Pimelia (Amblypteraca), Mas-Peinado, Buckley, Ruiz & Garcia-Paris, 2018

Mas-Peinado, Paloma, Ruiz, José L., Merkl, Ottó, Buckley, David & García-París, Mario, 2021, Taxonomy of the North Moroccan and Iberian species of the subgenus Amblypteraca (Coleoptera: Tenebrionidae: Pimeliinae: Pimelia), Zootaxa 4963 (3), pp. 457-482 : 469-471

Systematics of Amblypteraca

The analysis of the molecular and morphological data sets yielded quite similar views, with populations of Larache, Assilah, and Cádiz representing a fairly homogeneous unit corresponding to the mitochondrial group V, different to a point from the populations of Troia (mtDNA group IV), Tleta Rissana (mtDNA group IV) and La Mamora (mtDNA group III). All these populations are well differentiated from those from Sidi Boughaba (mtDNA group II), and also from Sidi Ifni —Sidi Boughaba (mtDNA group I). Morphological differentiation of Tleta Rissana and La Mamora specimens is relatively well marked, also compatible with the subdivision represented in their mtDNA and mtDNA- ITS2 ( Fig. 6 ).

In these groupings, however, there were some incongruences between the nuclear (ITS2) and mtDNA markers, suggesting the potential existence of gene exchange and/or mitochondrial introgression between some of those mitocondrial-morphological groups, or potential issues of incomplete lineage sorting (ILS) among markers. The nuclear tree showed that one specimen from Troia was nested among Assilah and Cádiz specimens, and was not sister to the other specimens from Troia. In addition, the specimen from Larache clustered together with specimens from Tleta Rissana, instead of being clustered with Assilah and Cádiz as they were in the mtDNA tree. Discordances between mtDNA and nuclear markers in a few specimens may suggest some gene exchange between mtDNA groups III, IV and V (Larache-Tleta Rissana; Troia-Cádiz/ Assilah). Specimens of La Mamora from mtDNA group III seem to retain their morphological and genetic identity, but they could differ by up to three mutations (ITS2) with respect to other populations of mtDNA groups IV and V. Since samples from La Mamora consisted of only three specimens, it is not unlikely that gene exchange could be effective among neighboring populations of clades IV and V. Consequently, we tentatively retain in a single evolutionary unit (species) all populations included in mtDNA groups III, IV and V. There is complete marker congruence, and no evidence of gene flow between clades A (I), B (II), and C (III-IV-V), and therefore with the data at hand we consider the existence of three main evolutionary units (species) within Amblypteraca : Clades A, B and C.

Geographic correspondence between patterns of morphological and mtDNA variability suggests that the potential gene exchange within Clade C must have been very limited or that it is relatively recent. The close geographic proximity between Larache and Tleta Rissana, and the apparent absence of environmental barriers, makes quite likely the existence of current gene flow between them. These two population groups were probably separated by the Lukus (or Loukkos) river during the Pleistocene. Nowadays, the Lukus river forms an estuary in the area between Larache and Lixus which may have favoured the connection between both populations ( Gharbaoui 1981; El Morhit et al. 2013). Specimens of Amblyteraca that inhabit sandy soils might be easily changing river shores as the flowing patterns changed when filled with sediments. On the other hand, the morphological differentiation of the Troia specimens with respect to other sequenced individuals is quite evident; however, these specimens are similar to those from populations of Ksar es Seghir (= Alcazarseguer), located at the beaches between Tangier and Ceuta (northern tip of the Tingitanian Peninsula). Ksar es Seghir was taken by the Portuguese in 1458 and remained for about 100 years as a main center for communications and transport between Lisbon and the Northern African Portuguese settlements ( Braga 1998). An appealing explanation for the current presence of this Amblypteraca lineage in continental Portugal, near Lisbon, is that it is the result of a fortuitous human-mediated introduction from populations located near Ksar es Seghir. Unfortunately, these populations are mostly gone because of the construction of Tanger-Med, a mega-portuary infrastructure, and we were not able to get specimens to test this hypothesis. Lack of congruence between nuclear and mtDNA in one specimen from Troia may suggest that the Portuguese specimens were in contact with coastal Moroccan populations, likely before to their arrival to their present location in Portugal.

Considering that gene flow is apparently limited, and that populations still retain their morphological singularity, we consider that the evolutionary unit corresponding to Clade C (mtDNA groups III, IVa , IVb and V), can be subdivided in four subspecific entities. By doing this, we retained an evolutionary species concept within an integrative taxonomy framework ( Padial et al. 2010) for the combined mtDNA groups III, IV and V, but we also considered that the population units of La Mamora, Troia, Tleta Rissana, and Larache-Assilah-Cádiz, might become isolated and independent from each other in the future. The consideration of these morphological groupings as subspecific units was already proposed by Koch (1941) and partially by Antoine (1949, 1951).

Figures

FIGURE 6. Bayesian phylogenetic tree based on the concatenated cox1, 16S and ITS2 (1580bp). Geographic location for lineages within Clade C [Pimelia (Amblypteraca) chrysomeloides Pallas, 1781] are shown in the map. Support for each node is provided by the posterior probabilities (PP) and ultrafast bootstrap support values (uBS).