Even if the blackfly fauna of La Gomera has already been examined several times (Crosskey, 1988; Crosskey & Báez, 2004; Reidelbach, 2004; Lüderitz et al., 2010), three holiday stays in 2013, 2016 and 2020 offered the possibility to also examine the preimaginal stages in selected flowing waters. All in all, 1207 individuals could the determined, the existence and distribution of which confirmed the results of the previous examinations (Table 1): in the longitudinal profile of the El Cedro stream, the hypocrenal areas as well as the upper course in the laurel forest are preferred by S. guimari Becker, 1908 and the following stream section is populated by S. intermedium Roubaud, 1906 and S. velutinum (Santos Abreu, 1922); S. tenerificum Crosskey 1988 mentioned in earlier examinations has been synonymised in the meantime (Adler et al., 2015). Outside the laurel forest, there is also S. ruficorne Macquart, 1838 which is able to even populate smallest trickles that are drying out such as in Barranco de Arure near El Guro (Table 1) (cf. Cherairia & Adler, 2018). Simulium pseudequinum Séguy, 1921 which is also mentioned for La Gomera could not be found as no samples were taken in the corresponding biotope, the lower stream sections of the lower altitudes.

The determination of the samples of 2013 coming from a spring brook at “Camino Forestal La Meseta”, further up Vallehermoso, came as a surprise: in addition to 20 S. guimari larvae, one young Rubzovia larva was found, which at first sight was determined as S. paraloutetense Crosskey, 1988. Due to its little size of 3.3 mm and the not yet developed pupal respiratory buds which correspond to the third instar (Harrod, 1964), the final determination was, however, postponed. Apart from that, it has been planned to visit the island again in the future aiming at finding more animals, particularly the pupae which are easier to determine.

The next stay on the island followed in February 2016 and was disappointing in as much as at the time of the sampling, the spring brook had high water level due to long lasting rainfall so that no successful sampling was possible (Fig. 1). This was particularly also due to the fact that the laurel leaves that had served as substrate in the first sampling had been washed away by the water current and the rock had also been uninhabited. In March 2020, the spring was finally visited once again within the scope of a Tenerife holiday, during a one-day trip to La Gomera. It is true that the hydrologic conditions were ideal (Fig. 2), the success of the sampling was, however, limited to the finding of one single pupa exuvia of S. guimari settling on a laurel leave lying in the water.

Rubzovia species are generally not only rare but also only few individuals of them are found (Crosskey et al., 1999). This fact is easy to confirm given the difficulties encountered in the previous, vain sample takings. As it was not clear how long it would take to make another finding, it seemed to be appropriate to intensively deal with the young larva once again to secure the faunistic importance of the finding from a taxonomic point of view.

The only existing determination key for the pre-imaginal stages of Rubzovia is based on the differentiation of the pupae (Crosskey et al., 1999). Based on own examinations and using the available literature data, Seitz et al. (2012) list the most importing diagnostic characters of all five known Rubzovia species (Adler, 2020) for the larvae in a synoptic table. As, however, a safe determination is based on examinations of the mature larvae with fully developed gill histoblasts, clear diagnoses are usually not possible when examining young larvae due to the lack of this characteristic. One thus tried to amend the table mentioned above by additional larval determination characteristics with the goal of ensuring the most exact and safe determination for S. paraloutetense possible (Table 2); Simulium vantshi Petrova, 1983 was not considered due to its range restricted to the central Asian Pamir Mountains of Tajikistan where the preimaginal stages settle in glacial streams (Petrova, 1983).

The most striking features which the larvae of the Rubzovia species have in common are the triangular (sagittate) posteromedian headspot as well as the small and/or missing postgenal cleft, which immediately strike the eye in the determination. There are minor differences between the species as regards the form of this headspot and as regards the extension of the cleft allowing for a rough classification of the species (Table 2). Comparison of these two characteristics immediately allows for the conclusion that our larva can be identified as S. paraloutetense (Figs. 3–4). While Crosskey et al. (1999) describe the important “median tooth of hypostoma” character which is important for many blackfly species in a slightly inaccurate way as “moderately long”, the “nipple-like projection” of the corner teeth of hypostoma described by these authors can also be confirmed for our larva (Fig. 5). In contrast, the corner teeth of hypostoma of S. joanae, S. knidirii and S. lamachi have a triangular form (Table 2).

The number of lateral serrations of hypostoma following the two paralateral teeth seems to be a useful diagnostic character which – for S. paraloutetense - have a pointed form. Due to its young age, our larva has three fully developed, pointed serrations, the remaining ones are only hardly or not visible at all, yet (Fig. 5, arrow). As with the other species, these serrations are blunt and rounded or completely missing, this allows for a differentiation from the other species (Table 2).

The inner preapical ridge of our young larva shows a large distal mandibular serration, which is long, triangular, and sharply pointed (Fig. 6). On one mandible this serration is followed by one subsidiary tooth, on the other mandible, there is moreover an additional, hardly visible small tooth; this corresponds to the characteristics of a mature S. paraloutetense larva. In contrast, S. knidirii has three subsidiary teeth, while with S. lamachi, the subsidiary tooth is closer to the serration and moreover bent. This simplifies the differentiation from the other species (Table 2).

The other diagnostic characters listed in Table 2 are typical meristic features, to which a certain range applies, depending on the individual. When it comes to the determination of the number of the primary fan rays of young blackfly larvae, there are more or less deviations from the reference value of the full-grown larvae (Schröder, 1987, 1988). Similarly, the number of the sublateral setae of hypostoma is clearly smaller with both, the young paraloutetense larva and the young S. joanae larva of identical size, which has been examined for comparative purposes, than the corresponding number of the mature larvae. Nevertheless, these two young larvae already have as many sublateral setae as mature S. knidirii and S. lamachi larvae.

For the number of rows of posterior circlet and the number of hooks per row it can be specified that with our larva, they were only developed about 80 percent due to its growth stage. This value complies with values that were for example also determined for corresponding young larvae of Prosimulium tomosvaryi (Enderlein, 1921) (Halgos & Jedlicka, 1973 sub. nom. P. nigripes).

Given the diagnostic characters described above, it is also possible to determine the exact species of young Rubzovia larvae with high probability by means of combination and exclusion. If posteromedian headspot, postgenal cleft, corner teeth and serrations of hypostoma are available collectively, S. paraloutetense can be differentiated from the other three species. The characters preapical ridge of mandible and sublateral setae of hypostoma moreover allow for the pairwise differentiation between the four similar species.

As to the ecological differentiation, it can be stated based on the cited literature information as well as on own examinations of S. lamachi that S. knidirii, S. lamachi and S. paraloutetense live in smallest spring brooks with little water and weak current which may already trickle away after a short section. In contrast to these species strictly bound to the hypocrenal, S. joanae is also found on Madeira in rhithral stream sections characterised by stronger current.

One decade after the first description of S. paraloutetense (Crosskey, 1988), the preimaginal stages were described (Crosskey et al., 1999). After another five years, however, it was reported that the only known aquatic finding spot was lost due to drying up and destruction (Crosskey & Báez, 2004). So from a faunistic point of view, the present finding is a stroke of luck as now, a second record could be determined for S. paraloutetense. The number of species determined on La Gomera is thus increased to six; consequently, the island has the largest species variety of all Canary Islands (López-Peña & Jiménez-Peydró, 2017).

Let us hope that for the existence of this species on La Gomera, the sampling site can be regarded as protected as the spring is situated within the “Parque Nacional de Garajonay”. The spring brook itself falls down a rock face already after few meters (Figs. 1–2); at the foot of which it originally accumulated and – after passing a trail – flowed down the valley. The major part of the discharge is, however, actually collected in the lower section of the small waterfall and led via a pipeline into a duct so that the lower section of the rock face is only kept wet by droplets.

As regards the water supply, it can be assumed that discharge is guaranteed throughout the year due to the North-East exposition favoured by the Trades and the related precipitation, which is – by the way – also confirmed by the existing syntopic taxa Dugesia gonocephala (Dugès, 1830) (Turbellaria) and Amphipoda.

All in all, the discovered situation at the spring largely corresponds to the descriptions of the habitat of S. paraloutetense on Gran Canaria stating that the spring there also showed a very low discharge, a short flowing section, North exposition and shading by trees; the larvae were mainly found in the phytal (Crosskey et al., 1999). To confirm more potential localities on La Gomera, it thus seems to be appropriate to look in the area of the laurel forest for very small, spring-fed trickles in a targeted manner.