Bull. nat. Hist. Mus. Lond. (Zool.) 61(1): 1-10 



Issued 29 June 1995 



A revised familial classification for certain 

 cirrhitoid genera (Teleostei, Percoidei 

 Cirrhitoidea), with comments on the group 

 monophyly and taxonomic ranking 



S THE NATURAL 

 HISTORY MU^EL 



-7 JUL 1995 



P.H. GREENWOOD 1 



Honorary Research Associate, J.L.B. Smith Institute of Ichthyology, Private Bag 1015, GraharrtSTtn 

 6140, South Africa and Visiting Research Fellow, The Natural History Museum, Cromwell Road, 

 London SW7 5 BD 



PHt:obivic:D 

 ZOOLOGY LIBRA 



CONTENTS 



Introduction 1 



Methods and materials 2 



Urohyal morphology and the suprageneric classification of the cirrhitoid fishes, particularly the genus 



Acantholatris and species currently referred to the genus Nemadactylus 2 



The familial classification of Acantholatris Gill, 1862, and Nemadactylus Richardson, 1839 5 



Possible lineages within the Latridae as now expanded to include the genera Acantholatris and 



Nemadactylus 5 



Taxonomic and phylogenetic conclusions 6 



Taxonomy 6 



The cirrhitoids as a monophyletic lineage 6 



Intragroup relationships within the cirrhitoids 8 



Acknowledgements 9 



References 10 



Synopsis. Previous suprageneric classifications of cirrhitoid fishes were based mainly on superficial characters. 

 Recent anatomical studies show that five morphologically distinct types of urohyal bone occur within the group, and 

 that on this basis certain taxa have been misplaced at both the generic and familial levels. The monophyly of the 

 cirrhitoid fishes, tacitly assumed by previous authors, is reviewed, and an hypothesis of their monophyly erected on 

 the basis of several synapomorphic features. It is also proposed, on the basis of those synapomorphies, that the 

 lineage be given subordinal rank within the Perciformes. 



INTRODUCTION 



During an investigation into the cranial and branchial muscles 

 associated with feeding in certain cirrhitoid fishes from South 

 African waters, it became apparent that five distinct morpho- 

 types of urohyal occur within the group. Two very distinct 

 types are found in genera currently classified as members of 

 the Cheilodactylidae; of these two forms, one also occurs in 

 genera referred to the family Latridae by Regan. Despite the 

 passage of over 80 years since Regan's (1911) paper was 

 published, it remains the most recent suprageneric classifica- 

 tion of these fishes, which he arranged in five families, viz 

 Cirrhitidae, Cheilodactylidae, Aplodactylidae, Chironemidae 

 and Latridae, grouping them informally as the 'cirrhitiform 

 percoids'. 



Like its predecessor, namely Gill's more detailed paper of 

 1862, Regan's later analysis was based mainly on external 



t Dr Greenwood died 3 March 1995. 

 ©The Natural History Museum, 1995 



features. Apart from some comments by Regan on cranial 

 features, vertebral numbers and vertebral morphology, nei- 

 ther paper included any other anatomical information on 

 these fishes. 



Although Regan (1911) expressed some doubts about the 

 familial status given by him to the five generic assemblages he 

 recognised (suggesting that subfamilial rank could be more 

 appropriate) he did not comment on Gill's (1862) earlier 

 classfication which recognised four subfamilies within a single 

 family, the Cirrhitoidae; indeed, and inexplicably, Regan 

 {op. cit.) makes no reference at all to Gill's paper. 



Gill's (op. cit) four subfamilies correspond, almost entirely, 

 to Regan's families, except that Gill's Latridinae included 

 two genera, Nemadactylus Richardson, 1839, and Cheilodac- 

 tylus Lacepede, 1803, which Regan incorporated in his family 

 Cheilodactylidae; Gill's list of included taxa in his concept of 

 the Latridinae also contained two genera not mentioned by 

 Regan, despite their being recognised at that time and still 

 being recognised today (viz Latridopsis Gill, 1862, and Acan- 

 tholatris Gill, 1862). 



