A taxonomic review of the genus Myrmelachista (Hymenoptera: Formicidae) in Costa Rica. Longino, J. T. Zootaxa 2006 1141 1 54 5WNS urn:lsid:biosci.ohio-state.edu:osuc_concepts:24785 Insecta Formicidae Myrmelachista CoL Animalia Myrmelachista Roger Hymenoptera 49 Arthropoda genus  The above are all the species and subspecies known from Caribbean islands. My knowledge of the Caribbean fauna is limited, but I have examined 20 collections of ramulorumfrom Puerto Rico, St. Croix, USA (Florida, possibly introduced and then extirpated, see Deyrup 2003), Santo Domingo, St. Thomas, and the Dominican Republic; three collections of rogerifrom Cuba; syntypes of rogeri mannifrom Cuba, and syntypes of rogeri rubricepsfrom Cuba. All appear to be similar to plebecula. All are bicolored or various shades of red brown. Unlike plebecula, all have long erect setae projecting from the sides of the head (workers and queens). Eight queens of ramulorumare very small with very narrow, rectangular heads. The largest of these have the narrowest heads, with HW around 0.70mm and CI around 74, a combination not found in any Costa Rican species except for the one small longiceps-like queen described under longiceps. Unlike ramulorum, the longiceps-like queen lacks erect setae on the sides of the head. The smallest ramulorumqueens and the queens of rogeriare in the same size range as plebeculaqueens, but with relatively narrow heads. All measured queens of plebeculahave CI 85 or greater. The highest CI among the ramulorumand rogeriqueens is 82. Myrmelachista kraatzii, ambigua, rogeri, and ramulorumare all older names than plebecula, and if plebeculaproves to be an allopatric variant of a widespread polytypic Caribbean species it will no doubt be a synonym of one of these older names. It is not clear that rogeriand ramulorumare distinct. When two Cuban rogeriqueens I have measured are compared to eight ramulorumqueens from other islands, they are at the small end of a continuum of measurements.  Myrmelachista ambiguawas described from a single worker from St. Vincent. Given the relative uniformity of workers, the published description and even examination of the type will be of little use. Queen and male-associated collections of Myrmelachistafrom St. Vincent will be needed to compare with material from other parts of the Caribbean. Wheeler (1908) considered the worker of ramulorumclose to ambigua.  Myrmelachista gagates, from Haiti, was described as being close to rogeribut solid black. It will be important to examine multiple collections of Myrmelachistafrom Cuba, to ascertain whether there are multiple sympatric species there. It is unknown whether kraatziiand the forms of rogeriare distinct or represent one variable species. Myrmelachista kraatziifrom Cuba and M. nigellafrom Venezuela are the two oldest names in the genus, kraatziibeing a 9-segmented form and nigellaa 10-segmented form. Thus kraatziiwould have priority among all the 9-segmented forms. One collection from El Yunque, Puerto Rico, is indistinguishable firom M. longiceps. It is a collection of workers and alate queens, collected by Juan Torres. I am reluctant to identify it as longicepsuntil more Puerto Rican material is obtained, but there is a large size gap between the queen of this El Yunque collection and the various queens of ramulorumfrom elsewhere in Puerto Rico. From these observations it is clear that more collections are needed from the Caribbean to better understand species boundaries in this group.