Integrative species delimitation
Integrating molecular and phenotypic data using different methodologies is widely assumed to be the most effective approach for delimiting species (Dayrat 2005; Will et al. 2005; Schlick-Steiner et al. 2010; Pante et al. 2015). Integrative taxonomy provides statistical rigor for species delineation and validation of species, as well as for assignment of specimens to a given species group, improving the detection of cryptic diversity and the inference of relationships among species (Leaché and Fujita 2010; Schlick-Steiner et al. 2010; Edwards and Knowles 2014; Leavitt et al. 2015). This approach allows the identification of concordant patterns of divergence based on different sources of information (e.g. unlinked genetic loci, morphology, behavior), thus revealing cases of full lineage separation, since it is highly unlikely that a coherent pattern of character agreement emerges by chance (Padial et al. 2010).
Our molecular-based analyses using the cox1 , 3RAD and UCEs data sets, together with analysis of CHC variation, yielded strongly congruent evidence for the existence of five different evolutionary lineages among the examined populations of E. ruidum . Our results support the existence of the four species suggested by Aguilar-Velasco et al. (2016) (E. ruidum spp. 1-4) and also a fifth species corresponding to what they suggested to be a hybrid population (E. ruidum 2x3). Four of these species were found in localities situated along the lowlands of the Sierra Madre del Sur in Oaxaca and Guerrero, Mexico. Moreover, the possibility of a sixth species in the area is plausible according to recent chemical, genetic and acoustic studies (Peña-Carrillo et al. 2021a, b).
We consistently recovered the specimens assigned to E. ruidum sp . 1 as a well-differentiated species, regardless of the data set and analyses employed. Our CHC and cox1 -based network showed the members of E. ruidum sp. 1 as a highly divergent cluster whose geographic distribution ranges from southeast Mexico to Ecuador, Colombia and Venezuela in South America, and in the Lesser Antilles in the Caribbean. This network also showed the existence of geographic structure within this species. The name E. ruidum should be applied to the populations assigned to E. ruidum sp. 1 based on the type locality of the species, which was restricted to Colombia by Kugler and Brown (1982), and on their morphological correspondence with the syntypes from this country (Aguilar-Velasco et al. 2016). The 3RAD, UCE and cox1 analyses also recovered the specimens of E. ruidum sp. 2 as a separate species, being composed of two geographically structured clades. One of these clades included specimens from southeast Mexico to Colombia and Ecuador in South America, whereas the other one was represented by specimens from southeast to northern Mexico. The cox1 data set, however, showed a considerable genetic distance between these clusters, suggesting that they could be separate species. This hypothesis is supported by the CHC analyses, which also showed a marked divergence between the two clades.
The remaining three species delimited here correspond to populations from the lowlands of Sierra Madre del Sur in the states of Oaxaca and Guerrero, in southeast Mexico. The two genome-wide data sets and the CHC profiles indicated that E. ruidum sp. 3 and E. ruidum sp. 4 are two closely related species whose geographic distribution is restricted to lowland areas of Oaxaca. Moreover, our results consistently support the hypothesis that the specimens of the putative hybrid population proposed by Aguilar-Velasco et al. (2016) from Pinotepa Nacional, Oaxaca, and those assigned to E. ruidum sp. 3 from the state of Guerrero, actually represent a distinct species. Further comprehensive sampling on foothills along the Pacific coast in southeast Mexico will reveal the actual geographic distribution of these three species. Moreover, taxonomic inferences based on these results should consider the previous name availability within the group.Ectatomma aztecum was described by Emery (1901) based on a single specimen collected in the state of Michoacán, Mexico, but without precise locality; however, it was subsequently regarded as a synonym ofE. ruidum by Kugler and Brown (1982). According to Aguilar-Velasco et al. (2016) the syntype of E. aztecum(CASENT0903841; MSNG, Genoa, Italy) is morphologically similar to the specimens assigned to E. ruidum sp. 3.