Aim: One of the most consistent global biogeographic patterns is the latitudinal diversity gradient where species richness peaks within the equatorial tropics and decreases towards the poles. Here, we explore the global biogeography of socially parasitic ants, which comprises the most diverse group of social parasites in the Hymenoptera. We test the biogeographic hypothesis that ant social parasites are distributed along an inverse latitudinal diversity gradient by peaking in diversity outside of the equatorial tropics.

Location: Global

Taxon: Ants (Hymenoptera: Formicidae)

Methods: We assembled a comprehensive biogeographic dataset for all 371 taxonomically described socially parasitic ant species. We used phylogenetic and taxonomic studies to compare species richness with the number of species representing independent evolutionary origins of social parasitism across a latitudinal gradient. In addition, we compared ant social parasite diversity across biogeographic regions using rarefaction to account for different sampling efforts. Finally, we tested for a correlation between latitude and the proportion of ant social parasite species within regional ant faunae.

Results: The geographic distribution records and the inferred 91 origins of socially parasitic life histories show that both species richness and the number of species representing independent evolutionary origins of social parasitism peak in the northern hemisphere outside of the equatorial tropics. Based on rarefaction curves, northern latitude regions harbour the most ant social parasite species, but the diversity of independent evolutionary origins is not significantly different between northern and southern hemispheres. The proportion of ant social parasite species within regional faunae is correlated with latitude only in the northern hemisphere.

Main conclusions: The inverse latitudinal diversity gradient of ant social parasites contrasts with the biogeographic pattern observed in free-living, non-parasitic ant species and appears to be driven by large species radiations as well as by the presence of specialized life histories exclusive to the northern hemisphere.

We assembled a biogeographic dataset consisting of 6001 occurrence records for all 371 taxonomically described socially parasitic ant species. Geographic occurrence records were assembled from the Global Ant Biodiversity database (GABI) via antmaps.org (Janicki et al., 2016; Guénard et al., 2017), AntWeb (antweb.org), the Museum of Comparative Zoology database (mczbase.mcz.harvard.edu), specimen records in our personal collections, personal communication with colleagues, and the primary literature. For each social parasite species, we gathered occurrence records at each latitudinal and longitudinal degree for every country the parasite was reported from. In the case of countries with greater than 2.78x10⁶ km² land area, which includes Russia, Canada, China, the United States, Brazil, Australia, India, and Argentina, we used occurrence records at each latitudinal degree at the scale of first-level administration to achieve a more detailed biogeographic resolution in our dataset, which is consistent with first-level administration data from GABI. For all occurrence data, we recorded the accuracy based on the precision of the occurrence record: country, first-level administrative division, locality, or GPS.

The raw data and metadata used in this study are in an Excel Workbook format. To use the raw data one should export only the "raw_data" sheet of the Excel Workbook. Note: the raw data includes special characters such as (?) and spaces that were filtered out during analyses.