New records of ants (Hymenoptera: Formicidae) for Colombia

Abstract Even though Colombia has high levels of ant species richness in the Neotropical region, this richness continues to increase. New records of the ant subfamilies Amblyoponinae, Dolichoderinae, Dorylinae, Myrmicinae, and Ponerinae are presented. Two species of Fulakora, two species of Azteca, one species of Cylindromyrmex, 25 species of Myrmicinae belonging to 12 genera (Acanthognathus, Basiceros, Daceton, Eurhopalothrix, Hylomyrma, Mycetomoellerius, Mycetophylax, Mycocepurus, Octostruma, Pheidole, Rogeria, and Talaridris), and one species of Leptogenys are registered for the first time for Colombia. Five species are new records for South America. For each species, the geographical distance of the record closest to the Colombian locality is offered. Several factors, such as access to previously unexplored conserved areas, sampling techniques that cover heterogeneous microhabitats such as leaf litter, and many more taxonomic researches have allowed the knowledge of ant fauna in Colombia to continue growing.


Introduction
Globally, 337 valid genera and 13,809 valid species of ants are known (Bolton 2020). With 105 genera and more than 1100 species, the ant fauna in Colombia is one of the richest in the World (Fernández et al. 2019). Some ant genera, however, are poorly represented in Colombia (e.g., Fulakora Mann, 1919) while others, such as Pheidole Westwood, 1839 exhibit high diversity (Fernández et al. 2019). Other taxonomic groups, such as the subfamily Martialinae Rabeling &Verhaagh, 2008, andthe genus Bothriomyrmex Emery, 1869, have yet to be found there. Because of this, Colombia ranks as the second most diverse Neotropical country, with a slightly lower ant genus richness than Brazil (105 vs 112 genera, respectively). The Colombian diversity, however, is relatively larger than Costa Rica's, whose myrmecofauna is much better known (Guerrero et al. 2018).
Recent field trips and routine curatorial activities in some major insect collections are helping to increase the knowledge of the ant fauna of this country. The recent ceasefire in Colombia has allowed field activities in areas that were previously too dangerous or areas under the protection of the National Parks Unit (Guerrero et al. 2018). This access to previously unsampled areas as the Colombian Amazon and forests in the Sierra Nevada de Santa Marta (northeastern Colombia) has allowed us to find 31 species of ants that we are registering for the first time for Colombia, and five of those ones are new records for South America.
The distance between the record in Colombia and the one in the nearest country was calculated with https://www.gps-coordinates. net/distance by calculating the straight line spacing between the geographic coordinates of both records. When a species has more than one record for Colombia, the Colombian record closest to the record in the neighboring country was used. The closest record to Colombia was extracted from electronic resources, such as antmaps. org (Janicki et al. 2016) and www.antweb.org, and in some cases from the original descriptions (e.g., Lattke, 1992). In the case of Antweb information, we record the specimen identifier (e.g., the FMNHINS0000095938 specimen of Eurhopalothrix schmidti (Menozzi, 1936) is the closest record to Colombia). In all cases, the distance is presented in kilometers.

Taxa richness
We identified 31 species of ants not reported from Colombia. These species are distributed among five subfamilies and 16 genera. The subfamily that contained the largest number of new records was Myrmicinae, with 12 genera, of which Pheidole with ten species, was the richest. The other subfamilies had only one or two new recorded species. We found only one genus in each of the following subfamilies: Amblyoponinae, Dolichoderinae, Dorylinae and Ponerinae.

Checklist of new ant records for Colombia
Comments. This species was previously known from Panama (Wilson 2003). The present record extends its distribution 979 Km to the South from the closest record in Panama, reported by Wilson (2003). This is the first record of Pheidole kuna for South America. Comments. Pheidole socrates was previously known from Bolívia, Brazil, Ecuador, French Guiana, Guyana, Lesser Antilles, Mexico, Panama, Suriname, and Trinidad and Tobago (Forel 1912, Fernández & Sendoya, 2004, LaPolla & Cover 2005, Mamani-Mamani et al. 2012, Guenard et al. 2017. The present record is located 201 Km to the North from the closest record in Ecuador, reported in antmaps.org (Guénard et al. 2017).

Comments.
Pheidole midas was previously known from Brazil, Ecuador, French Guiana, Lesser Antilles, Panama, Peru and Venezuela (Alonso et al. 2001, Basset et al. 2012, Fichaux et al. 2019, Filho et al. 2003, Prado et al. 2019, Wilson 2003. The record from San Juan Nepomuceno is located 526 Km to the Southeast from the closest record in Panama reported by Guénard et al. (2017).
Pheidole midas was previously identified as P. veletis Wilson, 2003, as both species are differentiated by very subtle traits that can lead to misidentification. However, Pheidole midas can be differentiated by the dorsal surface of the head that is largely smooth and relatively shining, with notable foveas and a large-celled patch of rugoreticulum present to the side and behind each antennal fossa (Wilson 2003), while P. veletis with the dorsal surface of the head dull and rough, and carinulae originating on frontal lobes extend halfway between level of eye and occiput.
Comments. Pheidole wallacei was previously known from Brazil and French Guiana (Mann 1916, Groc et al. 2009). The present record is located about 1000 Km to the Northwest from the closest record in Rondonia, Brazil (Mann 1916).

Discussion
In the Neotropical region, the number of genera and species currently known is 131 and 3463, respectively (Fernández et al. 2019), and approximately 81% of these genera are registered in Colombia (Fernández et al. 2019). The thirty-one new ants registered here raise the number of specific taxa known from Colombia to 1197 species. This species richness represents more than one third (~34%) of the known Neotropical species.
In the "Hormigas de Colombia" book, Fernández et al. (2019) reported 50 genera and 535 Myrmicinae species for Colombia, but a recent phylogenetic work added two more genera, Mycetomoellerius Solomon et al., 2019 andParatrachymyrmex Solomon et al., 2019 replacing the former genus Trachymyrmex Forel, 1893 (Solomon et al. 2019). The thirty-one new records we provide here increase the Myrmicinae composition to 560 species. This species richness is relatively higher than that of countries such as Costa Rica (535 native species sensu antmaps.org (Janicki et al. 2016) or 539 according to Antweb), whose forests have been widely sampled resulting in a much better-known ant fauna. Within the Myrmicinae, the genus Pheidole is the most species-rich, with more than 1100 species throughout the world. Currently, there are 620 species of Pheidole known for the Neotropical region, with just two invasive species, Pheidole indica Mayr, 1879 and Pheidole megacephala (Fabricius, 1793). After Guerrero et al (2018), the species richness of Pheidole was 118 species, including Pheidole indica (there reported for the first time for Colombia) and Pheidole megacephala (previously reported by Chacón de Ulloa & Achury 2011). Here, we provide ten new species records elevating the species richness of Pheidole in Colombia up to 128 species. The species richness of Pheidole in Colombia is almost half of the specific richness reported for the Mesoamerican wet forests (Longino 2019), however, the records for Colombia could equal or exceed those numbers, due to the study of Pheidole specimens coming from different types of forests, both from lowlands (e.g., tropical dry forest and sub-xerophytic formations) as Andean and sub-paramun forests.
Among the thirty-one new ants for Colombia, we found species whose updated distributions may be the basis for present or future biogeographic questions. We recorded for the first time five species that were previously known from Central America, Azteca quadraticeps, Azteca snellingi, Eurhopalothrix schmidti, and Eurhopalothrix xibalba and Pheidole kuna. The two Azteca species extend their distribution to the northernmost area of South America, both with populations inhabiting the lowlands of the Sierra Nevada de Santa Marta (northern Colombia).
In general, Azteca species have great dispersal capacity, which for both cases could have allowed them to reach a wide distribution dispersing through the Darien mountain range and the favorable habitat that extends along the Caribbean plain, until reaching the lowlands of the mountainous massif. In the case of the Eurhopalothrix species registered here, both present an Andean distribution, one in the eastern Cordillera (Eurhopalothrix schmidti) and the other in the central Cordillera (Eurhopalothrix xibalba). Eurhopalothrix schmidti was collected at 2000 m of altitude, matching the altitude distribution in Mesoamerica that ranges from 1100 m to 2200 m, while Eurhopalothrix xibalba was collected at high elevations, above 1700 m within the altitudinal range of this species in Central America, from 50 m to above 1600 m (Longino 2013a). On the other hand, Pheidole kuna, spreading through the Chocoan biogeographic region, adding to other arthropods from the Chocó-Darién province (Morrone 2014).
The records of Fulakora agostii in Colombia expands its distribution from Brazil to almost 4500 km north-west of South America. The first records of Fulakora agostii come from the Brazilian Atlantic Forest (Lacau & Delabie 2002), however, the Fulakora agostii population in Colombia comes from the Andean landscape at 1500 m altitude. The Colombian Andean and the Brazilian Atlantic forest populations are allopatric, with a large discontinuity in its distribution, as currently there are no records in the Amazon basin, despite the abundant sampling in the Brazilian Amazon. Several hypotheses may explain the current distribution range of this species, but only two will be outlined. First, its occurrence in those two kinds of forests could suggest wide ecological plasticity to adapt to contrasting habitats. The second hypothesis would take into account the taxonomic validity of the Andean population, in this sense, this Andean population could correspond to an undescribed different evolutionary lineage. Although the diagnostic characteristics in this population match completely those offered by Lacau & Delabie (2002) it is necessary to carry out molecular analysis at the population level in F. agostii to try and differentiate between these hypotheses.
Recently, Guerrero et al. (2018) indicated that the growing knowledge of Colombian ant diversity may be due to factors such as the possibility of sampling in forests under the protection of the National Parks Unit, using sampling techniques that produce large volumes of biological material (e.g. Winkler extractors). However, the possibility of sampling in areas that were previously in armed conflict has also had a positive effect on myrmecological studies in the country. In this case, 50% of the new reports here come from forested areas where the armed conflict has completely ceased (e.g., all Octostruma species and five species of Pheidole recorded here); the other 50% correspond to ants that were collected in forests protected by the network of national parks or inside indigenous reserves. These factors expose the importance of collecting in unsampled areas for 1) better resolution of the distribution of biota and to support stronger biogeographic hypotheses, 2) access to more populations of known species to have a clearer picture of their genetic variability, tokogenetic relationships of these populations (i.e., phylogeography) and understanding the phenotypic variability of some species, and 3) increase records within national biological collections to help complement biodiversity inventories and decision-making by government entities at different spatial scales.

Acknowledgments
Special thanks to Dr. Rodrigo Feitosa for his comments on the early version of the manuscript. Thanks to John Lattke for reviewing the English and comments to improve the quality of the manuscript. Many thanks to two anonymous reviewers for the invaluable comments and corrections that improved the scientific quality of this contribution. To Dr. Dimitri Forero and Dr. John Cesar Neita for their help in facilitating the access to the ants deposited in the MPUJ and IAvH collections. This is contribution No. 14 in the series Ants of Colombia. This article is part of the master's work of the first author: Taxonomy and Distribution of Pheidole ants (Formicidae: Myrmicinae) in Colombia. We thank the Office of "Editorial Unimagdalena" for the financial support to pay the publication fees. This is scientific contribution No.17 from the Colecciones Biológicas at the Universidad del Magdalena (CBUMAG).

Author Contributions
Emira I. García contributed in the preparation of the manuscript, as well as to the identification of several ant species and the collection of distribution data from museums and biological collections. Also, she critically reviewed the different versions of this manuscript, incorporating arguments and hypotheses contained in the latter.
María C. Tocora contributed to the preparation of the manuscript, as well as to the identification of ant species and the collection of distribution data from museums and biological collections. She reviewed and corrected the different versions of this manuscript.
Gian P. Fiorentino contributed to the preparation of the manuscript, as well as to the identification of ant species and the collection of distribution data from museums and biological collections. He reviewed and corrected the different versions of this manuscript.
Fernando Fernández contributed to the preparation of the manuscript, as well as to the identification of ant species and the collection of distribution data from museums and biological collections. He reviewed and corrected the different versions of this manuscript.
Mayron E. Escárraga contributed in the preparation of the manuscript, as well as in the identification of some ant species. He was responsible for managing electronic resources to calculate geographic distances and to refine the geographic coordinates of all records; he critically revised the first draft and the final version of the manuscript.
Roberto J. Guerrero contributed substantially in the preparation of the manuscript, as well as to the identification of several ant species and the collection of distribution data from museums and biological collections. He wrote and added intellectual arguments and hypotheses contained in the manuscript; moreover, critically reviewed and corrected the different versions of this one.