Background and aims – Diatoms from tropical Central Africa are rarely studied and documented. Waters of the Congo Basin are often acidic and thus may be dominated by diatoms from the genus Eunotia, usually found in the benthos and periphyton, but rarely in the plankton. We investigated planktonic Eunotia species from the Eunotia asterionelloides complex to determine their identity and distribution.

Material and methods – Samples were collected in fishponds in the vicinity of Kisangani, D.R. Congo, from herbarium material collected from the Bonkele River, D.R. Congo, and in the Kouilou River, Republic of the Congo. These samples were prepared for diatom analysis, examined using standard methods, and documented using both light and scanning electron microscopy.

Key results – Several taxa belonging to the Eunotia asterionelloides complex were observed. Taxa from this complex are recorded in literature as either E. asterionelloides or E. zasuminensis. The taxon from fishponds in the D.R. Congo was identified as E. tukanorum, previously described from Brazil. The structure of the valve outline and raphe appear similar, particularly the raphe position and length. A second species was found on aquatic macrophyte (Utricularia and Nymphaea) herbarium specimens from the D.R. Congo and a third species from the Kouilou River in the Republic of the Congo. These latter two taxa superficially resemble E. zasuminensis. Although similar in valve outline, the taxa differ morphologically. The shape of the apices in larger cells is indented to a larger degree, and the raphe differs in its structure, particularly the shape of the terminal endings. The structure and position of the rimoportula also differ. Also, one of these taxa is sometimes strongly flexed around the apical axis. Based on morphology, we consider these taxa distinct from E. zasuminensis and from each other, and thus describe Eunotia globicephala sp. nov. and Eunotia magnaparva sp. nov.

diatoms, D.R. Congo, Congo, Eunotia zasuminensis, Eunotia tukanorum, Eunotia magnaparva, Eunotia globicephala, new species, taxonomy, tropical Africa

Diatom studies, and those encompassing many other groups of the aquatic tropical fauna and flora in Central Africa, are sporadic and rather rare given the scale of the territory and the number and extent of water bodies found in this region. In 2010, a large expedition carried out extensive surveys along sections of the Congo River to study the biodiversity in non-protected areas. Known as the Belgian-Congolese Boyekoli Ebale Congo 2010 expedition, 250 km of the Congo River and some of its tributaries between Kisangani and Bumba were surveyed (Nagy et al. 2011; Virgilio et al. 2011; Cocquyt et al. 2013). This initial expedition gave the impetus for various other studies that followed in the Democratic Republic of the Congo (D.R. Congo) and the results are gradually appearing in the scientific record. Moreover, a biodiversity centre (“Centre de surveillance de la biodiversité”) was established at the University of Kisangani (https://centresurveillancebiodiversite.org) to continue this work.

Of particular importance is the study of the eunotioid diatoms from the region. The waters of the Republic of the Congo (Congo) and D.R. Congo are generally acidic “black waters” coloured by humic substances (except for the Congo River and its major tributaries). These waters are often dystrophic but may range in quality from oligotrophic to eutrophic. Acidic water bodies, as in other regions in the world, are often dominated by Eunotia Ehrenb. and related taxa, which are adapted for survival under these low pH conditions. Investigations of material from the D.R. Congo have already led to the description of new species within this genus, e.g. Eunotia leonardii J.C.Taylor & Cocquyt (Taylor et al. 2016), E. pierrefuseyi J.C.Taylor & Cocquyt (synonym of E. papilio var. africana (Fusey) Fusey nom. inval. and E. fuseyi J.C.Taylor & Cocquyt nom. illeg.) (Taylor et al. 2016), and E. rudis Cocquyt & M.de Haan (Cocquyt et al. 2016), and to a detailed discussion on the morphological variability within E. zygodon Ehrenb. (Taylor et al. 2016). Moreover, Eunotia enigmatica L.F.Costa & C.E.Wetzel, recently described from the Amazon Basin in South America was observed in D.R. Congo and documented by Cocquyt et al. (2019).

Most of the recent work on the diatoms of Central Africa has focussed on benthic diatom taxa; this is in contrast to many of the studies on large lakes in Africa, which focussed on the phytoplankton. In the present work, we examined samples containing planktonic Eunotia that form colonies in order to maintain their position in the water column. Several cells link at the corners forming irregular colonies that provide some resistance to sinking. This complex, broadly referred to as the Eunotia asterionelloides Hust. complex, has been discussed at length by Wetzel et al. (2010).

Wetzel et al. (2010) provided a systematic review of the E. asterionelloides complex and discussed the various challenges encountered while studying this complex. They provided illustrations of a population of E. zasuminensis (Cabejsz.) Körner from Finland, an important contribution to augment the illustrations given by Körner (1970) from a population of Lillsjön in Sweden, as the original slide and samples from Lake Zasuminskie in Poland have been lost. Wetzel et al. (2010) also briefly discussed the taxa in this complex occurring in Africa (e.g. Cholnoky 1958), expressing some doubt about the fact that Körner (1970) had synonymised Asterionella africana Cholnoky with E. zasuminensis. Given the morphology of the illustrated cells, both taxa seem to represent distinct taxa. Unfortunately, both the material and slides of Cholnoky have been lost, making it impossible to determine the identity of the taxon he observed. Wetzel et al. (2010) also expressed some doubt regarding the distribution records for E. zasuminensis within Africa, which the present work bears out. Records of the African distribution of the taxa in the Eunotia asterionelloides complex are given in Table 1. All records remain doubtful until the material is further examined/resampled.

Diatom material from herbarium sheets held at Meise Botanic Garden (BR) was investigated as part of a study to infer past water quality in the region (Okito et al. 2021). During this study, a number of cells similar in morphology to Eunotia asterionelloides were noted and it was decided to study them further. Sampling in the fishponds of Ngene-Ngene, Kisangani (D.R. Congo) was done for a preliminary study of their diatom community and diversity to determine the suitability of these habitats for bachelor and/or master thesis topics at the University of Kisangani (UNIKIS). These fishponds are owned by UNIKIS and are used for experiments to study fish productivity.

The broader Kouilou Plain and neighbouring Noumbi Basin in Congo were surveyed as part of an environmental impact assessment to record a freshwater baseline for the area in 2012. This area shows a high freshwater habitat heterogeneity, ranging from littoral forest and mangroves along the coast, to swamp forests, forest and savannah streams, lakes and large rivers inland. Epiphytic diatom material was collected at freshwater sites to compliment traditional water quality sampling, with the intention to use diatoms as a tool in future monitoring and evaluation studies of the area. These samples were sent to the South African National Diatom Collection (SANDC) after the studies were completed.

Based on our observations of the material from D.R. Congo and Congo, we conclude that the specimens from the Ngene-Ngene fishponds in D.R. Congo have a morphology consistent with that of E. tukanorum C.E.Wetzel & D.C.Bicudo. We however consider that the valves observed in the material from D.R. Congo on the herbarium material to have morphological characteristics that distinguish it from E. zasuminensis, including the shape of the apices, structure and position of the raphe, and the position of the rimoportula, and it is therefore described here as Eunotia magnaparva J.C.Taylor & Cocquyt sp. nov. Similarly, the species observed in the lake outflow in Congo can be distinguished from both E. zasuminensis and E. magnaparva and is described here as Eunotia globicephala J.C.Taylor, Cocquyt & G.Walsh sp. nov.

Material used in this project was collected from Central Africa from both the D.R. Congo and the Republic of the Congo. samples were collected from a variety of habitats including fishponds and the outflow of a lake. In addition, samples originally collected in 1936 and 1946 from standing waters in the D.R Congo were taken from herbarium sheets with preserved aquatic macrophyte plants. These herbarium sheets are housed at Meise Botanic Garden (BR). Details of accession numbers and sampling localities are given below.

Eunotia globicephala sp. nov. and E. magnaparva sp. nov. are morphologically most closely related to E. zasuminensis described originally from a Polish lake (Cabejszekówna 1937). The type material of E. zasuminensis has been lost. However, Wetzel et al. (2010) examined and illustrated a population of this taxon from Finland using both LM and SEM observations. Using these observations, we can differentiate both novel taxa from E. zasuminensis based on the following morphological characteristics (summarised in Table 2): the position of the raphe is closer to the apices than in E. zasuminensis, the raphe is consistently longer in E. globicephala than that of E. zasuminensis. The raphe of E. magnaparva is rather variable but longer in smaller cells. The shape of the apices differs from E. zasuminensis, cordiform with a larger swelling of the apex to the ventral side in E. globicephala and cordiform with a deeper cleft in E. magnaparva. In E. zasuminensis, the rimoportula is located in the cleft of one bilobate apex, whereas in E. globicephala and E. magnaparva the rimoportula is located towards the ventral margin near the helictoglossa. Eunotia zasuminensis has twisting around the apical axis whereas, however, in E. globicephala the degree of twist (up to 90° in some instances; Fig. 5G, N) seems to be higher. The helictoglossa is much reduced in E. zasuminensis and much more pronounced in both E. globicephala and E. magnaparva. All three taxa have a slight swelling of the valve in the centre of the cell as well as some degree of dorsiventral symmetry.

Eunotia is a genus in which considerable morphological variation can be noted over a diminution series from larger cells to smaller cells. One such example, Eunotia zygodon, is illustrated and discussed by Taylor et al. (2016). In the Eunotia asterionelloides complex, Wetzel et al. (2010) also illustrate similar distinct changes in size over the cell cycle of E. tukanorum. At the extremes of such a diminution series, cells may appear entirely distinct from each other in terms of apex shape, symmetry, and a variety of other characters. For this reason, we illustrate an exhaustive diminution series of Eunotia magnaparva (Fig. 1A–AS) based on as many cells as we could observe from a single sample. In addition, other samples were observed and comparable results were obtained (not illustrated here). These additional samples demonstrated a similar size range and continuity from larger to smaller cells. This series is also supported by a selection of SEM micrographs (Figs 2–4) that show what we consider to be key defining characteristics for this taxon. We have paid attention to the striation of the apices in terms of angle, curvature, and how the striae are interrupted, all of which remain constant over the cell cycle. In addition, the number and position of the rimoportulae remains constant. The occlusion of the round areolae by hymenes is also consistent in structure between the large and smaller cells.

As to why raphe length differs over the cell cycle, it could be surmised that this phenomenon is connected to the amount of contact the cells will have with a substrate. The majority of taxa within the Eunotia asterionelloides complex are considered planktonic with cells linking to form branched chains that remain suspended in the water column (Wetzel et al. 2010). The raphe is a structure generally accepted to be associated with diatom motility, while suspended in the phytoplankton cells would need to have little or no motility. As the cell cycle advances and cells are reduced in size, surface area and thus buoyancy would decrease. As buoyancy decreases, cells will be more likely to settle out of the water column, come into contact with a substrate and at this point motility may be useful in order that cells could then move towards the light. Regarding the presence of a raphe in planktonic Eunotia, Furey (2021) discusses at length the almost complete reduction of the raphe in some taxa. Furey (2021) postulates that although movement in such environments would be minimal the presence of a raphe indicates that movement would be possible. As discussed above, this becomes particularly important for smaller (shorter) cells which may gain a competitive advantage by having a larger raphe system and being more motile. Although the type sample (CCA 4521) for Eunotia magnaparva was collected from standing waters, it seems that the substratum (in this case Utricularia) possibly provided a habitat for small motile forms of this species. This would have to be tested though by sampling true open water populations vs near shore or benthic populations. There are enough stable characters (e.g. stria density, structure and position of the rimoportulae, structure of the striae at the apices, disjunct striae, etc.) to demonstrate that all the cells in the diminution series belong to a single species. If, for example, the cell illustrated in Fig. 1A and Fig. 1AO had been observed in separate samples, it would have been easy to assume that these are distinct species not related to each other, especially with LM observations.

Collecting diatoms from herbarium samples has proved interesting in the past for reconstructing environmental conditions (Yallop et al. 2009; Okito et al. 2021) but as the present study has shown they may also be a valuable source of diatom samples for taxonomic work. Dried herbarium sheets would in most cases provide an ideal environment for diatom preservation as the silica cell walls of the diatoms on dried material can persist indefinitely in a protected environment.

Sampling of the fishponds at Ngene-Ngene, Kisangani was done in the framework of the project Apport de la biodiversité à la formation et la sécurité alimentaire dans le Bassin Nord-Est du Congo (Kisangani, R.D. Congo) financed by the Vlaamse Interuniversitaire Raad – Universitaire Ontwikkelingssamenwerking (VLIR-UOS). The South African Institute for Aquatic Biodiversity is the custodian of the South African National Diatom Collection (SANDC) and provides financial support for the running and day-to-day operation of the collection.