4.4.42 Lasius tebessae Seifert 1992

Lasius tebessae Seifert 1992 [type investigation]

Type material: Holotype plus 8 paratype workers labelled ” Algeria Tebessa 16.5. 1967, 950 m ex coll. Cagniant“; 8 paratype workers labelled ” Algeria: 50 km S Alger, 17.6.1964, 1200 m ex coll. Cagniant “; depository SMN Görlitz.

All material examined. A total of 6 nest samples with 28 workers were subject to NUMOBAT investigation. These originated from Algeria (2 samples) and Morocco (4). For details see supplementary information SI1.

Geographic range. Occurring from North Morocco (34.915°N, 5.393°W) to East Algeria (35.400°N, 8.117°E) at elevations between 888 and 1750 m.

Biology. The species is apparently bound to habitats with trees. Reported were Quercus ilex, Pinus halepensis and Picea forests and a open pasture with scattered trees.

Diagnosis (Tab. 8, Figs. 81 –82; key):

Absolute size rather large (CS 966 µm). Head length index large (CL/CW 900 1.090), scape length index large (SL/CS 900 1.037); postocular distance small (PoOc/CL 900 0.218), eye and torulo-clypeal distance large (EYE/CS 900 0.249, dClAn 900 5.20); terminal segment of maxillary palp long (MP6/CS 900 0.230). Number of mandibular dents of average level (MaDe 900 8.13). Pubescence on clypeus denser than in L. emarginatus (sqPDCL 900 4.07); frontal pubescence rather short (PLF 900 27.6). Setae length slightly larger than in L. emarginatus but their number very similar (PnHL/CS 900 0.155, GuHL/ CS 900 0.136, nOcc 900 11.6, nGen 900 6.3, nGu 900 7.0, nSc 900 8.8, nHT 900 14.7). Coloration: variable from concolorous medium to dark brown to moderately bicolored with mesosoma lighter than head and gaster. The concolorous pale yellowish-reddish brown pigmentation of the type specimens is possibly caused by bleaching during 25 years of ethanol storage.

Comments. The separation of 21 specimens of L. tebessae from 53 specimens of sympatric L. mauretanicus sp. nov. is mainly given by the higher setae numbers in the latter. A PCA considering all 16 standard NUMOBAT characters provides a complete separation of these 74 specimens. This classification is confirmed by a LOOCV-LDA with an error of 0% in 74 specimens after character reduction to CL/CW 900, PoOc/CL 900, nOcc 900, nSc 900, and nSt 900 in order to prevent character-number overfitting. Furthermore all four variants of NC-clustering confirm the PCA classification.