as the result of DNA contamination from chamois (Rupicapra) in the laboratory. (2) Hassanin et al. (2001) revealed that some horns assigned to P. spiralis are simply cow horns that had been artificially carved and twisted. Two DNA markers were sequenced from four trophies of P. spiralis collected in Indochina during 1925: a 243-bp fragment of the mitochondrial cytochrome b and a 327-bp fragment of the nuclear lactoferrin gene. The phylogenetic results showed that the enigmatic horns of the linh duong belonged to domestic cattle (Bos taurus) (Hassanin et al., 2001). Morphological inspection indicated that horn sheaths, originally smooth, were carved to create the annulations, while the twist in the upper part of the horns was made by artificial torsion (Thomas et al., 2001). This raised the question of whether all horns of the linh duong are fraudulent or not. In other words, did the species P. spiralis ever really exist (see review in Brandt et al., 2001)? (3) Most recently, Kuznetsov et al. (2001) suggested that P. spiralis was a new species of buffalo based on a 962-bp fragment of the mitochondrial 12S rRNA gene. However, one of us (Hassanin, 2002) demonstrated that the putative sequence of P. spiralis was a chimera obtained from three different species: B. taurus, Bubalus bubalus (domesticated Asian water buffalo), and Saiga tatarica (saiga antelope). In addition, several factors indicated that their specimen was artificially made using horns and a frontlet from domestic cattle, B. taurus. The ongoing and increasingly heated controversy surrounding P. spiralis has been covered in Science (Malakoff, 2001), Nature (Whitfield, 2002), and the New York Times (Mydans, 2002), with emphasis on both the folkloric portrayals of a snake-eating cow-like creature Molecular Phylogenetics and Evolution 27 (2003) 545–548