A NEW SPECIES OF CHIBCHANOMYS 



127 



Table 2. Matrix showing character state distributions among 12 



ichthyomyine species (for details see text and Voss, 1988). The character 

 state assessments for all taxa other than C. orcesi were taken directly 

 from Voss (1988 Table 45, page 441). Character states for the new 

 species were assessed by PJ except for those of the visceral and 

 reproductive systems (characters 15-18) which were unobservable in 

 the new taxon, so scored as '?'. 



















Characters 



















1 



2 



3 



4 



5 



6 



7 



8 



9 



10 



11 



12 



13 



14 



15 



16 



17 



18 



Taxa 







































Ale 



1 



1 







1 



1 







1 



1 



1 











1 







1 



1 



1 







1 



Ctr 



1 



1 







1 



1 









1 



1 















1 



1 



















Cor 



1 



1 











1 









1 



1 























? 



7 



? 



7 



Ihy 







1 





















1 







1 



1 



2 







1 







1 



1 







Ipi 







1 





















1 







1 



1 



2 







1 







1 



1 







Itw 







1 





















1 







1 



1 



2 







1 







1 



1 







Nmo 



1 





























1 







































Nve 

































1 















1 























Rme 







1 



1 



1 



1 





2 



1 



1 











2 







1 



1 



1 







2 



Rha 







1 

















1 



1 















2 







1 







1 







2 



Rtt 







1 

















1 



1 















2 







1 







1 







2 



Run 







1 



1 



1 



1 





2 



1 















2 







1 







1 







2 



choice has been avoided in the current study, since it is aimed at 

 determining the level of affinity of the new taxon to other 

 ichthyomyines, rather than seeking to add any new dimension to the 

 phylogenetic status of the ichthyomyines as a group. Instead an 

 hypothetical outgroup was constructed in which all character states 

 were assessed as primitive, which was used to root the trees. 



Using branch and bound algorithms, a search was made of the 

 character data summarised in Table 2. In the analysis in which all 

 character states were unordered, the length of the shortest tree was 

 equal to 32 character state transformations and six trees were 

 retained. In each of the other analyses (with the multistate characters 

 7 and 12 ordered, and character 18 varyingly ordered) the length of 

 the shortest tree was equal to 33 character state transformations but 



only three trees were retained. The variation in treatment of charac- 

 ter 18, was not considered to be particularly important in this study, 

 since character state 2 is exhibited only by taxa of the genus 

 Rheomys. In both of the latter analyses C. orcesi and C. trichotis are 

 non-monophyletic in all three trees and also in the semistrict consen- 

 sus of these trees (see Fig. 3). The only evidence of a monophyletic 

 generic grouping shown in the semistrict consensus tree is for 

 Ichthyomys, and this tree is similar in most respects to the most 

 parsimonius hypothesis of ichthyomyine relationships shown by 

 Voss (1988: Fig.88). 



DISCUSSION 



There is obvious conflict in that the results of the phylogenetic 

 analysis do not support the generic classification currently in use. 

 The morphological data is sufficiently persuasive to conclude that, 

 on the available material, the new taxon is correctly attributed to the 

 genus Chibchanomys as currently construed. 



Acknowledgements. We are grateful to Jim and Theresa Clare for 

 unpublished information on Chibchanomys and to the late Professor Gustavo 

 Orces for support and encouragement during fieldwork. We are indebted to 

 Robert Voss (American Museum of Natural History) who initially recognised 

 the uniqueness of the new species and who provided much helpful advice 

 when commenting on drafts of the manuscript. Particular gratitude is owed to 

 Darrell Siebert (Natural History Museum) for guidance with the phylogenetic 

 analysis, and for constructive comments and criticism of the manuscript. 

 Mark Hafner (Museum of Zoology of Louisiana State University) kindly 

 provided photographs and measurements of the specimen of C. trichotis from 

 Peru. As always the staff of the Natural History Museum freely provided 

 support; in particular we thank Richard Sabin (Mammal Group) and Deryck 

 Jones (Electronics) for the X-rays, Phil Hurst (Photographic Unit) for the 

 photographs, while Clive Moncrieff (Biometrics) patiently assisted with the 

 PAUP analysis. 



Ale 

 Ihy 

 Itw 

 Ipi 

 Rha 

 Rtt 

 Rme 

 Run 

 Ctr 

 Cor 

 Nmo 

 Nve 



Fig. 3 Semistrict consensus tree showing hypothetical phylogenetic relationship of Chibchanomys orcesi to other taxa of ichthyomyine rodents. 

 Consistency Index 0.636, Retention Index 0.786. 



