Bumblebee species (Bombus Latreille, 1802) were initially identified by their colour patterns. However, different species can show convergent colour patterns and furthermore, one given species can display a high intraspecific variation in different geographical locations (Williams, 2007). This results in the presence of cryptic species very similar or identical in morphology which cannot be readily distinguished (Rasmont, 1984). To overcome this, a molecular approach based on the sequence variation of a region of the mitochondrial cytochrome c oxidase subunit I (cox1) gene known as DNA barcode, has proved useful for revealing species identical to those recognised by traditional methods (Schmidt et al., 2015). DNA barcodes are included together with other data in an integrative taxonomy approach to delimit, discover and identify meaningful, natural species and taxa at all levels (Will et al., 2005), and have been widely used in bumblebee identification (Williams et al., 2012).

Within Bombus, the subgenus Bombus s. str. has presented more problems than other subgenera when it comes to species identification (Williams et al., 2012). In Western Europe, only two species were widely accepted until the mid-twentieth century: B. terrestris (Linnaeus, 1758) and B. lucorum (Linnaeus, 1761). However, a detailed review of the species B. lucorum based on morphological methods led to the confirmation of the species B. magnus (Vogt, 1911) by Krüger (1954) and B. cryptarum (Fabricius, 1775) by Rasmont (1984). Their status as species was supported by cross-breeding experiments (De Jonghe & Rasmont, 1983) and mating observations that indicated reproductive isolation between them (Bučánková et al., 2011). Later, molecular data and male labial gland secretions contributed enough evidence to accept all three as separate species (Bossert, 2015; McKendrick et al., 2017). This grouping is now known as the B. lucorum complex.

Nevertheless, there are not enough morphological characteristics to discriminate the species reliably. The first collar may be the only characteristic that allows discrimination of the three species (Williams, 2000; Bossert, 2015), but it is only useful for queens. Ecological data cannot distinguish the species as they live sympatrically, and there are no comparative studies about their habitat preferences over a wide geographic area (Bossert, 2015). Consequently, Rasmont et al. (2015) has considered B. magnus as a taxonomically problematic species.

In the Iberian Peninsula, only two species from this complex are known to be present: B. lucorum and B. magnus. Bombus magnus in particular has been located in the northern half of the Iberian Peninsula: between altitudes of 100 and 2000 m in the Cantabrian Range, the Pyrenees, Sierra de Guadarrama and provinces of Guadalajara and Teruel (Rasmont, 1984; Castro, 1996; Ornosa et al., 2017). However, these reports have not been confirmed with molecular assays and have been the topic of further discussion (Williams et al., 2012; Ornosa et al., 2017). Only one report from a specimen collected in 1965 in Soria (also in the northern half of the Iberian Peninsula) has been confirmed with DNA barcoding (Williams et al., 2012).

The lack of certainty in the B. magnus records from the Iberian Peninsula south of the Pyrenees (Bossert, 2015; Ornosa et al., 2017) reveals the need for further sampling to clarify the distribution of this species in the Iberian Peninsula. In addition, specimens from the Iberian Peninsula are of particular importance since there are indications that queens of B. lucorum exhibit a yellow thoracic collar coloration similar to B. magnus queens in central Spain (Bertsch, 2009). Bombus lucorum queens in Spain can also be exceptionally large and this could lead to wrong identifications as B. magnus queens are supposedly larger, although specimens with medium and small size have been also observed (Bertsch, com. pers., 2017). In this study, we used the DNA barcode to molecularly identify individuals of this species complex. Although this species is not considered to be threatened in the IUCN Red List of European Bees (Rasmont et al., 2015), distribution models project a reduction of suitable areas by 2050, therefore an update of the distribution of B. magnus in the Iberian Peninsula is needed to propose conservation measures.

Bumblebees were sampled from different areas of the northern half of the Iberian Peninsula between the summers of 2013 and 2017 (Table 1). Individuals (N = 113) of different castes (Seven queens; 47 workers; 59 males) were preserved in absolute ethanol until analysed.

The posterior left leg was removed to extract the DNA with the Chelex method (Walsh et al., 1991). A fragment of the mitochondrial gene cox1 of 614 bp was amplified with MyTaq™ Red Mix (Bioline) using two different methods. For the samplings between 2013 and 2016, we used the primers LCO/HCO (Folmer et al., 1994) and the following program: initial denaturation at 94 °C for 3 min; 40 cycles of 94 °C for 30 s, 48/50 °C for 30 s, 72 °C for 1 min, and a final extension step at 72 °C for 10 min. For the sampling of 2017, we used the mini-barcode primers Barbee and MtD9 (Françoso & Arias, 2013) and the following program: initial denaturation at 94 °C for 5 min; 35 cycles of 94 °C for 1 min, 46 °C for 1 min 20 s, 64 °C for 2 min, and a final extension step at 64 °C for 10 min.

PCR products were sequenced in Secugen (Madrid). Sequences were edited with Geneious 7.1 and species identification was performed by BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi).

Both amplification methods amplified the same region of the gene cox1. DNA barcodes of two individuals showed an identity of 100% with the sequences GU705915 (B. magnus from Germany) and JN872621 (B. magnus from Denmark) from Genbank, therefore of 113 individuals sampled, 111 were found to be B. lucorum and two individuals firstly identified as B. lucorum by morphometrical data were found to be B. magnus. One of the individuals was a male caught in San Nicolás de Bujaruelo (Huesca) in the Pyrenees, at 1591 m above sea level, 9 VIII 2014 (P. De la Rúa leg.). The second individual was a worker sampled in Naveros de Pisuerga (Palencia), at 800 m above sea level, 20 VI 2017 (C. Ornosa leg.), foraging on Rubus ulmifolius Schott (Fig. 1).

These new reports of B. magnus confirm the current distribution of the species in the Iberian Northern Plateau. The fact that only two individuals were sampled in four years strongly suggests that the species is rare and much less abundant than B. lucorum. Although both species are more widely distributed in northern Europe, previous works have already reported that B. lucorum is more frequent than B. magnus in mainland Europe (Bossert, 2015).

These reports agree with previous studies, which have pointed to a patchy distribution for B. magnus, with an association to heathlands with low diversity of plant species in Europe (Bossert, 2015) and other Mediterranean landscapes in the Iberian Peninsula, where it was previously reported based on morphological data. These results highlight the need for sampling more individuals in the Iberian Peninsula, and the utility of the DNA barcode region for identifying them. Obtaining more samples could help pinpoint both the period of the year when B. magnus nests are active and their habitat, facilitating the discrimination of the species from B. lucorum. It could also help to assess its conservation status in Spain which is unknown at present.