John T. Longino, The Evergreen State College, Olympia WA 98505 USA.
23 April 1997
Brown (1959) observed that the highlands of Costa Rica produced melanic variants of many dacetonine species. My own observations of many ants of forest leaf litter, and not just dacetonines, confirm a general pattern. Within a set of morphologically very similar specimens (similar enough to be proposed as one species), specimens from montane forests tend to be larger and darker than their lowland counterparts. But "species" for which I have abundant material along an elevational transect allow me to examine patterns in greater detail. I have not observed cases of gradual, continuous variation from lowlands to highlands. Patterns I observe are (1) variation is continuous, but the transition from the lowland form to the upland form is concentrated in a narrow elevational band, (2) variation is discontinuous, and the switch from the lowland form to the upland form occurs in a narrow elevational band, (3) variation is discontinuous, and the lowland and upland forms have an elevationally broad zone of sympatry, and (4) variation is continuous, but the intermediate forms are rare, and the overall pattern is of broad sympatry of discrete forms.
In these Web pages on the Costa Rican fauna, I may discuss such elevational patterns of variation. When variation appears discontinuous and/or sympatric, I may treat the forms as distinct "species," even though existing and future taxonomic revisions may treat them as one species. Also, the patterns I discuss are usually not well-substantiated, because they emerge from collecting that was not designed to capture or quantify the patterns, and because my examination of the specimens may have been fairly cursory. As specimens accumulate from various collecting efforts, the perceived pattern may change. Proposed discontinuous variation may change to continuous when intermediates are found. Proposed parapatry may change to sympatry when discrete forms are found together at a greater range of sites.
My purpose is not to quibble over the definitions of species, but to present what I know about geographic variation of particular ant lineages.
Geographic patterns such as these are relevant to the study of clines and speciation. Something is causing correlated patterns of morphology across many ant lineages. There must be selection for larger, darker workers at higher elevations, and the counteracting forces of selection and gene flow are playing themselves out independently in hundreds of ant lineages. For a discussion of the evolutionary ecology of clines and their relevance to taxonomy and species definitions, see Mallet (1995) and related papers.
Brown, W. L., Jr. 1959. A revision of the Dacetine ant genus Neostruma. Breviora 107:1-13.
Mallet, J. 1995. A species definition for the modern synthesis. Trends in Ecology and Evolution 10:294-299.