Confirmation of the presence of Ischnura senegalensis (Rambur, 1842) on the Canary Islands R. A. Sánchez–Guillén & A. Cordero–Rivera

Confirmation of the presence of Ischnura senegalensis (Rambur, 1842) on the Canary Islands.— The pres ence of one or two species of damselflies of the genus Ischnura in the Canary Islands has been a matter of debate in recent years. The first published records listed I. senegalensis as the only zygopteran inhabiting the archipelago, but this proved to be wrong, and until recently, all specimens of Ischnura captured in the islands were unanimously regarded as belonging to I. saharensis . Recent photographic evidence, however, is compatible with the presence of I. senegalensis. In this study, we give morphological and genetic evidence of the presence of I. senegalensis in the Canary Islands, and we discuss the importance of voucher specimens to correctly identify very similar species. Resumen


Introduction
The genus Ischnura is one of the most speciose genera of the family Coenagrionidae, with around 70 species with a worldwide distribution (Dumont, 2013). The Mediterranean is home to nine species: I. elegans (Vander Linden, 1820); I. evansi (Morton, 1919); I. fountaineae (Morton, 1905); I. genei (Rambur, 1842); I. graellsii (Rambur, 1842); I. intermedia (Dumont, 1974); I. pumilio (Charpentier, 1825); I. saharensis (Aguesse, 1958); and I. senegalensis (Rambur, 1842) (Boudot et al., 2009). Given the inherent dynamic nature of animal distributions, the species found at particular localities are expected to change over time, especially in a scenario of global warming and range expansions (Hickling et al., 2005;Sánchez-Guillén et al., 2013, 2014b. To date, 14 species of Odonata (three Zygoptera and 11 Anisoptera) have been recorded in the Canary Islands, including one Macaronesian endemic genus (Sympetrum nigrifemur, also found in Madeira) (Weihrauch, 2013). However, only one zygopteran has been reported regularly on the islands. It was first identified as Ischnura senegalensis by Valle (1955), but this identification was questioned by Hämäläinen (1986), who examined Valle's specimens preserved in the Zoological Museum of the University of Turku, and confirmed that they belong to Ischnura saharensis. For many years, the Canarian Ischnura was considered to be represented by only one species, namely I. saharensis, and consequently the monograph by Báez (1985) lists it as the only species of Zygoptera in the Archipelago.
In most of the Mediterranean area, I. elegans and I. pumilio occur in sympatry. Other species overlap locally: I. elegans and I. graellsii in Spain, I. elegans and I. genei in the islands of Elba and Giglio (Italy), and I. graellsii and I. saharensis in Morocco (Boudot et al., 2009 (Boudot et al., 2009). Given that I. senegalensis has been recorded from Mauritania (Boudot et al., 2009) and Cape Verde Islands (Aistleitner et al., 2008), its presence in the Canary Islands is possible due to the frequent winds coming from the desert. These winds have been invoked as the explanation for the only record of Platycnemis subdilatata for the Canaries (Kalkman & Smit, 2002).
In recent years, there has been an on-going discussion about the presence of I. senegalensis in the Canaries (Weihrauch, 2013), particularly because several pictures uploaded by amateurs to web servers suggested its presence. Nevertheless, scientific evidence needs voucher specimens (Corbet, 2000), because this is the only way to re-exam data in the light of future taxonomic knowledge. Our initial goal, therefore, was to obtain a sample of Canarian I. saharensis for molecular analyses, but we only found specimens which morphologically resembled I. senegalensis. The main purpose of this work was therefore to determine whether mtDNA sequencing would confirm that our Ischnura specimens were I. senegalensis. Additionally, we investigated the phylogenetic relationships of seven Mediterranean Ischnura species (I. elegans, I. fountaineae, I. genei, I. graellsii, I. pumilio, I. saharensis and I. senegalensis) and discuss our findings in the light of the recent review by Dumont (2013).

Material and methods
We collected five Ischnura samples on 20 V 2007 near San Andrés (Santa Cruz de Tenerife, Canary Islands) (28° 30' 51.26'' N and 16° 11' 31.53'' W). Three males and two gynochrome females, which we identified as For the molecular identification, we combined two mitochondrial markers, cytochrome oxidase II and cytochrome b, because the use of several genetic markers improves species identification (Bergmann et al., 2013). DNA of three samples of putative Ischnura senegalensis from the Canary Islands, and one to three samples of each of the seven Mediterranean Ischnura species: I. elegans, I. fountaineae, I. genei, I. graellsii, I. pumilio, I. saharensis, and I. senegalensis was extracted from the thorax using a standard phenol/chloroform extraction protocol (Sambrook et al., 1989). Additionally, we included three samples of I. asiatica and one of Enallagma basidens, E. cyathigerum and Telebasis salva as outgroups (Gene bank accession numbers: AF067669.1, AF067670.1, AF067681.1, AF067689.1, AF067690.1 and AF067701.1). Table 1 gives the localities of capture of these samples and the accession numbers. Extracted DNA was amplified by PCR for part of cytochrome oxidase II (673 bp), with the primers TL2-J-3037 and C2-N-3494 and C2-J-3400 and TK-N-3785 (Simon et al. 1994) (Gene bank accession numbers KC430114-KC430232), and cytochrome b (457 bp) with the primers CB-J-10933 and TS1-N-11683 (Simon et al., 1994) (Gene bank accession numbers KC430114-KC430232). The PCR program had an initial cycle of 95ºC for 3 min, followed by 34 cycles at 95ºC for 30 s, with annealing for 45s, an elongation phase at 72ºC for 45s, and   Clade pumilio 100 a final extension phase at 72ºC for 10 min. It was performed in 10 ml and amplification conditions were as follows: 1-2 ng of DNA (2 mL), 5.0 mL of 2X Ready Mix TM PCR Master Mix (1.5 mM MgCl 2 ), 1mL 10Î of BSA, 0.3 mL of MgCl 2 (50 mM), 1.1µL of distilled water, and 0.3 mL of each primer (10 pmol) in a 'GeneAmp PCR system 2700' thermocycler (Applied Biosystems). Bidirectional sequencing reactions were conducted using the Bigdye TM terminator cycle sequencing kit (Applied Biosystems) using the automatic sequencer ABI3100. Forward and reverse sequences were edited and aligned with Clustal X (Thompson et al., 1997) implemented in Mega version 6 (Tamura et al., 2013). Variable positions were revised by eye, and only high quality sequences were considered.
We generated a neighbour-joining tree using Mega version 6 (Tamura et al., 2013) by using the consensus sequence of mtDNA cytochrome oxidase II and Fig. 2. Consensus neighbour-joining tree based on Kimura 2-parameters (K2P) with the rate of variation between sites being modelled with a gamma distribution (shape parameter = 1). Branches corresponding to partitions reproducing more than 50% of bootstrap replicates are shown.

Results
All five samples we collected were morphologically identified as I. senegalensis. Figure 1 shows the male anal appendages in the posterior view of an immature (A) and a mature male (B), both from Tenerife, and samples of I. senegalensis from Namibia (C) and I. saharensis from Morocco (D) for comparison.
For the DNA analysis, after deleting gaps and missing data, we obtained a total of 660 informative positions in the final dataset. The resulting neighbour joining tree ( fig. 2) grouped our sample of Ischnura species in two main clusters. In the first cluster, one branch included all species of the I. elegans group and I. fountaineae. The second cluster includes all the samples of I. senegalensis, irrespective of their geographical origin. Therefore, our morphological identification was confirmed by the analyses of DNA sequences.

Discussion
The five Canarian samples were similar to the samples of I. senegalensis from Namibia although they were larger. The easiest way to distinguish between I. saharensis and I. senegalensis is the different orientation of lower appendages, which are clearly divergent in saharensis (as is typical of the Ischnura elegans group, to which it belongs; Dumont, 2013; see also below) and convergent in senegalensis ( fig. 1).
The neighbour-joining tree ( fig. 2) of the eight Ischnura species showed two main clusters, with all species of the I. elegans group (I. elegans, I. genei, I. graellsii and I. saharensis) and I. fountaineae in the first cluster. However, not all species formed monophyletic groups in our analysis. This is expected for taxa which have recently diverged, and is consistent with the hybridization processes between I. elegans and I. graellsii in Southern Europe (Sánchez-Guillén et al., 2011), I. graellsii andI. saharensis in Morocco (Sánchez-Guillén et al., 2014a), and I. elegans and I. genei in the Tyrrhenian Islands Elba and Giglio (Sánchez-Guillén et al., 2014a). The second group of the first cluster included all I. senegalensis samples, those from Namibia and Japan, and also from the Canary Islands. Finally, the second cluster included I. pumilio samples from the Azores and Germany, and I. asiatica from Japan. Our results are consistent with the tree constructed by Dumont (2013) using cytochrome oxidase I and ITS DNA fragments. Ischnura pumilio and I. elegans appear separated in the two main clades that form the genus, and I. elegans, I. genei, I. graellsii, and I. saharensis appear as closely related species.
In a recent paper, Peels (2014) presented photographs of Ischnura individuals from the Canary Islands, morphologically identified as I. senegalensis. He also provided useful characters that allow the two Ischnura species found in the islands to be separated. Our research highlights the importance of voucher specimens in taxonomy and ecology (Corbet, 2000). Had no specimens been captured by Håkan Lindberg in 1949, studied by Valle (1955) and preserved in a museum, their identity would never have been confirmed. We agree that restraint should be exercised when collecting insects, or other animals, and that photography is preferable for identification purposes. This is indeed the safe approach for some species, particularly when the number of possible species is low (Peels, 2014). For instance, the first record of Erythemis vesiculosa for the Galapagos islands is based on an unpublished picture examined by expert taxonomists (Muddeman, 2007). Photo-identification is also possible for species that have a unique appearance, like Ischnura nursei, and Indian species recently found in the United Arab Emirates, but even in this case a voucher specimen was collected (Feulner & Judas, 2013). However, as photographs can not be used to extract DNA or to conduct a detailed study of morphology, they are of limited use in science. We do not know where I. senegalensis came from, although it likely colonized from northwestern Mauritania where the nearest populations of the species are found (Boudot et al., 2009). This question might be answered by a detailed genetic analysis, including samples from different geographical regions. Again, voucher specimens would be needed to do so.