Monomorium sahlbergi Emery, 1898 (Formicidae, Hymenoptera): a cryptic globally introduced species

Abstract The discovery in the Netherlands in a shipping container of the ant Monomorium sahlbergi Emery, 1898, a species similar to the invasive pharaoh ant M. pharaonis (Linnaeus, 1758), led to a quest to better define the distribution of this species, which was initially obscure due to uncertain specimen identifications. Here it is shown that M. sahlbergi, like M. pharaonis, is found worldwide, almost certainly as a result of introductions. Including quarantine interceptions, this species is recorded from seven global biogeographic regions, but its established outdoor distribution is currently limited to the tropics and subtropics. Monomorium dichroum Forel, 1902 is here presented as a junior synonym of M. sahlbergisyn. nov. based on morphometric and CO1 analyses.


Introduction
Broadening transport networks and rising demand for commodities have led to increases in alien species worldwide (Hulme 2009), including ants (Suarez et al. 2001;Bertelsmeier et al. 2017). In the Netherlands, for example, a relatively large number of non-native ant species are being recorded owing in part to the shipments of plant material imported into the country (Boer and Vierbergen 2008).
A concerted effort is underway to identify ant species introduced into the Netherlands, whether they are established or found during import inspections. Thus far 120 species have been identified (Boer et al. 2018). Many of these introduced species are poor colonisers and have not been able to establish and/or spread after arriving (Boer and Vierbergen 2008). The actual number of introduced species is almost certainly greater; some specimens are impossible to identify due to a lack of suitable identification keys and uncertainty about the origin of the ants. Limited identification tools and training increase the chances that species names are ascribed incorrectly, especially in the case of closely related species. In this work we describe an example of one invasive species remaining hidden in the guise of another, more common species. The case concerns two closely related species of the genus Monomorium, of which one, M. pharaonis (Linnaeus, 1758), is considered the most notorious pest ant species in the world (Wetterer 2010). In the Netherlands, M. pharaonis is the first recorded tramp ant species; the oldest specimen is dated 1877 (Boer and Vierbergen 2008).
On 2 June 2014, the pest controller A.J.A. Heetman intercepted ants found in a shipping container at a distribution company in the Netherlands and sent them to the first author. The shipping container, filled with glycine for the food industry, came from a chemical plant in Wuyi, Hengshui, Hebei, China. The intercepted ants appeared similar to the well-known and globally common tramp species M. pharaonis, but differed in their black gaster. While trying to identify the specimens, we came across images of identical specimens on AntWeb (http://www.antweb.org), where they were recorded under the provisional name M. pharaonis_nr (CASENT0173275, CASENT0246074) and M. bicolor complex (CASENT0178876).
Further comparison of our specimens with the images from AntWeb convinced us that the ants discovered in the Hebei shipping container were a previously described species, M. dichroum Forel, 1902 (Figs 1-3). Monomorium dichroum was reported as only known from India (type locality) (Imai et al. 1984, Bharti 2015 and China (Guénard and Dunn 2012).
Further exploration of similar species on AntWeb, however, suggested our specimens, and M. dichroum for that matter, were identical to M. sahlbergi Emery, 1898, a little-known species described from Israel. We set out to ascertain the true identity of our specimens and determine whether dichroum and sahlbergi are two distinct species.

Materials and methods
Available descriptions of all Monomorium species occurring in the area between Saudi Arabia in the west and China in the east were consulted. Syntype material of M. dichroum and M. sahlbergi were requested and investigated. Monomorium pharaonis, M. cf. pharaonis, M. nr. pharaonis, and M. bicolor-complex ants identified from the collection of CASC and RMNH were investigated. In total, we examined hundreds of specimens from the Netherlands, France, Germany, Israel, Saudi Arabia, United Arab Emirates, Oman, Yemen, Seychelles, Papua, Nepal, New Zealand, Western Australia, Myanmar, Taiwan, China, Ivory Coast, Cameroon, Madagascar, Indonesia, Panama, Mexico, Trinidad, Netherlands Antilles, and the United States of America.
For morphometrical comparisons, 16 workers of M. pharaonis were examined (all in the collection of Naturalis Biodiversity Center, RMNH). The size and shape characters of these workers were quantified (Table 1) and reported as lengths or indices. All measurements are in millimetres. The numeric characters and abbreviations are defined below.

CL
Maximum cephalic length in median line.

CW
Maximum cephalic width, across eyes.

Omm
Number of ommatidia across the widest diameter of the eye.
The examined specimens in this study are deposited in the following institutions:  ) and Naturalis Biodiversity Center (Leiden, the Netherlands), following the protocol described in Fisher and Smith (2008). All sequences are available at GenBank and Appendix 1. Phylogenetic analyses also included 20 Monomorium sequences from GenBank and two sequences as outgroup (Huberia striata and Podomyrma sp.), see Appendix 1 for sequence details.

CASC
Molecular phylogenetic inference. Sequences were aligned using Geneious 11.1.5 (Biomatters Ltd.). The phylogenetic tree was inferred in MEGA7 using maximum likelihood and 100 bootstrap replicates. Nucleotide substitution model selection and genetic p-distance calculation were also performed using MEGA7 (Kumar and Tamura, 2016). The best fit model selected under the corrected Akaike Information Criteria (AICc) was GTR+G+I.

CO1
The phylogenetic tree recovered sequences of M. dichroum and M. sahlbergi in the same clade (Fig. 5), showing low within-clade genetic distance (1.0%). Genetic distance among sequences previously identified as M. dichroum and M. sahlbergi was also low (1.3%). All M. pharaonis sequences clustered together, showing 0.3% within genetic distance and 16.5% genetic distance between this and the M. dichroum + M. sahlbergi clade.

Morphological comparisons
Monomorium dichroum and M. sahlbergi show similar colouration, especially with regard to the infuscate genae and the light spot on the posterior side of the gaster. Morphometrically, these ants are identical. None of the regression analyses of various morphometrical data, such as cephalic width versus cephalic length, scape length, maximum width of postpetiole, width of postpetiole versus width of petioles, comparisons between the cephalic index versus eye index, versus petiole index, versus scape index, and versus promesonotal index, showed any difference. The number of ommatidia across the widest diameter of the eye was the same. Nor could we find any differences in pilosity and pubescence. The surface sculpturing of the head, mesosoma, nodes, and gaster were the same.      Both Monomorium sahlbergi and M. pharaonis belong to the salomonis group, as defined by Bolton (1987). For a detailed description of M. pharaonis see Heterick (2006). Morphometrically M. sahlbergi is similar to M. pharaonis (Table 1). Compared to the workers of M. pharaonis, 1/4 instead of 2/3 of the first gastral tergite (abdominal segment 4) is light-coloured; the structure of the frontal side of the head is strigulate rather than reticulate; the mesonotal groove is shallower; the pronotum and metanotum are higher than the propodeum in M. pharaonis as opposed to equally high in M. sahlbergi and promesonotal setae are missing on the mesosoma, in M. pharaonis two to six (Figs 1-4). Note that in Monomorium specimens the setae are quite stiff and break easily, thus reducing utility of this character in some specimens. Based on the distribution of other species in the salomonis group, the native distribution would include specimens from the Indomalaya region (Nepal, India, Thailand). Our data came from the following main geographic regions: Palearctic (China, Israel, Netherlands (interception)), Australian (New Zealand, from likely interceptions), Nearctic (USA, in part interceptions), Neotropical (Panama, Galapagos), Afrotropical (Reunion, Madagascar) and Oceania (Hawaii) (Fig. 6).

Discussion
The global distribution of Monomorium sahlbergi suggests a history of introductions. Although the native distribution requires further evaluation, specimen records from disturbed habitats suggest that, like the introduction in the Netherlands, this species has already been introduced to other regions. Some distribution records suggest that M. sahlbergi could indeed be a successful invasive species, and is already successfully established in areas such as disturbed areas on the islands of the Galapagos (Ecuador) and urban areas in Texas, USA, Panama-City, Hawaii, Madagascar, and Reunion. It is easy to confuse M. sahlbergi with the well-known pharaoh ant M. pharaonis, because the former also lives near or in human settlements and looks very similar to M. pharaonis. Therefore, we suspect that M. sahlbergi has more than once been misidentified as M. pharaonis, a view supported by the misidentifications encountered in this study. These findings suggest that M. sahlbergi is likely more common than we realise. partly supported by US National Science Foundation grant DEB-1655076. ACL was supported by Conselho Nacional Científico e Tecnológico (CNPq, Brazil; grant numbers 306772/2019-1 and 300737/2020-3).