Biogeographia – The Journal of Integrative Biogeography

Following the Apennines: updating the distribution of Formica clara and Formica rufibarbis in Italy (Hymenoptera,


INTRODUCTION
Formica clara Forel, 1886, with its huge Eurosiberian range, appears to be one of the most widespread species of its genus (Seifert 2018). Yet, despite this and its relatively early description (Forel, 1886), its taxonomy has remained confused until 12 years ago, when 2 Seifert & Schultz (2009) redescribed it and defined a reliable set of diagnostic characters (also see Seifert 2018). Currently, the nominal form is thought to occur from Western Europe (westernmost limit in Spain, see Arcos 2020) east to Central Asia, south to the Himalayas and north to Scandinavia, while the subspecies F. clara sinae Emery, 1925, probably requiring further taxonomic investigation, inhabits eastern China (Seifert & Schultz 2009). The rarity of F. clara in both France and Spain as well as its absence from the British Isles suggests postglacial spreading from a SE European refuge (Seifert 2018). Eventual nomenclatural changes concerning F. clara would imply a different understanding of its relationship with other Asian species, but the status of the European F. clara as a different species from F. rufibarbis is undisputed (Czechowski et al. 2012).
However, before the revision published by Seifert & Schultz (2009), F. clara was broadly confused with F. rufibarbis Fabricius, 1793 and probably even F. cunicularia Latreille, 1798. Presently, F. rufibarbis is thought to occur at least from the Iberian Peninsula to West Siberia, showing less thermophilic attitudes in Central Europe where it occurs sympatrically and appears more common than F. clara (Seifert 2017;2018).
In Italy, the presence of F. clara has long gone unnoticed due to confusion with F. rufibarbis, which is the sole of the two names to be present in the existing -and rather outdated -Italian checklists (Baroni Urbani 1971;Poldi et al. 1995). In particular, 23 records of F. rufibarbis were published between 1834and 1969(Baroni Urbani 1971, recording this taxon from the Alps to the southernmost part of the Italian peninsula (Aspromonte, Calabria) in addition to Sardinia (Fig. 1). These records are all unreliable under the contemporary F. rufibarbis definition and may as well refer to misidentified F. clara, which was first officially recorded in Italy by Rigato & Toni 2011). During the last decade, some new data were published, starting to clarify what the true distribution ranges of F. clara and F. rufibarbis in Italy may be. Formica clara was recorded from the Alps (Glaser et al., 2012), the Po Plain (Castracani et al. 2020) and Sardinia (there probably replacing F. rufibarbis entirely, see Rigato & Toni 2011;Schifani et al. 2021a). On the other hand, F. rufibarbis was only found in the Alps (Glaser et al. 2012). Dubious data under "Formica prope clara" were also published from Salento (Apulia, SE Italy) (Scupola 2016; Fig.  1). However, according to the most recent discriminant functions (Seifert 2018) these Apulian specimens key out as F. cunicularia (Antonio Scupola, pers. comm.).
Here we report and discuss new distribution records of F. clara and F. rufibarbis from Italy, significantly extending their ascertained distribution range and clarifying their occurrence throughout the Apennines.

MATERIALS AND METHODS
Ant collection was performed by direct sampling. Specimens were stored under 96% ethanol (apart from a single older sample collected in 1993 and kept under 70% ethanol) and deposited in the authors' personal collections (E. Schifani pers. coll., Palermo, Italy -ESPI; A. Alicata pers. coll., Catania, Italy -AACI; M. Menchetti pers. coll., Barcelona, Spain -MMBS). Their study was performed with the aid of a stereoscopic microscopes at 45-180x magnification, in addition to photographybased morphometry. Species-level identification was achieved based on the keys provided by Seifert & Schultz (2009) and Seifert (2018).

RESULTS
We identified F. clara from 8 sites across 5 Italian regions (Abruzzo, Calabria, Emilia-Romagna, Sicily, Tuscany) and F. rufibarbis from 4 sites and 3 regions (Emilia-Romagna, Tuscany, Veneto). The southernmost Formica clara record is 37.9° N in Sicily, while that of F. rufibarbis is 44.0° N in Tuscany (Figs. 2, 3). Almost all sites of both species are from hilly to montane from 660 to 1,470 m asl, with the exception of the two Po Plain sites (Emilia-Romagna) which are lowland sites at ≤ 50 m asl. Nests were found in meadows neighboring streams, forests or urban habitats. Detailed data are given in Tab. 1. 4 Figure. 3. Formica clara worker from the southernmost site of its Italian distribution in Sicily: lateral view (above) and head view (below). Table 1. New records of Formica clara and F. rufibarbis in Italy. Altitude data for samples whose GPS error is >10 m are preceded by the symbol "~" and indicatively report the altitude corresponding to the given latitude and longitude without taking into account the error.

Species
Region

DISCUSSION
Our new data considerably extend the geographical range of F. clara along the peninsula by about 900 km, demonstrating its presence through the Apennines south to Sicily (Fig. 3). The verified range of F. rufibarbis in Italy (here referring to Glaser et al. 2012) is also expanded, by about 225 km, as its presence is recorded south of the Alps for the first time. Our F. clara sites in Tuscany and Calabria are very close to historical sites of F. rufibarbis (Baroni Urbani 1964;Zangheri 1969). On the other hand, the discovery of F. clara in Sicily represents the first record from the island for any of the two species (Schifani et al. 2021b). Our results may suggest that F. clara extends through the Central and Southern Apennines, while F. rufibarbis is restricted to more northern latitudes. Such biogeographic differences could be reasonable in light of the different ecological niches the two species occupy across their sympatric Central European range (Seifert 2017;2018;Wagner 2020). It is notable that Seifert (2017;2018) describes F. cunicularia as a less thermophilic species of F. clara in Central Europe, while in Italy the first is not only more common but also found at much lower latitudes even in the warmest regions (see for example recent data from Sicily by Schär et al. 2020). Data from Greece and Poland also suggest that F. cunicularia may be adapted to a wider range of environmental conditions compared to F. clara (Lech Borowiec, pers. comm.).
The results we presented show how unreliable old distribution data have become as ant taxonomy evolved in the last decades. The distribution range of F. rufibarbis was greatly overestimated while the presence of F. clara remained unrecognized for a long time (Baroni Urbani 1971). Unfortunately, lack of distribution data verified under modern taxonomy is a common issue for Italian ants, well beyond the case examined in this study, with the distribution of many diverse and ecologically important genera such as Lasius strongly requiring new investigation (e.g. Schifani et al. 2021b).