Water mites from West Africa (Acari: Hydrachnidia)

Abstract. Ten new species of water mites are described from Ghana, i.e. Limnesia (Limnesia) stagnalis n. sp. (Limnesiidae), Hygrobatopsis (?) convexipalpis n. sp., Hygrobatopsis (Hygrobatopsis) pauciglandulosa n. sp., Hygrobates (Hygrobates) dentipalpis n. sp., Hygrobates (Hygrobates) pseudoniloticus n. sp. (Hygrobatidae), Neumania (Soarella) fusiformis n. sp., Neumania (Soarella) ghanaensis n. sp., Unionicola heversi n. sp. (Unionicolidae), Arrenurus (Megaluracarus) chutteri ankasa n. sp. and Arrenurus (Micruracarus) circulodorsalis n. sp. (Arrenuridae). Three subspecies are raised to full species, i.e. Mamersa expansa Cook, 1979 nov. stat. (Hydryphantidae), H. coriaceus (Lundblad, 1952) nov. stat. (Hygrobatidae) and Nyangalla (Ecpolopella) acuticaudata K. Viets, 1916 nov. stat. (Unionicolidae). Ecpolus dorsofenestratus Lundblad, 1949 is synonymized with Neumania (Alloneumania) marginata (K. Viets, 1916). The adults of Hygrobatopsis (Hygrobatopsella) inflatus (K. Viets, 1925), a species know thus far as deutonymph only, the females of Tubophora limnesioides Walter, 1935, Hygrobates niloticus Walter, 1922, Arrenurus (Megaluracarus) geniculatus Koenike 1898 and Momonides ghanaensis Smit, 2012, and the male of Piona damasiella Cook, 1966 are described for the first time. Many new records are presented for Ghana and the Gambia.


Introduction
Research on water mites from Western Africa started at the beginning of the 20th century with a series of publications of Karl Viets (1911Viets ( , 1912Viets ( , 1913Viets ( , 1916Viets ( , 1925. All the papers of Viets were dealing with water mites from Cameroon. Cook (1966) published one of the most comprehensive studies with the water mites from Liberia. A review on all African papers on water was given by K. Viets (1953) and K.O. Viets (1970). Since then, relative few papers have been published on the water mites from West Afrika (see Table 1). A number of publications are dealing with a wider range in Africa but include material from West Africa (Table 2). This paper deals with water mites from Ghana and the Gambia not treated in any of the papers from Table 1 or 2. Not included are most specimens of the genera Eylais Latreille (current state of knowledge of the genus insufficient) and Hydrodroma, Atractides and Torrenticolidae (these will be published in separate papers). Despite a long history in the study of water mites from West Africa, our knowledge is still poor. The aim of this study is, therefore, to increase the knowledge of water mites from West Africa.

Material and methods
The material described in this paper has been collected in 1998 (the Gambia), 2011 and2013 (Ghana). All material from this study is collected by the author, unless stated otherwise.
Holotypes, paratypes and all non-type material will be lodged in Naturalis Biodiversity Center, Leiden (RMNH). The following abbreviations are used: a.s.l. -altitude above sea level; Cx-I-IV -first to fourth coxae; Cxgl-1-42 -coxoglandularia 1-4; Dgl-1-4 -dorsoglandularia 1-4; IV-leg-6 -sixth segment of fourth leg; L -length; NHRS -Swedish Museum of Natural History, Stockholm; NP -National Park; suture line Cx-III/IV -suture line between third and fourth coxae; P1-5 -palp segments 1-5; SMF -Senckenberg Museum, Frankfurt a. Main; Vgl-1-3 -ventroglandularia 1-3. All measurements are in μm, measurements of palp and leg segments are of the dorsal margins, measurements of paratypes are given in parentheses. Ventral length is measured from the tip of Cx-I till posterior idiosoma margin. Numbers are given as male/female/deutonymph or adult/deutonymph. All coordinates are taken with a GPS. Coordinates given as degrees, minutes and seconds are taken from Google Earth and are by approximation. Data on the world distribution are taken from Smit (2020). Table 1 References after 1986 dealing with West African water mites.

Reference Country(ies) in W Africa Taxa
Diplodontus schaubi (Koenike, 1893)  Mamersa expansa Cook, 1979 -nov. stat. Remarks. Cook (1979) described this taxon as a subspecies of M. testudinata Koenike, 1898. Now the nominate taxon and its subspecies have been found on the same locality, they cannot belong to the same species. Moreover, the differences between the two taxa are quite large. Cook (1979) mentioned already the separation of Cxgl-1 and -2 by the genital plates. Mamersa expansa is much longer than M. testudinata and the genital plates are much longer (see Table 1). Consequently, the number of acetabula is much larger in M. expansa.
For differences between males and females see under M. testudinata. Distribution. Uganda (Cook 1979), Ghana (this study). Remarks. Females have a wider pre-genital sclerite than males, in the latter the pre-genital sclerite is more triangular. Moreover, the genital plates have a short anteromedial extension in the females, which is absent in the males.
Distribution. Widespread in Africa south of the Sahara.

Family Lebertiidae Thor
Genus Lebertia Neuman, 1880 A genus with numerous species known, but from the Afrotropical region only two species have been described, one from South Africa and one from Liberia (Cook 1966).
Remarks. Cook (1966) illustrated the female only, but the male is more characteristic with the long setae on the posterior part of the genital flaps ( Figure 1). This species has many swimming setae, also on II-leg-5 (one). III-leg-5 has up to seven swimming setae. The integument is lineated.

Genus Oxus Kramer, 1877
Oxus (Oxus) stuhlmanni (Koenike, 1895)  Remarks. In all but one aspects similar to the description given by Cook (1966). The posterodorsal platelet of the males of this study has a straight anterior margin instead of a Distribution. Liberia (Cook 1966 Remarks. The male matches the description of Walter (1946) and Lundblad (1949), but almost all female specimens have the Cx-I separated (Figure 2), unlike the descriptions of the latter authors. In a few females of this study Cx-I are near-fused. Lundblad (1949) described the subspecies L. monodi processifera with Cx-I separated, but this subspecies has smaller acetabula. Other subspecies have Cx-I fused, or the female is unknown. For the time being I assigned the specimens from this study to the nominate taxon.
Males of L. maglioi and L. monodi are easy to separate in the large (monodi) or small (maglioi) ventral extension of P4, but females of L. monodi with Cx-I separated are very similar to L. maglioi. The only differences are the more slender palp of maglioi, and possibly Cxgl-4 are closer to the associated seta in maglioi.
Distribution. Widespread in Africa south of the Sahara. Diagnosis. P2 ventrally with a setal tubercle, seta longer than tubercle; Cx-I fused medially, Cxgl-4 close to accompanying seta.
Etymology. Named for its occurrence in a wetland.

Remarks.
The new species is close to L. lacustris K.O. Viets, 1973 from Kivu Lake, Congo. It differs in the following characters (in parentheses L. lacustris): Cxgl-4 close to accompanying seta (widely separated), genital field of male rounded (with an indistinct to distinct posterior extension), seta of P2 longer than setal tubercle (about as long as setal tubercle). Moreover, the palp of the new species is much longer, e.g., P4 of the male measures 160 in the new species Distribution. Mediterranean countries, Mauretania (Walter 1946), Kenya (Lundblad 1942), South Africa (K.O. Viets 1964), Togo (Bader 1981), Ghana (this study).

Tubophora limnesioides
Remarks. Thus far, known only from the male holotype, and the female is described here for the first time.

Family Hygrobatidae Koch
Genus Atractides Koch, 1837 Atractides (Atractides) vietsi (Cook, 1966) ( Figure  Material for comparison. Atractides rectipes (K. Viets, 1924), holotype male, Múke-Fluß, Kamerun, 15 Jan. 1916 (slide 45938, SMF). Atractides vietsi (Cook, 1966) Remarks. According to Cook (1966), the differences between males of A. rectipes and A. vietsi are the single dorsal platelet in the former and the paired dorsal platelets in the latter. Unfortunately, the female of A. rectipes is unknown. The female of vietsi has a pair of dorsal platelets, as in the male ( Figure 5A). A character not discussed by Cook (1966), is the shape of the two distal setae of I-leg-5. These are slightly curved in both species, but rounded at its tip in rectipes, and truncated in vietsi (and not pointed as illustrated by Cook 1966) ( Figure 5E). Given these differences, the female of this study must be assigned to A. vietsi.
Distribution. Liberia (Cook 1966 (Lundblad, 1927). However, Lundblad (1952) reported H. coriaceus from the same locality as the nominate taxon, and therefore, they cannot belong to the same species. Also, in the Supon stream I collected both taxa. They are easily separated by the shape of the projection of P2: relatively short in coriaceus, and long in spathuliferus. Males of coriaceus have a ventral shield, but this can be absent in young specimens. More characters to separate the two species can be found in K.O. Viets (1963).
Female: Unknown. Etymology. Named for the presence of relatively large denticles on P3.
Remarks. The new species is similar to H. niloticus Walter, 1922 in the presence of an anteroventral extension of P2, with several denticles on this extension. However, H. niloticus has strongly tapering anterior coxae. Males of Hygrobates sudafricanus K.O. Viets, 1963 from South Africa, have P3 with 5-6 smaller denticles, P2 has a less extending anteroventral extension with larger and fewer denticles (about 2-3), in the new species there are about 6-7 small denticles on the anteroventral extension. Females of H. sudafricanus have P2 with a larger anteroventral extension and P3 with larger denticles, but the female of the new species is unknown.
Male: Unknown. Etymlogy. Named for its similarity with H. niloticus. Remarks. Hygrobates pseudoniloticus n. sp. is similar to H. niloticus in long and tapering anterior coxae. Unfortunately, H. niloticus is insufficiently described. The figures provided by Walter (1922) are sketchy and lack details, and only one male was known to Walter. The holotype is apparently lost. It is, therefore, difficult to decide which of the two species with long and tapering anterior coxae of this study belongs to H. niloticus. The latter species is quite large (the male is 700 long), and the anterior coxae are extremely tapering. Therefore, I assigned the largest of the two species, with the longest and most tapering anterior coxae to H. niloticus.

Remarks.
The new species is close to Hygrobatopsis levipalpis. The female of this species is not well-described. In the original description (K. Viets 1924), only the male has been described. Lundblad (1927) reported the female of this species, but in his material no males were present. I examined his specimens with the help of some photographs. P3 of Lundblad's specimens have a more or less straight ventral margin (like the male), P4 has a more distinct ventral bulge ( Figure 10A-B), Dgl-1 are on small sclerites and the gonopore is relatively longer. Moreover, P3 and P4 of Lundblad's specimens measure 93 and 129, respectively, much shorter than the specimens from this study.
Remarks. K. Viets (1925) described H. inflata based on the deutonymph only. Thus far, the adults of this species were unknown. The palp of the deutonymph of H. inflata as described by Viets agrees well with the palp of the adults and deutonymph collected in Ghana, especially the bulging ventral margin of P4 in combination with P2 with a ventral projection is characteristic.
The male of the new species is remarkably similar to the South African H. bella Cook, 2005, given the distance between the two countries. The anterior dorsal plate of H. bella is narrower and more elongated, and the posterior dorsal plate of bella is more or less as long as  I-leg-4-6: 79, 90, 68. I-leg-5 anteriorly with a few fine setae, one of these elongated (but lost due to mounting), no downturned seta present ( Figure 13D). Length of IV-leg-4-6: 88, 112, 88. Fourth legs much heavier than other three legs ( Figure 13E). Swimming setae: III-leg-5 and IV-leg-5 with two swimming setae, II-leg-5 with a short swimming seta.
Etymology. Named for the presence of few glandularia on the dorsal shield of the male. Remarks. Thus far, two species were known of the nominate subgenus. The male of Hygrobatopsis levipalpis K. Viets, 1925 (Cameroon) has a more shallow genital bay, Vgl-3 is not fused with the ventral shield and P4 is more bulging ventrally. The male of Hygrobatopsis sudafricana Cook, 2005 (South Africa) has the dorsal shield with three pairs of glandularia and the genital bay is shallow too. Both these species lack swimming setae. The presence of swimming setae will separate the female from other species of the genus.

Genus Tetrabates Thor, 1922
A genus with three or four species known from the Afrotropical region.

Piona damasiella
Remarks. The male of this species has not been described previously. Compared to P. damasi Lundblad, 1949, the genital plates are longer and extend more laterally. As in the female, the ventral margin of P2 is slightly convex, but straight in P. damasi.
Piona seyrigi continentalis Lundblad, 1952 New records.  Cook (1966) the only reliable character to separate the nominate form and continentalis is the number of acetabula, 30-38 in the seyrigi seyrigi and 16-26 in seyrigi continentalis. This study shows there can be an overlap in the number of acetabula, and that this character is not always reliable.
Distribution. Uganda (Lundblad 1952), Liberia (Cook 1966), Ghana (this study). Remarks. The assignment of this species is uncertain. Cook (1966) tentatively assigned specimens from Liberia to this species, which was described by Walter (1937) based on the female only. The specimens from Ghana match the description of Cook (1966) well.

Family Unionicolidae Oudemans
Distribution. Angola (Walter 1937), Liberia (Cook 1966 (Koenike, 1893), a species described initially based on the female sex only. However, Encentridophorus females show few species-specific characters, and therefore I agree with Bader (1981) that the assignment of Viets (1916) is questionable. Bader (1981) described his specimens from Cameroon as E. vietsi Bader, 1981, and postulated that the specimens of Viets might belong to this species. My male from the Gambia has the posterior idiosoma margin with 4+1 (in the holotype 5+1) pairs of stout setae, but is otherwise similar to E. vietsi. In my opinion the male from the Gambia and the specimens from Viets (1916) belong all to E. vietsi. The number of stout setae at the posterior idiosoma margin is, therefore, 4-6 pairs of setae + a separated seta on each side. Bader (1981) was the opinion that the specimens reported by Walter 1939) from Chad do not belong to E. vietsi, but belong possibly to a new species. Bader based this opinion on the measurements of leg segments and the number of swimming setae. I disagree with Bader, and I don′t see any marked differences between the specimens from Chad and Cameroon.

Subfamily Pionatacinae K. Viets
Genus Neumania Lebert, 1879 Neumania (Alloneumania) marginata (K. Viets, 1916) -nov. comb. Ecpolus dorsofenestratus Lundblad, 1949 Remarks. Cook (1966) suggested that Neumania marginata K. Viets, 1916 might belong to the subgenus Alloneumania. Males of Neumania dorsofenestrata (Lundblad) and females of N. marginata have been collected together in this study, and I consider them belonging to the same species (both with hook-like lateral extensions of coxae and extensive secondary sclerotization). Therefore, the suggestion of Cook (1966) was correct. I do not see any differences between the males from Ghana and Congo, and therefore I propose to synonymize N. dorsofenestrata with N. marginata. Young males do not have a dorsal shield. Walter (1935) reported this species from Niger and Burkina Faso. His illustration of the female shows a narrow strip of secondary sclerotization along Cx-IV. However, N. marginata has very broad and distinct secondary sclerotization along Cx-IV, almost completely filling the medial space between Cx-IV. In my opinion the specimens of Walter (1935) must be assigned to N. separata Cook, 1966. The female of this species has a genital field which is very similar to that of N. marginata, but lacks the broad secondary sclerotization. Walter & Bader (1952) reported N. marginata from Kenya, but they illustrated the deutonymph only. Therefore, it is not clear to which species their material belongs.
Remarks. The male specimen from the Gambia differs somewhat from the holotype and only known specimen thus far. The large claw of II-leg-6 is 86-94 long (121 in the holotype), and IV-leg-3 and IV-leg-4 have only one trifurcate ventral seta (versus IV-leg-3 with two or three and IV-leg-4 with three in the holotype). As nothing is known about the variation in these characters, I assign the specimen from West Africa for the time being to falcifera. The female has not been described previously.
Male: Unknown. Etymology. Named for the large spindle-shaped tubercles.
Remarks. The new species is somewhat similar to N. tuberculata Cook, 1966 andN. ecphyma Cook, 1996, but differs in the short apodemes of the anterior coxae, and the excretory pore is not on a large tubercle.
Etymology. Named after the country where the new species was collected. Remarks. The new species is somewhat similar to N. ecphyma Cook, 1966 from Liberia. The new species has smaller tubercles, and three of the dorsal tubercles are located medially (in ecphyma all tubercles are located near the lateral idiosoma margin). Moreover, the lateral extension of the genital field is absent in N. ecphyma.
Remarks. The male matches the description of Cook (1966), apart from the apodemes of the anterior coxae. According to Cook, they extend to the anterior part of Cx-IV, the specimens from Ghana have longer apodemes, extending to the middle of Cx-IV. Cook (1966) was uncertain about the assignment of the female, but in my opinion this is correct. Females have the venter with less secondary sclerotization compared to the male, and the apodemes of the anterior coxae extend to the middle of Cx-IV. Some measurements are given below. Males: Integument smooth. Idiosoma dorsally 437-454 long and 421-441 wide, ventrally 429-470 long. Females: Idiosoma dorsally 514-591 long and 454-531 wide, ventrally 535-591 long.
Neumania (Neumania) marginata K. Viets, 1916 New records. The Gambia.  Walter (1935) reported this species from Niger and Burkina Faso. His illustration of the female shows a narrow strip of secondary sclerotization along Cx-IV. However, N. marginata has very broad and distinct secondary sclerotization along Cx-IV, almost completely filling the medial space between Cx-IV. In my opinion the specimens of Walter (1935)  assigned to N. separata Cook, 1966. The female of this species has a genital field which is very similar to that of N. marginata, but lacks the broad secondary sclerotization. Walter & Bader (1952) reported N. marginata from Kenya, but they illustrated the deutonymph only. Therefore, it is not clear to which species their material belongs.
Remarks. Previously this taxon was considered a subspecies of N. tesselata. As already mentioned by Viets (1916), the two taxa can be found on the same locality, and therefore, they cannot belong to the same species. Also, in Ghana the two taxa co-occur. Moreover, there are distinct differences between the two taxa. The female of N. acuticaudata has not been described previously. The female of the N. tesselata has a larger central platelet with two pairs of glandularia and the two pairs of platelets posterolateral and posterior to the central platelet are absent. The male of N. tesselata has posterior to the central platelet a separate platelet with a truncated posterior extension, in N. acuticaudata there is a no such separate platelet, but there is a pointed extension posterior to the central platelet. Given the large differences between the two taxa, in males as well as in females, and the co-occurrence of the two taxa, I propose to consider N. acuticaudata a full species.
Unionicola heversi n. sp. Diagnosis. Cx-II and Cx-III with anterolaterally with a pointed extension; palp stocky, especially P4, this segment with two setal tubercles and anteromedially a stout pointed seta, P5 with large claws, ventrally with two claws but only visible in a skew position.
Description. Female: Idiosoma dorsally 745 long and 656 wide, ventrally 689 long. Anterior coxae with short apodemes, extending to Cx-III. Cx-II anterolaterally with a pointed triangular extension, Cx-III anterolaterally with a broader, pointed extension ( Figure 22A). Posterior margin of Cx-IV with a short projection. Genital field with five pairs of acetabula on two pairs of platelets, all platelets with long posterior setae. Anterior platelets with two acetabula each, posterior pair with three acetabula each. Unfortunately, the genital field is distorted in the slide. Palp stocky, especially P4, the latter segment with two setal tubercles and a short, stout pointed seta anteromedially. P5 with large claws, ventrally with two claws but only visible in a skew position as these are lying behind each other ( Figure 22B-C). Length of I-leg-4-6: 203, 227, 158. Long setae of first leg not fluted or grooved ( Figure 22D). Length of IV-leg-4-6: 251, 316, 284. IV-leg-5 with three, IV-leg-4 with two swimming setae.
Male: Unknown. Etymology. Named after Jürgen Hevers for his excellent work on the genus Unionicola.
Remarks. The palp with its large claws is somewhat similar to palps of U. dentifera Cook, 1966 andU. cooki Bader, 1981, but these species have the coxae with very large anterolateral extensions.
Distribution. Cameroon (K. Viets 1913, Burkina Faso (Walter 1935), Angola (Walter 1937), Liberia (Cook 1966), Ivory Coast (Hevers 2020), Ghana (this study). Outside Africa reported from Israel (Por 1968). Distribution. Cameroon (K. Viets 1911Viets , 1912, Liberia (Cook 1966), Ivory Coast (Hevers 2010), Ghana (this study). Outside Africa reported from Israel (Por 1968 Smit, 2012. I separated the two species based on the presence or absence of a two-lobed dorsal shield, and in lateral view the petiole is straight or upturned. With new material available of A. chappuisi, the slender idiosoma, and the more slender and longer petiole (both characters in dorsal view), are better characters to separate these two species Distribution. Sudan, Congo (Bader 1968), Ghana (this study).
Etymology. Named after Ankasa National Park, the name is a noun in apposition.
Remarks. The male of the new species differs from A. chutteri K.O. Viets, 1966 andA. chutteri longipes Smit, 2012 in a much more laterally expanding posterior part of the cauda. The female of A. chutteri has shorter genital plates, not extending to the lateral idiosoma margin. The female of A. chutteri longipes has undulating genital plates.
Remarks. As stated by Gerecke (2009), attribution of males and females in arrenurids can be problematic due to the strong sexual dimorphism. This applies to the specimen of this