Integrative Description of Temnothorax siculus sp. n.: A New Ant Species from Sicily, Italy (Hymenoptera, Formicidae) ^†

Abstract

The mostly Holarctic genus Temnothorax (Hymenoptera, Formicidae) is the most diverse ant genus in temperate regions. The Mediterranean, a biodiversity hotspot of rare ant species, hosts over 150 Temnothorax taxa, including several short-range endemics. Over the last few years, phylogenomic reconstructions and integrative taxonomy have significantly improved the understanding of global Temnothorax diversity, but much taxonomic work is still needed in the Mediterranean region. Here, we present the integrative description of a new species of the genus, discovered in the central Mediterranean island of Sicily: Temnothorax siculus sp. n. is defined and compared to congeneric species integrating morphometrics and phylogenomics. It is a ground-nesting, lowland species, of which workers were regularly observed foraging on bushes and small trees. In the global phylogeny, covering all the main lineages of the region, it belongs to the Palearctic clade and is related to the tuberum and unifasciatus complexes. Morphological separation from other Sicilian Temnothorax species can generally be achieved on qualitative characters, but we also provide morphometric discriminant functions to separate it from T. apenninicus and especially T. unifasciatus. Temnothorax siculus has been rarely collected but appears to be widespread in Sicily, and may occur in neighboring regions.

1. Introduction

The ant genus Temnothorax (Hymenoptera, Formicidae, Myrmicinae), part of the tribe Crematogastrini [1], currently includes over 500 valid species [2]. It is the largest ant genus to have a mostly Holarctic distribution, with only a few species residing in the Neotropical and Afrotropical regions [3,4,5]. Most species are minute ants that live in small colonies, foraging on the ground or climbing on plants, and nesting in a variety of microhabitats, including soil, rock crevices, moss, empty shells, or deadwood, with some species being entirely arboreal [4,6].

Defining species groups in Temnothorax for taxonomic and evolutionary purposes has repeatedly proven difficult, with exclusively morphological attempts often failing to recover monophyletic groups [7]. However, the last few years have seen a tremendous improvement in the understanding of the genus, with the first global phylogenomic reconstruction published in 2017 providing a fundamental basis [4]. Overall, the taxonomy of the genus has a long history of challenges and problems. It still sees a considerable number of novelties published each year, including new species being described [2]. Especially in the western Palearctic, synonymizations have been equally impactful, with modern taxonomic revisions often demonstrating a nomenclatural inflation [8,9,10]. Integrative taxonomy and quantitative methods are playing a pivotal role in finally establishing a clearer and more reliable taxonomic framework [11,12], with some major revisions published over the last decade [9,13,14,15,16,17].

The Mediterranean region, a global biodiversity hotspot for rare ant species with small distribution ranges [18], hosts over 200 species of Temnothorax, a considerable proportion of the known global diversity [19]. Temnothorax is notably richer in small-range species than any other western Palearctic ant genus, with the complex Mediterranean biogeography favoring allopatric speciation [7,20]. Major islands host a significant proportion of this endemic diversity, with the highest amount found on Crete [7,21,22]. Sicily, which is the largest island in the Mediterranean, is estimated to harbor 19 Temnothorax species, of which 5 (26%) are deemed to be endemic or only shared with the neighboring Maltese archipelago [7,10,14,23].

In recent years, many efforts have been spent describing the Sicilian ant fauna, which had long been overlooked [7,24,25,26,27,28,29,30]. With a mixture of biogeographic influences from continental Europe and North Africa, as well as from both the Eastern and Western Mediterranean sectors, and a wide range of ecological conditions due to its climatic and geological diversity, the island contains about 120 taxa, with many still requiring further investigation. Since 2018, six Temnothorax species have been recorded [23,25,26,28], five have been or are being described, and two others have been redescribed [7,9,10,14].

Here, we present the description of T. siculus sp. n., a new distinctive species of Temnothorax discovered in Sicily. We provide a detailed morphometric characterization and phylogenomic data to define its identity in comparison with similar congeneric taxa and reconstruct its position in the global phylogeny of the genus.

2. Materials and Methods

2.1. Examined Materials

Worker specimens of T. siculus sp. n. were collected across Sicily, mostly by AA, as a part of the authors’ broader efforts to document the ant fauna of the island and of neighboring regions, using hand collecting and pitfall trapping, from 1992 to 2020 (see [7,28] for further details on the sampling effort). Sampling in circumsicilian and Maltese islands, as well as Calabria and Tunisia, did not result in additional specimens being collected (but see the Distribution and Ecology in Section 3.3.3). Two workers from distant localities within Sicily were sequenced (see the Morphometrics and Phylogenomics sections below). In addition, we visually inspected two singleton specimens belonging to T. siculus sp. n.: one from the Natural History Museum of Milan (Italy), originally collected and identified as T. affinis (Mayr, 1855) by Christophe Galkowski and recorded as such by Schifani and Alicata [25]; a second one from the ant collection of Roger Vila at the Institute of Evolutionary Biology (CSIC-Univ. Pompeu Fabra) of Barcelona (Spain). A list of all investigated material is provided in the Supplementary Table S1. The terminology in the descriptions follows Schifani et al. [7]. The holotype material and most paratypes were deposited at the Hungarian Natural History Museum, Budapest, Hungary (HNHM).

2.2. Morphometrics

2.2.1. Measured Specimens

Morphometric data of T. siculus sp. n. were obtained via SC from 1–3 specimens per colony sample (based on their availability), in total 24 workers from 13 colonies. These were compared with morphometric data from the only two Sicilian species with some similarities in pigmentation: 73 workers of T. unifasciatus (Latreille, 1798) and 38 of T. apenninicus Csősz, Schifani, Seifert, Alicata, & Prebus, 2024, all from Italy and mainly from Sicily, and originated from previous studies [9,14].

2.2.2. Protocol for Morphometric Character Recording

All measurements were made with a cross-scale graticule at μm precision using a pin-holding stage, permitting rotations around the X, Y, and Z axes with an Olympus (Tokyo, Japan) SZX16 stereomicroscope with a 1.6× Plan Apochromat objective at a magnification of ×192 for each character. Definitions of morphometric characters are listed in

Table 1. Verbatim trait definitions for morphometric character recording.

Abbr.	Definition of the Trait
CL:	Maximum cephalic length in median line; the head must be carefully tilted to the position with the true maximum. Excavations of hind vertex and/or clypeus, if any, reduce cl.
CS:	Cephalic size; the arithmetic mean of CL and cwb.
CW:	Maximum width of head, including eyes; measured across the most distant contour lines of the two compound eyes when the head is in full face view.
CWb:	Maximum width of head capsule, measured posterior to the eyes.
EL:	Maximum diameter of the compound eye.
FRS:	Minimum distance between the frontal carinae.
ML:	Mesosoma length from caudal-most point of propodeal lobe to the transition point between the anterior pronotal slope and anterior propodeal shield (preferentially measured in lateral view; if the transition point is not well defined, use the dorsal view and take the center of the dark-shaded borderline between the pronotal slope and pronotal shield as the anterior reference point).
MW:	Maximum mesosoma width; pronotal width in workers
NOH:	Maximum height of the petiolar node, measured in lateral view from the uppermost point of the petiolar node perpendicular to a reference line set from the petiolar spiracle to the imaginary midpoint of the transition between the dorso-caudal slope and dorsal profile of the caudal cylinder of the petiole.
NOL:	Length of the petiolar node, measured in lateral view from petiolar spiracle to dorso-caudal corner of caudal cylinder. Do not erroneously take as reference point the dorso-caudal corner of the helcium, which is sometimes visible.
PEH:	Maximum petiole height. The chord of ventral petiolar profile at node level is the reference line perpendicular to which the maximum height of the petiole is measured.
PEL	Diagonal petiolar length measured from the base of the anteroventral subpetiolar process to the dorsocaudal corner of the caudal cylinder.
PEW:	Maximum width of petiole
POC:	Postocular distance. Use a cross-scaled ocular micrometer and adjust the head to the measuring position of cl. Caudal measuring point: median occipital margin; frontal measuring point: median head at the level of the posterior eye margin.
PLST	Distance between the most dorsocaudal point of the propodeal lobe and the center of the propodeal stigma.
PPW:	Maximum width of postpetiole.
SL:	Maximum straight line scape length excluding the articular condyle.
SPBA:	The smallest distance between the lateral margins of the spines at their base. This should be measured in dorsofrontal view, since the wider parts of the ventral propodeum do not interfere with the measurement in this position. If the lateral margins of spines diverge continuously from the tip to the base, the smallest distance at the base is not defined. In this case, SPBA is measured at the level of the bottom of the interspinal meniscus.
SPST:	Distance between the center of the propodeal stigma and spine tip. The stigma center refers to the midpoint defined by the outer cuticular ring but not to the center of the real stigma opening that may be positioned eccentrically.
SPWI:	Maximum distance between outer margins of spines; measured in the same position as SPBA.

and follow Csosz et al. [9]. Raw morphometric data are given in μm in the Supplementary Table S1.

2.2.3. Exploratory Analyses Through sPCA in Combination with Gap Statistics

We have limited knowledge about a new species at the time of its description. Hence, to check for the presence of cryptic species within the examined material, we investigated the structure of the morphometric dataset, though a scatterplot generated via principal component analysis (PCA) [31]. PCA identifies discontinuities in continuous morphometric data and visually represents patterns, aiming to capture the maximum variation in the dataset. However, it does not provide an estimate of the ideal number of clusters, meaning that the classification of objects must be interpreted subjectively. Therefore, to determine the optimal number of clusters, we employed the Gap statistics partitioning algorithm [32] based on the principal component scores, utilizing the “clusterGenomics” package [33] with the function “gap”. Our parameters were set as follows: Kmax = 5 (the maximum number of clusters), B = 1000 (bootstrap iterations), and ref.gen = “PC” (reference data generated uniformly over a box aligned with the principal components of the dataset). This method estimates the optimal number of clusters using statistical thresholds and automatically assigns objects to the identified partitions.

2.2.4. Hypothesis Testing via Confirmatory Analyses

To validate the species hypothesis, we conducted hypothesis testing through cross-validated linear discriminant analysis (LOOCV-LDA). Backward stepwise analysis was applied to reduce the number of characters, allowing us to provide the most suitable and easy-to-use numeric tool for users.

2.3. Phylogenomics

2.3.1. Molecular Data Generation

To generate the genetic data, we extracted DNA, prepared genomic libraries, and performed targeted enrichment of ultraconserved elements (UCEs) based on six samples of Temnothorax. We extracted DNA nondestructively from adult worker ants by using a flame-sterilized size 2 stainless steel insect pin to pierce the cuticle of the head, mesosoma, and gaster on the right side of the specimens and then used a DNeasy Blood and Tissue Kit (Qiagen, Inc., Hilden, Germany) following the manufacturer’s protocols. We verified the DNA extract concentration using a Qubit 3.0 Fluorometer (Invitrogen, Waltham, MA, USA). We inputted up to 50 ng of DNA, sheared to a target fragment size of 400–600 bp with a QSonica Q800R sonicator (Qsonica, Newtown, CT, USA), into a genomic DNA library preparation protocol (KAPA HyperPrep Kit, KAPA Biosystems, Wilmington, MA, USA). For the targeted enrichment of UCEs, we followed the protocol of Faircloth et al. [34] as modified by Branstetter et al. [35] using a unique combination of iTru barcoding adapters (BadDNA, Athens, GA, USA [36]) for each sample. We performed enrichments on pooled, barcoded libraries using the catalog version of the Hym 2.5Kv2A ant-specific RNA probes (Arbor Biosciences, Ann Arbor, MI, USA [35]), which targets 2524 UCE loci in Formicidae. We followed the library-enrichment procedures for the probe kit, using custom adapter blockers instead of the standard blockers (BadDNA, Athens, GA, USA [36]), and left enriched DNA bound to the streptavidin beads during PCR, as described in [34]. Following post-enrichment PCR, we purified the resulting pools using SpeedBead magnetic carboxylate beads (Sigma-Aldrich, St. Louis, MO, USA [37]) and adjusted their volume to 22 μL. We verified the enrichment success and measured size-adjusted DNA concentrations of each pool with qPCR using a SYBR-FASTqPCR kit (Kapa Biosystems, Wilmington, MA, USA) and a Bio-Rad CFX96 RT-PCR thermal cycler (Bio-Rad Laboratories, Hercules, CA, USA), subsequently combining all pools into an equimolar final pool. We sequenced the final pool in one lane at Novogene (Sacramento, CA, USA) on Illumina HiSeq 2500 150 cycle Paired-End Sequencing v4 runs (llumina, San Diego, CA, USA), along with other enriched libraries for unrelated projects.

2.3.2. Dataset Construction

To place the new species within the context of the Palearctic Temnothorax species, we combined our newly generated sequence data with previously published sequences for a total dataset of 30 samples (Refs. [7,10] see Supplementary Table S1 for voucher specimen data). At the current stage, the phylogenomic data over global and Palearctic Temnothorax fauna cover all the European and Mediterranean lineages, including almost all the species groups occurring in the investigated region and almost all Sicilian species [7]. This allowed us to compare the new species with satisfactory coverage of possible synonyms. We followed the standard PHYLUCE protocol for processing UCEs in preparation for the phylogenomic analysis [38], aligning the monolithic unaligned FASTA file with the phyluce_align_seqcap_align command, using MAFFT [39] as the aligner (--aligner mafft) and opting not to edge-trim the alignment (–no-trim). We trimmed the resulting alignments with the phyluce_align_get_gblocks_trimmed_alignments_from_untrimmed command in PHYLUCE, which uses GBlocks ver. 0.91b [40], using the following settings: b1 0.5, b2 0.5, b3 12, b4 7. After removing UCE locus information from taxon labels using the command phyluce_align_remove_locus_name_from_nexus_lines, we examined the alignment statistics using the command phyluce_align_get_align_summary_data and generated a dataset in which each locus contains a minimum of 90% of all taxa using the command phyluce_align_get_only_loci_with_min_taxa. Newly generated sequence reads were deposited in the NCBI sequence read archive (SRA) under BioProject PRJNA1227512.

2.3.3. Phylogenetic Inference

In our initial analysis, we inferred a phylogeny with maximum-likelihood analysis in IQTREE v2.1.2 [41], using the unpartitioned matrix as input, setting the substitution model to GTR+G, and using 1000 ultrafast bootstrap replicates (-B 1000) to estimate statistical support for the inferred topology.

Because the assumption that the evolutionary rates of sequence data are homogenous is often violated in empirical data [42], we partitioned our UCE loci into sets of similarly evolving sites. To achieve this, we used the command phyluce_align_format_nexus_files_for_raxml, which concatenates loci into a single alignment and generates a partition file for input into the SWSC-EN method [43], as implemented in the CURE pipeline [44]. We used the resulting datablocks as input for partitioning in IQTREE2 using the command -m MFP+MERGE. Because the combination of gamma and proportion of invariable sites (+I+G) has been demonstrated to result in anomalies in the likelihood estimation [45,46], we set the rate heterogeneity models to a subset that includes everything except the combination of gamma and proportion of invariable sites (-mrate E, I, G) and set the search algorithm to -rclusterf 10. We used the resulting partitioned dataset as input for maximum likelihood tree inference in IQTREE, using 1000 ultrafast bootstrap replicates (-B 1000).

3. Results

3.1. Morphometrics

Our PCA analyses returned three principal components (PCs) that explain a significant portion of the variation in the data, with each PC accounting for over 10% of the variance: PC1 contributes 43.2%, while both PC2 and PC3 each contribute 16.6%. Notably, the patterns did not reveal any clearly defined groups among the set workers (Figure 1A). Furthermore, the gap statistics indicated a single cluster within the data (Figure 1B), reinforcing the absence of distinct patterns shown in the PCA. This evidence indicates that there are no unexplored cryptic species present within the dataset of T. siculus sp. n.

Morphometric Separation from Similar Sympatric Species

In the separation between T. siculus sp. n. and T. apenninicus, the LDA confirmed the classification by returning an overall accuracy = 1.0 at the individual worker level. The more pessimistic Leave-One-Out Cross-Validation (LOOCV) yielded a 100% classification success (every case was correctly classified). In challenging cases, a shorter function after drastic character reduction using a combination of two traits (D2 = + 0.054 * ML − 0.090 * SL + 2.710) still provides a powerful tool for identification offering a satisfactory 98.4% classification success (Figure 2).

siculus (n = 24) = +2.038 [+0.384, +4.895]

apenninicus (n = 38) = −2.038 [−4.389, +0.239]

In the separation between T. siculus and T. unifasciatus, the LDA confirmed the classification and returned an overall accuracy = 0.99 at the individual worker level. The more pessimistic Leave-One-Out Cross-Validation (LOOCV) yielded a 99% of classification success (resulting in a single misclassification out of the total of 97 cases). In challenging cases, a shorter function after drastic character reduction using a combination of four traits (D4 = +0.073 * MW + 0.092 * FR − 0.090*SL − 0.078*PEH +13.032) still provides a powerful tool for identification offering a staggering 99% classification success (Figure 3).

siculus D4 (n = 24) +1.993 [−0.654, 3.851]

unifasciatus D4 (n 73) = −1.993 [−4.051, +0.016]

3.2. Phylogenomics

3.2.1. Sequence Capture and Processing

We sequenced ultraconserved elements from six new samples of Temnothorax and incorporated several published sequences for a total dataset of 30 samples. The 90% complete matrix contains 2134 UCE loci, with a mean locus alignment length of 1421 bp and a total length of 3,040,494 bp. We deposited raw reads sequenced for this study in the NCBI sequence read archive (SRA), under BioPoject ID PRJNA1227512 (see Supplementary Table S1 for additional sequence data).

3.2.2. Phylogenetic Inference

Partitioning of the dataset with CURE and clustering with IQTREE resulted in a 140-partition scheme. The unpartitioned and partitioned analyses differed only slightly with respect to topology and bootstrap support, mainly in the placement of Temnothorax alienus (compare Figure 4 and Supplementary Figure S1). Temnothorax siculus sp. n. belongs to the Palearctic clade which includes most of the Paleartic species of the genus [4] and all of those occurring in Sicily (except for a few species belonging to the rottenbergii and rugulatus clades [4,7,14]). Within it, it belongs to a group that includes the tuberum and unifasciatus complexes, as well as T. nigriceps (Mayr, 1855), with T. unifasciatus being the sole other species occurring in the study area.

3.3. Description of Temnothorax siculus sp. n.

Zoobank: urn:lsid:zoobank.org:act:8C30C9B8-0F99-4D81-A043-341FE185FFEC

Holotype: 1 worker, Fossa della Garofala, Palermo, Sicily (Italy), 38.106420, 13.350403, 18.V.2021, 41 m asl, E. Schifani legit, codename ITA006a (upper specimen of its pin), Hungarian Natural History Museum, Budapest, Hungary (HNHM).

Paratypes: A total of 32 workers are designated as paratypes. Eight workers with the same data as the holotype, ITA006b (HNHM), ES (ES21A058), and IT_P899 (MP personal collection). One worker, Macconi Santa Lucia, Caltanissetta, 37.04614444, 14.28351667, 8 m asl, 6.X.2004, A. Alicata legit, ITA:AAS8E3 (HNHM). One worker, Vallone Piano della Corte, Enna, 37.64834, 14.49314, 524 m asl, 19.XII.2017, A. Alicata legit, ITA:AAS12C (HNHM). One worker, Vallone Piano della Corte, Enna, 37.64943, 14.49164, 527 m asl, 19.III.2018, A. Alicata legit, ITA:AAS14H (HNHM). One worker, Cozzo Viglio, Villapriolo, Enna, 37.626167, 14.172367, 620 m asl, 29.VI.2007, A. Alicata legit, ITA:AAS3D1 (HNHM). Three workers, Vallone Mazzarò, Messina, 37.855422, 15.297211, 45 m asl, 24.IV.2018, A. Alicata legit, ITA:S10F8 (HNHM). Two workers, Bosco di Milo, Catania, 37.707878, 15.116311, 600 m asl, 12.VI.1992, A. Alicata legit, ITA:AAS4B3 (HNHM). One worker, Bosco di Passopomo, Catania, 37.673346, 15.127222, 370 m asl, 15.III.1992, A. Alicata legit, ITA:AAS4I5 (HNHM). Three workers, Timpa Santa Maria delle Grazie, Catania, 37.594235, 15.172545, 60 m asl, 9.V.1992, A. Alicata legit, ITA:AAS89E1 (HNHM). One worker, Linera, Catania, 37.669278, 15.137008, 290 m asl, 16.VI.1992, A. Alicata legit, ITA:AAS4G8 (HNHM). 1 worker, Bosco di Santo Pietro, Catania, 37.075862, 14.510289, 252 m asl, 2.VI.2019, A. Alicata & S. Csősz legit, IT_P578 (MP personal collection). Three workers, Saline di Priolo, Siracusa, 37.144536, 15.214746, 6 m asl, 2002, A. Alicata legit, ITA:SalinePriolo (HNHM). Three workers, Bosco di Baulì, Siracusa, 37.03136944, 14.93931111, 620 m asl, 16.V.1993, A. Alicata legit, ITA:AAS5I7 (HNHM). Two workers, Sortino, Siracusa, 37.161148, 15.042615, 447 m asl, 25.VII.1993, G. Silluzio legit, ITA:AAS11E8 (HNHM).

3.3.1. Etymology

The species is dedicated to the island of Sicily where it was discovered, siculus being the Latin adjective that translates as “Sicilian”.

3.3.2. Formal Description of the Worker

Head, mesosoma, and nodes yellow to light ferruginous. Gaster yellow, with a dark transversal band mainly restricted to the posterior half of the first tergite. Appendages are yellow, except for the dark antennal club (at least the last two segments). Head subrectangular, with head margins, clypeus, and mandibles rounded. Antennae of 12 segments, antennal club of three segments. Compound eyes ovoidal. The mesosomal dorsal profile is curved and lacks a marked metanotal impression. Propodeal spines are very variable in length and slightly recurved when longer. Petiolar dorsal profile variable from rounded to angulate. Postpetiole ordinary rounded in profile view, subrectangular in dorsal view. Surface sculpturing is relatively weak, with a possible smooth area on the frons, the background sculpturing on the rest of the body being areolate-rugose, with some irregular longitudinal and transverse striae on the mesosoma. Gaster and appendages smooth. See Figure 5.

3.3.3. Distribution and Ecology

It is widely distributed across Sicily but seems rather elusive to collect and we know that it only originates from a limited number of localities, sparse through the provinces of Agrigento, Caltanissetta, Catania, Enna, Messina, Palermo, and Siracusa (Figure 6). Despite our lack of records, based on this pattern, its occurrence in Calabria and/or in the Maltese Islands should be further investigated [24]. Across this range, it has always been collected from habitats with trees or shrubs (oak forests, riparian vegetation, olive groves, a seminatural urban area) and never from more arid and open habitats. All sites are relatively thermophilous, lowland areas (from 8 to 620 m in elevation). Nests, although were located only once, are apparently built on the ground, but workers have been consistently observed climbing on tall herbaceous plants, bushes, or small trees for foraging: on bear’s breeches (Acanthus mollis L.), on spiny brooms (Calicotome sp.), on the extrafloral nectaries of the invasive tree of heaven (Ailanthus altissima [Mill.] Swingle) together with Temnothorax mediterraneus Ward, Brady, Fisher & Schultz, 2014, and tandem-running on olive trees [25]. Shadier and colder sites even within this elevational range are likely unsuitable.

3.3.4. Qualitative Diagnostic Traits and Worker-Based Key to the Sicilian Temnothorax Fauna

Most Sicilian Temnothorax species have typically different color patterns, either being dark-colored [i.e., T. laestrygon (Santschi, 1931), T. mediterraneus, T. rottenbergii (Emery, 1870), and its sister species under description [10]], or, if light-colored, having concolor light antennal clubs (i.e., T. clypeatus (Mayr, 1853), T. flavicornis (Emery, 1870), T. lagrecai (Baroni Urbani, 1964), T. lichtensteini (Bondroit, 1918), T. marae Alicata, Schifani, & Prebus, 2022, T. nylanderi (Foerster, 1850), T. poldii Alicata, Schifani, & Prebus, 2022, T. recedens (Nylander, 1856), T. vivianoi Schifani, Alicata, & Prebus, 2022, and the social parasites T. algerianus (Cagniant, 1968) and T. muellerianus (Finzi, 1922)]. Antennal clubs tend to be slightly infuscated in T. apenninicus, but the latter is otherwise lighter in color, and it lacks a fully developed transversal band on the gaster. The only species that often has a very similar color pattern is T. unifasciatus, which is usually characterized by darker antennal clubs, a straight rather than rounded dorsal profile of the mesosoma, and a coarser sculpturing. Apart from pigmentation, most of these species are also morphologically very different, and should be easily distinguished by anybody familiar with the morpho-taxonomy of Mediterranean Temnothorax.

In qualitative terms, T. apenninicus and especially T. unifasciatus are the most similar species occurring on the island. The first is indeed not related at all, as it belongs to the luteus group from the rottenbergii clade and should be very easy to distinguish by a trained eye. On the other hand, T. unifasciatus is more closely related. Both species were collected at much higher elevations of at least 1180 m asl [9,14]. In the event of confusion, a safe morphological identification can be attained using the two discriminant functions illustrated in the Section Morphometric Separation from Similar Sympatric Species of this paper.

Furthermore, we provide a key to navigate through all the species of Temnothorax currently known in Sicily:

1.	Frontal carinae extending to the level of the posterior eye margin, ventral postpetiole with an acute denticle………………………………………………………T. muellerianus [6]
	- Frontal carinae short, ventral postpetiole without denticles……………………………2
2.	Petiole in lateral profile view with a very broad lobiform process on the ventral side… …………………………………………………………………………………T. algerianus [17]
	- Petiole in lateral profile view only with very small, non-lobiform processes…………3
3.	Antennae normally with 11 segments……………………………………T. flavicornis [7,13]
	- Antennae with 12 segments…………………………………………………………………4
4.	Very long scapi, metanotal impression deeply marked with a very strong discontinuity in the lateral profile of the mesosoma between the propodeum and the rest, body covered with several long setae, sculpture extremely reduced……………T. recedens [6]
	- Shorter scapi, lateral profile of the mesosoma less discontinuous, shorter and thicker setae on the body………………………………………………………………………………5
5.	Whole head and body dark-blackish, only the mesosoma may be occasionally reddish ……………………………………………………………………………………………………6
	- Head and body yellowish to ferruginous, only the gaster may be partly or entirely blackish, and sometimes the frons……………………………………………………………8
6.	Head and mesosoma with a very strong sculpture, petiole in profile view subrectangular…………T. rottenbergii or its cryptic sister species under description [10]
	- Head sculpture much reduced, petiole in profile view subtriangular…………………7
7.	Long propodeal spines, clypeus with a middle carina, petiole lateral profile triangular, without a clear anterior peduncle. Arboreal species, typically wood-nesting…………… ………………………………………………………………………………T. mediterraneus [48]
	- Propodeal spines short, clypeus without a middle carina, petiole typically pedunculate. Ground nesting………T. laestrygon (pending a revision of the exilis group)
8.	Antennal club normally darker than the rest of the antenna, mesosoma lateral profile not consistently characterized by a metanotal impression…………………………………9
	- Antennae concolor light, a metanotal impression may or may not be present………11
9.	Long scapi [SL/CW: 0.87 (0.82–0.91)]. Mesosoma dorsal profile in lateral view rounded, antennal club only very slightly infuscate, gaster with a weak dark band, body otherwise yellowish, high elevation species (1400–2000 m asl).………T. apenninicus [14]
	- Shorter scapi. Either a lowland species or the dorsal profile of the mesosoma is largely straight. Antennal clubs normally clearly darkened or blackish, and gaster more extensively dark-colored………………………………………………………………………10
10.	Dorsal profile of the mesosoma straight, surface sculpturing coarser. Mid-high elevations (1180–1975 m asl). Can be told apart using the discriminant function illustrated in Section Morphometric Separation from Similar Sympatric Species: D4 (+0.073 * MW + 0.092 * FR − 0.090 * SL − 0.078 * PEH + 13.032) = −1.993 [−4.051, +0.016] …………………………………………………………………………………T. unifasciatus [9]
	- Dorsal profile of the mesosoma rounded, sculpture weaker. Low elevations (8–620 m asl). D4 = +1.993 [−0.654, 3.851]…………………………………………………T. siculus sp. n
11.	Bulky, high and short petiole in lateral view, with a slightly convex frontal face forming a rounded-rectangular transition with the dorsal crest, and straight mesosoma dorsum. Rare, relatively large-sized arboreal species………………………………T. clypeatus [6,28]
	- Character combination different……………………………………………………………12
12.	In lateral view, petiole narrow and high, without or with a very minimally developed subpetiolar process, the dorsal profile of the mesosoma is interrupted by a marked metanotal impression………………………………………………………nylanderi group, 13
	- Character combination different……………………………………………………………14
13.	Long propodeal spines deviating by <28.5° from the mesosoma axis. Lower altitudes …………………………………………………………………………………T. lichtensteini [13]
	- Spies shorter (SPST/CS), deviating by 32–42° from the mesosoma axis. High altitudes …………………………………………………………………………………T. nylanderi [13,28]
14.	Normally larger in size (CL: 636–771, ML: 671–905), color ferrugineous, stronger sculpture with several striae except for a notable smooth shiny area in the frons, spines long……………………………………………………………………………………T. poldii [7]
	- Smaller size, color yellow, sculpture much weaker but more homogeneous (no smooth area in the frons), spines variable……………………………………………………………15
15.	Spines long, subpetiolar process forming a small denticle……………………T. lagrecai [7]
	- Spines short, subpetiolar process reduced…………………………………………………16
16.	Dorsal profile of the petiole blunt, including a horizontal plane. South-western Sicily… …………………………………………………………………………………………T. marae [7]
	- Profile of the petiole forming a sharp triangle, with no horizontal dorsum. North- western Sicily………………………………………………………………………T. vivianoi [7]

4. Discussion

The last few years have seen growing efforts spent on documenting the ant fauna of Sicily, producing the discovery of over 40 new species on the island, including native and non-native ones, and the description of new taxa [7,10,14,23,24,25,26,27,29,30,49,50,51]. They have also contributed to uncovering the biogeography of the island’s ants to a new extent, documenting a larger contribution of Maghrebian influences than previously documented [23,24,27]. In this context, T. siculus sp. n., possibly a Sicilian endemic, is an enigmatic new piece of the puzzle that contributes to the distinctiveness of the island’s faunistic assemblage [20]. The phylogenomic reconstruction places it within the Palearctic clade and close to the unifasciatus and tuberum complexes, which include T. unifasciatus and T. tuberum in Italy, only the first of which extends its distribution south and reaches Sicily. Further researchers should look at North Africa, the Maltese Islands and Calabria for possible extensions of the distribution range of T. siculus sp. n., while its apparent absence from most circumsicilian islands could have ecological reasons. In biogeographic terms, T. siculus may appear to be a lineage of European origins that isolated south in Sicily, but it is a thermophilous species unlike T. apenninicus or other Sicilian endemic ants of which ancestors must have colonized the island from the north, suggesting the alternative hypothesis of a southern origin [14,49]. Expanding the phylogenomic understanding of all Mediterranean Temnothorax, including from currently underrepresented regions, such as North Africa and the Eastern Mediterranean, will be important to further clarify its evolutionary and biogeographic origins.

Several faunistic and taxonomic novelties have concerned the Sicilian Temnothorax fauna. The list of species presented by Schifani et al. [7] is already outdated. Temnothorax affinis must be removed given the re-identification of the only record [25], it could be restricted to the northern regions of the country, and its Italian distribution should be revised considering past confusion with T. aveli or T. italicus [52]. Furthermore, the recent record of T. ravouxi (André, 1896) should instead be attributed to T. algerianus, also recently recorded [23], based on the recent revision of the corsicus species group (formerly Myrmoxenus) [53]. Finally, T. apenninicus was described in 2024 [14]. At present, we consider the following 17 species to be present: T. algerianus, T. apenninicus, T. clypeatus, T. flavicornis, T. laestrygon, T. lagrecai, T. lichtensteini, T. marae, T. mediterraneus, T. nylanderi, T. poldii, T. recedens, T. rottenbergii and its sister species under description [10], T. siculus sp. n., T. unifasciatus, and T. vivianoi. In this list, we consider all the few mentions of T. exilis in Sicily under T. laestrygon, which was described from the island. However, the exilis group needs to be revised to check whether T. laestrygon is a good species and if more than one species of the group occurs in Sicily. As for most species in the genus, the queen and male morphology in T. siculus sp. n. still needs to be uncovered [7].

Modern ant taxonomy is increasingly relying on quantitative methods, including morphometry and molecular approaches, to minimize the chances of error that subjective traditional approaches often imply, but also to offer reproducible procedures and comparable datasets for hypothesis testing [12]. In particular, integrative taxonomy further reduces error chances and produces more informative accounts on the evolution and natural history of the species, while also offering more types of data for future comparisons and studies [11,12]. While taxonomic revisions of species groups and complexes are fundamental to advance ant taxonomy at this stage, isolate species descriptions can still have a place when a new taxon can be adequately defined and is not part of a cryptic complex. The morphometric and phylogenomic data produced in this study will offer the possibility of further investigating the evolutionary and biogeographic history of T. siculus sp. n. in future studies.