Pollimyrus species

4.5 | Biogeography of the Pollimyrus species diversity

Based on the current study of most available type specimens, a first overview of the biogeography of the known Pollimyrus species has been summarized (Table 8). This overview is only preliminary and needs more verified non-type specimen distribution data. Nevertheless, it shows that, with 11 species, the highest species diversity in Pollimyrus can be found in the Congo IP, whereas only a few species are found in the Nilo-Sudan (two or three), Lower Guinea (two or three), Zambezi (three; but see Schedel et al., 2024 for a potential new species), East Coast (one), and Great Lakes (one) IPs. Except for their absence from the Angolan and Ethiopian Rift valley IPs, this known distribution for Pollimyrus coincides with that of the family Mormyridae as a whole ( Lévêque & Paugy, 2017). Noticeable is also the dominance of slender-tailed species in the Congolese IP, with seven(eight) slender-tailed versus two(three) thick-tailed species. In addition, slender-tailed species only occur in the Congo IP, except for P. adspersus , whose original collecting locality remains unclear ( Günther, 1866). The precise type locality of P. adspersus is unknown and therefore placed, tentatively, in three different IPs in western Africa. On the contrary, thick-tailed species occur in most IPs from which Pollimyrus species have been reported, but only in low numbers.

In some species complexes and morphological groups, the lack of a clear morphological distinction between type specimens of nominal species complicated the species delineations. However, in some cases, multivariate statistical analysis (PCAs) detected subtle differences in morphology between (hydro)geographically distinctly occurring specimens. This was observed, for instance, in the P. pulverulentus , the P. tumifrons , and the P. stappersii / P. castelnaui species complexes (see Results). These examples show that the Pollimyrus alpha-taxonomy might be strongly linked to allopatry, although, at present, the distribution ranges of several species in the Congo basin seem to overlap, such as those of P. tumifrons , P. pulverulentus , and P. schreyeni (Figure 2).

As this summary was compiled using mostly data from type specimens only and, as a result, does not take into account the full geographical distribution of the species identified, it could be possible that more species co-occur in some of these IPs. So far, this is unknown, and each species seems to be restricted to one IP due to the low number of specimens available. Remarkably, P. krameri is the first Pollimyrus species described from the East Coast IP. So far, only

IP Thick-tailed morphological group Slender-tailed morphological group Other morphological groups Total

Nilo-Sudan Pollimyrus petricolus Pollimyrus adspersus ? Pollimyrus isidori (thick tail) 2 or 3 Upper Guinea P. adspersus ? 0 or 1

Lower Guinea Pollimyrus guttatus P. adspersus ? Pollimyrus vanneeri (thick tail) 2 or 3

Congo Pollimyrus brevis Pollimyrus ibalazambai Pollimyrus fasciaticeps (thick tail) 11 Pollimyrus stappersii Pollimyrus maculipinnis Pollimyrus tumifrons (slender tail) Pollimyrus osborni Pollimyrus pedunculatus ( Cyphomyrus plagiostoma ) Pollimyrus pulverulentus Pollimyrus schreyeni

Zambezi Pollimyrus castelnaui Pollimyrus marianne Pollimyrus weyli

East Coast Pollimyrus krameri Great Lakes

3

1

Pollimyrus nigricans (thick tail) 1

Note: Due to the unknown type locality of P. adspersus in “West-Africa,” this species is tentatively placed in three IPs from this region.

P. nigricans is known to occur in one of the great lake basins in Africa, namely the mouth of the Katonga River at Lake Victoria (Great Lakes IP). As the type specimens from P. isidori and P. nigricans are morphologically similar and were found in the Nile River and Lake Victoria basin, respectively, two systems that are hydrologically connected, although the Ripon Falls and Owen Falls separate the Lake from the Nile basin (with presently the Nalubaale Dam), both these species could be phylogenetically closely related. Junior synonyms of P. isidori have been described from several river basins, P. isidori gaillardi from Lake Chad (Nilo-Sudan IP), P. isidori rudebeckii from the Gambia River (Nilo-Sudan IP), and P. isidori vanderbilti from the Central African Republic (Nilo-Sudan/ Congo IP). However, these junior synonyms were not included in the present morphological study, as the main scope of this study was to work with the valid species. Therefore, it remains unclear whether these junior synonyms are indeed conspecific with P. isidori . If correct, that would mean that the geographic range of P. isidori is the largest known range for a Pollimyrus species so far, and that one species can occur in multiple river basins and IPs. However, further study is needed to assess the conspecificity and re-evaluate the distribution of these nominal species. It is to be noted that other nominal species once thought to be junior synonyms of P. isidori occurring in different IPs are now recognized as valid species, that is, P. fasciaticeps and P. osborni from the Congo IP.

The little amount of data on the distribution of species, combined with the small number of specimens known per species, which often come from the same river(basin), might indicate that Pollimyrus species, in general, are not very abundant. Due to their restricted distribution, they might be more vulnerable than more common and widespread species. However, for most species (15), the IUCN Red List ( IUCN, 2024) categorized them as Least Concern (LC) ( P. adspersus , P. brevis , P. castelnaui , P. isidori [including P. osborni and P. fasciaticeps as subspecies], P. marianne , P. nigricans , P. nigripinnis , P. petricolus , P. pulverulentus , P. schreyeni , P. stappersii , P. tumifrons , and C. plagiostoma ), as there is little indication for any threats or population declines ( Diouf, 2020; FishBase team RMCA and Geelhand, 2016; Moelants, 2010a, 2010b, 2010c, 2010d, 2010e, 2010f, 2010g, 2010h; Olaosebikan & Lalèyè, 2020; Olaosebikan & Moelants, 2020; Tweddle & Marshall, 2007; Tweddle et al., 2019), whereas three others remain Data Deficient (DD) ( P. pedunculatus , P. guttatus , and P. maculipinnis : see Moelants, 2010i, 2010j, 2010k). The four new species were not assessed yet, but there are indications for severe environmental impacts on the type localities of P. krameri and P. weyli (R.B. personal communication). The available assessments, however, consider the range of the species based on identifications of non-types (e.g., see presence of P. adspersus in the Congo basin), which are, most likely, considering our presented results, not using the correct species identifications (also see Hopkins et al., 2007). As such, the current assessments might not be reflective of the species vulnerability (see also Palacio et al., 2023). Similarly, the distribution details provided for each species on FishBase ( Froese & Pauly, 2024) have been compiled from various existing publications. However, considering the numerous identification issues identified, the former distribution data should be handled with care. A revision of their current conservation status is certainly needed, which will only be possible after reidentifying museum specimens and summarizing the new resulting distribution for each of these.