Myrmica carbonaria Smith 1858

Seifert, Bernhard, 2025, The Monomorium carbonarium species group in the Nearctic and Europe (Hymenoptera: Formicidae), Soil Organisms (Toronto, Ont.) 97 (1), pp. 55-84 : 67-72

Myrmica carbonaria Smith 1858 [type investigation]

This species has been described from Madeira. Investigated were four worker syntypes labelled “ Madeira ”, “ Smith coll. pres. by Mrs. Farren White 99-303.”, “ SYNTYPE ” and “ANTWEB CASENT0902279”; depository NHM London. The type series is centrally placed within the Nearctic-European Monomorium carbonarium cluster. For more details see below un- der Taxonomic Comments .

Monomorium minimum ( Buckley 1867)

[syn. nov., type investigation].

This taxon has been described as Myrmica (Monomarium) [sic!] minima Buckley 1867 from Texas. Buckley (1867) gave only the terra typica: “Common in Central Texas ”. According to Creighton (1950) and DuBois (1986) no type material could be identified in any collection in the USA including that of the Academy of Sciences Philadelphia. Accordingly, DuBois (1986) fixed a neotype for Monomorium minimum (Buckley) in a winged gyne from Texas labelled “Bastrop St. Pk., Tex. June 9, 1954 W. Clayd, T-119”, “ Monomorium minimum (Buckley) Neotype M. DuBois 1983” and “M.C.Z. NEOTYPE 32921”, depository MCZ Cambridge. Unfortunately there is no worker material associated to the neotype to have a more reliable classification. Monomorium carbonarium develops large and small winged gynes (see Tab. 5). The latter are most similar to winged gynes of M. ergatogyna and the neotype with CS= 0.736 mm may belong to both species ( Tab. 5). Twelve worker-associated small winged gynes of M. carbonarium but only two worker-associated winged gynes of M. ergatogyna were available for comparison. Running the neotype of M. minimum as wild-card in a stepwise LDA considering all characters shown in Tab. 5 it is allocated to M. carbonarium with p>0.98. Yet, this is a weak suggestion because of strong character overfitting and the high variance and small sample size in the comparison samples. A better argument in favor of a synonymization with M. carbonarium is its higher abundance in Texas: 11 evaluable worker samples of Monomorium , all collected before 1947, were available from Texas of which 91% belonged to M. carbonarium and only 9% to M. ergatogyna . The situation today may differ. The third supporting argument is the higher frequency of winged gynes in the studied Nearctic material of M. carbonarium : the ratio of winged gynes vs. wingless gynes was 14: 5 in M. carbonarium but 2: 24 in M. ergatogyna . These arguments multiply to a fair probability that M. minimum from Texas should be a junior synonym of M. carbonarium . A really convincing argumentation is not possible at the current stage.

Monomorium minimum emersoni ( Gregg 1945)

[syn. nov., type investigation]

This taxon has been described from Globe / Arizona (holotype series) and Austin / Texas (paratypes) and San Marcos / Texas (paratypes). In NMNH Washington are 1 paratype winged gyne plus 51 worker paratypes on 17 pins all labelled “Austin Tex. IV-6-1937 A Emerson” [printed] and “ Monomorium minimum emersoni subsp. nov. R. Gregg det R Gregg” [hand-written by Gregg himself]. As it is not clear if this large material represents a nest sample, two worker series with remote placement in the box and the gyne were morphometrically investigated. The gyne data allow no decision – is either a small winged gyne of M. carbonarium or a winged gyne of M. ergatogyna . However, the two worker series were allocated to M. carbonarium with p=0.9999 and 0.9982. Nevertheless, the synonymization with M. carbonarium stated here is risky as the holotype series from Globe was not studied. Based on abundance of species in the region the holotype series could also belong to M. ergatogyna or M. emarginatum . If belonging to the latter, M. emersoni would become a senior synonym. The holotype series from Globe was not available.

Monomorium trageri DuBois 1986

[syn. nov., type investigation]

This taxon has been described from Gainesville in Alachua County, Florida based on a nest population collected from beneath a board in a cow pasture. The specimens investigated from the holotype nest are as follows. Holotype wingless gyne labelled “ Florida: Alachua C., Gainesville April 5, 1980 J.C.Trager ” and “ Monomorium trageri Holotype M. Dubois 1983”; 1 paratype wingles gyne, 1 paratype worker with same locality label, depository NMNH Washington. 1 worker and 1 wingless gyne from holotype nest labelled “ Florida: Alachua C., Gainesville April 5, 1980 J.C.Trager ” and “ Monomorium trageri Paratype M. DuBois 1983.” [blue, bleached label, year not clearly visible], depository AMNH New York. 1 worker and 1 wingles gyne from the holotype nest labelled “ Florida: Alachua C., Gainesville April 5, 1980 J.C.Trager ” and “ Monomorium trageri Paratype M. DuBois 1983.” [blue, bleached label, year not clearly visible], depository CAS San Francisco. The wingless gyne from CAS has a malformed propodeum and waist segments and is not included in the analysis. Monomorium trageri is synonymized here with M. carbonarium because the worker sample is allocated to the M. carbonarium cluster with p = 0.9989 if run as wild-card in a LDA and because the three evaluable type wingless gynes are placed in a PCA close to wingless gynes of M. carbonarium ( Fig. 7 ). For more details see below under Taxonomic Comments.

All material examined. Numeric phenotypical data were taken in 45 samples with 141 workers and 27 gynes. The material came from France (12 samples), Germany (1), Mexico (1), Netherlands (3), Portugal (3) and USA (25). For details see supplementary information SI1, SI2 and SI3 .

Geographic range. Widely distributed over the USA south to Mexico. Introduced to West and Central Europe. Altitudinal records range from sea level up to 2900 m (at 20°N in Mexico) .

Description: - -Worker ( Tab. 3; Figs. 23, 24 ; key). Very small, CS 470 ±22 µm. Head moderately long, CL/CW 1.191 ± 0.027. Dents on clypeal margin moderately long (ExCly/CS 4.80 ± 0.58%), not very acute and moderately distant (ClSpD/CS 11.42 ±0.97%). Frontal carinae moderately distant and slightly diverging frontad (FL/CS 0.285 ± 0.008, FR/CS 0.276 ± 0.008, FL/FR 1.031 ± 0.021). Eye and preocular distance moderately large (EL/CS 0.205 ± 0.007, PrOc/ CS 0.224 ±0.010). Scape and funiculus segments moderately long (SL/CS 0.807 ± 0.021, Fu2/CS 6.38 ± 0.21 %, Fu3/CS 4.82 ± 0.25%). Mesosoma rather short but relatively wide (ML/CS 1.179 ± 0.028, MW/CS 0.581 ± 0.014). Metanotal groove moderately deep, MGr/CS 5.24 ± 0.66%. Dorsum and slope of propodeum in profile forming a convex curve. Petiole rather wide and high (PeW/CS 0.289 ± 0.018, PeH/CS 0.378 ± 0.013). Postpetiole moderately wide and rather high (PpW/CS 0.316 ± 0.017, PpH/CS 0.294 ± 0.011). All body surfaces with exception of few rugulae on frontal lobes, ventral mesopleuron, ventrolateral area of metapleuron and genae glabrous and shiny. All body parts homogenously dark brown to jet black; mandibles, lateral clypeus and antennal funiculus lighter brown.

--Gyne ( Tabs. 4, 5): The Nearctic population is extremely polymorphic in gynes, developing winged and wingless gynes with strong size variation in each morph. The sparse material available does not allow to propose diagnostic determination characters. The large winged gynes are probably better to identify as a combination of short scape and long petiole ( Tab. 5). Taxonomic comments. The description above does not present a single diagnostic character which results in the uncomfortable situation that species delimitation from M. ergatogyna and M. gallicum n.sp. is only possible by multivariate analyses. A species hypothesis was formed by NC-part.hclust considering the 12 characters CS, CL/CW, SL/CS, EL/CS, ClSpD/CS, MW/CS, PpW/CS, PeH/CS, PeL/CS, PpH/ CS, MGr/CS, Fu3/CS. Within a total of 96 samples, it proposed three clusters and five outliers. In the controlling LDA, these clusters were accepted as species hypothesis and the outliers were run as wild-cards. The final species hypothesis of these runs is given in Fig. 33. It indicated that NC-part.hclust misclassified only one sample of M. gallicum n.sp. whereas the clusters later identified as M. ergatogyna and M. carbonarium were incompletely separated ( Fig. 31 ). This result was repeated by a PCA in which M. gallicum n.sp. formed a well separated and constricted cluster ( Fig. 32 ). One sample of M. carbonarium that approached M. gallicum n.sp. cluster was run as wild-card in the LDA and confirmed as the former species. In order to improve the separation of M. carbonarium and M. ergatogyna , M. gallicum n.sp. was excluded from analysis and NC-part.hclust was run using the 11 characters FR/CS, PeH/CS, SL/CS, ClSpD/CS, PpW/CS, MGr/CS, PeW/CS, EL/CS, PpH/CS, FL/CS, MW/CS. The analysis proposed two clusters and seven outliers within a total of 73 samples ( Fig. 34 ). The final species hypothesis indicated that NC-part.hclust misclassified two samples. A PCA and NC-NMDS.kmeans run with the same data misclassified three and five samples respectively. Note: The cluster named above M. carbonarium was divided by NC-part.hclust into three subclusters. I decided to give these sublusters no taxonomic rank because NC-part.hclust tends to oversplit and because the resulting sample size of these subclusters became too small to reasonably check the data by the controlling LDA. Anyway, these data show the need for further research and that some of the synonymizations proposed in this paper under M. carbonarium might possibly be reversed.

Biology. The unresolved species delimitation in the past does not allow to use literature sources on life history and habitat selection referring to the situation in the New World. Some 200 samples in American museum collections labelled by DuBois as “ Monomorium minimum ” are a random mixture of M. carbonarium , M. ergatogyna and M. emarginatum . These three species are abundant and widely distributed in the USA which prevents any probabilistic interpretation of published biological data based on zoogeography. The more advanced species delimitation presented here and published information on particular classified samples allows to extract some fragments of life history: M. carbonarium is extremely gyne polymorphic in America with the winged gyne most certainly representing an ability for long-range flight-dispersal and independent colony foundation. Most colonies seem to be polygynous. Winged gynes are present in polygynous colonies but the wingless gynes are the dominant gyne morph in very polygynous (in Europe supercolonial) colonies. In a polygynous colony from Gainesville / Florida (type location of M. trageri ), 10% of the queens were winged gynes and 90% wingless gynes and the latter were mated intranidally ( DuBois 1986). All over the USA, young winged gynes with wings and males were found in the nests from 6 June to 24 July with the bulk in July. Wingless gynes were observed to perform female calling, dispersed on the whole plant of a waist-high Ailanthus bush and were accompanied by a retinue of workers. One to multiple alate males engaged on a single female but the mating count was uncertain (SaNo 141, 24 June 2023, Washington D.C., Brendon Boudinot pers. comm. 2024). Only wingless gynes are known so far from Madeira and West Europe which seems to contrast the very polymorphic situation in the Nearctic. This reduction to wingless gynes and the very recent occurrence as invasive pest ant in urban areas of West Europe support the assumption that the European population was founded by introduction from America. Supercoloniality has not been reported so far from the USA but in five of the 16 European locations. One supercolony in a greenhouse in the Netherlands caused a lot of problems there and another supercolony in a school yard in southern France expanded its territory by 18 m within a year. The distribution over main habitat types in Europe was 14 findings in urban, semi-urban or rural habitats and 2 findings in “natural” habitats (i. e., habitats with lowest anthropogenic influence such as coastal grey dunes or pastures). This distribution seems to differ from the situation in the sympatric and biologically very similar Monomorium gallicum n.sp. where 16 findings were made in the first habitat group but 7 in the latter. The soil of the urban habitats can be fully sealed with almost no vegetation being present.

Monomorium ergatogyna Wheeler 1904

Figures

Figure 7. Principal component analysis of wingless gynes of Monomorium carbonarium (black dots), M. ergatogyna (white rhombs) and M. gallicum n.sp. (black crosses). Acronyms indicate the placement of type specimens: W – holotype of M.wheelerorum, C – paratype of M. cyaneum, T – types from holotype nest of M. trageri, G – types from holotype nest of M. gallicum n.sp., E – types from holotype nest of M. ergatogyna.

Figure 24. Lateral aspect of Monomorium carbonarium; Netherlands: Zwartevaal, 2022

Figure 31. NC-part.hclust analysis of workers of Monomorium carbonarium (black bars), M. ergatogyna (blue bars), and M. gallicum n.sp. (red bars). The empty bar refers to outliers.

Figure 32 (left). Sample means of principal component analysis considering 12 characters measured in workers of of Monomorium carbonarium (black dots), M. ergatogyna (white rhombs), and M. gallicum n.sp. (black crosses). Figure 33 (right). Sample means of a linear discriminant analysis considering 12 characters measured in workers of of Monomorium carbonarium (black dots), M.ergatogyna (white rhombs),and M.gallicum n.sp. (black crosses).Acronyms indicate the placement of type specimens: CA – type sample of M.carbonarium, CY – paratype sample of M. cyaneum, EM – paratype sample of M.emersoni, ER – type sample of M. ergatogyna, GA – type sample of M.gallicum n.sp., TR – type sample of M.trageri, WH – type specimen of M. wheelerorum.

Figure 34. Results of three exploratory data analyses (EDAs) of worker samples ofMonomorium carbonarium (blackbars) and M.ergatogyna (grey bars).Shown is the NC-part.hclust dendrogram and results of NC-NMDS.kmeans and Principal Component Analysis.The empty space in the NC-part.hclust bar refers to outliers. The mean error of the three EDAs relative to the final species hypothesis is 5.0%.