Scientific data laundering: Chimeric mitogenomes of a sparrowhawk and a nightjar covered-up by forged phylogenies

Abstract

In 2017, Gang Liu and colleagues published a paper in this journal in which they described the first mitochondrial genome of Japanese Sparrowhawk Accipiter gularis (Aves, Accipitriformes). The paper included two phylogenies (one based on full mitogenomes, the other on mitochondrial 12S/16S rRNA sequences) that placed this species among the goshawks and sparrowhawks (genus Accipiter). These phylogenies also included a previously published mitogenome of Grey Nightjar Caprimulgus indicus. Here, we show that the published mitogenome of A. gularis is a chimera of three species (a sparrowhawk, a buzzard and a dove), and that the published mitogenome of C. indicus is a chimera of two owl species. Phylogenetic re-analysis showed substantial differences from those published by Liu et al. (2017): the A. gularis sequence was placed among Streptopelia doves in the 12S/16S rRNA tree, and the C. indicus sequence was placed among Otus owls in the mitogenomic and mitochondrial 12S/16S rRNA trees. The phylogenies in Liu et al. (2017) showed several other peculiarities: (i) the inclusion of Javan Hawk-eagle Nisaetus bartelsi in the mitogenome tree (whereas no mitogenome is available of this species and no such sequence has been described in the literature), (ii) reciprocal monophyly of the genera Buteo, Leucopternis and Buteogallus (whereas multiple previous studies have shown these to be para- or polyphyletic), and (iii) the placement of Black-chested Buzzard Eagle Geranoaetus melanoleucus and Crane Hawk ‘Geranoaetus’ (=Geranospiza) caerulescens as sister species (whereas all previous studies have shown these to be distantly related). Close examination of the figures in Liu et al. (2017) revealed that both phylogenies have been manually prepared or modified. The evidence indicates that Liu et al. (2017) published phylogenies that were not based on existing data but were fabricated to reflect preconceived ideas about phylogenetic relationships. The phylogenies published by Liu et al. (2017) give the impression that the mitogenomic sequence described in their paper (A. gularis) and another one (C. indicus) described by Zhao et al. (2016) are authentic and correctly identified, whereas in reality these sequences are highly problematic chimeras. The term ‘scientific data laundering’ is proposed for the practice of hiding poor or erroneous research data with fabricated or falsified figures.