Megadrile Earthworm Taxa Introduced to South African Soils (Oligochaeta: Acanthodrilidae, Eudrilidae, Glossoscolecidae, Lumbricidae, Megascolecidae, Ocnerodrilidae)

ABSTRACT The occurrence of 50 earthworm species of 22 genera from six families, namely Acanthodrilidae (Acanthodrilinae, Benhamiinae), Eudrilidae (Eudrilinae, Pareudrilinae), Glossoscolecidae, Lumbricidae, Megascolecidae, and Ocnerodrilidae, not native to South African soils, is reported. Some of the species of Pareudrilinae may be listed temporarily, possibly being indigenous to South Africa. Brief information on family status, species origin and broader distribution is included. Various laboratory and field experiments conducted in South Africa on some species are selectively indicated.


INTRODUCTION
The southern part of the African continent may be ranked as extremely biologically diverse terrain. Soil diversity caused by variable factors in geology, climate, landscape, etc interest. However, our knowledge of the terrestrial earthworm fauna in South Africa is indigenous families: Acanthodrilidae (5 genera with 146 species in Acanthodrilinae) and Microchaetidae (6 genera, 136 species), with numerous new species expected to be found. Data on the exotic taxa occurring in South African soils are usually restricted to inadequate notes. During the last century only three records of exotics were published. Michaelsen (1913b), identifying indigenous earthworms from Zululand, noted seven foreign species in Natal and one in the Free State. Ljungström (1972) recorded 23 nominal taxa introduced to this country intentionally or coincidentally during past centuries. Van Bruggen (1978) repeated similar information. The present record of the exotic terrestrial megadrile earthworms occurring in the soils of South Africa is drawn from various publications, supplemented by records kept in the Natal Museum database and observation gathered during years of earthworm collection and taxonomic study. Fifty nominal species (including subspecies) are accredited to 22 genera of six families: Acanthodrilidae: Acanthodrilinae (1 genus, 2 species); Benhamiinae (1 genus, 6 species); Eudrilidae: Eudrilinae (1 genus, 1 species); Pareudrilinae (2 genera, 4 species); Glossoscolecidae (1 genus, 1 species); Lumbricidae (8 genera, 19 species); Megascolecidae (4 genera, 12 species); Ocnerodrilidae (4 genera, 5 species).
It is usually accepted that earthworms may spread naturally, although very slowly, being carried by water (rivers, streams, rainfall), possibly by birds, and certainly transported intentionally or accidentally by humans. The majority of the species not native to South African soils are known worldwide, being distributed mostly by humans, as was pointed by Ljungström (1972). Pareudrilinae species listed in this paper, being ori-ginally indigenous to West or Central Africa, might spread to South Africa naturally, or are possibly developed from ancestral taxa extending the family's distribution to the southern part of the African continent. Some ocnerodrilid species, with confusing taxonomic and distributional status, are recorded with uncertainty, but their status may formation on the species' frequent or rare occurrence, indicating possible uncontrolled spread after accidental or intentional introduction.

MATERIAL AND METHODS
The information on megadrile earthworm species occurring in the soils of South Africa was drawn from publications containing taxonomic and distributional information, in some cases enriched with new data from the Natal Museum database, and the author's observations. The material, however, was not studied, and the location of though possibly mentioned in the original papers, is not given here. Data on species distribution are occasionally enriched with some ecological observations. Data pubwith no preserved material, are given selectively. The whereabouts of material used indicated. Short information is provided on family status and the original description and distribution of the species. The systematic record of families, subfamilies, genera and species is presented in alphabetical order. The names of provinces are updated to those currently used, and are abbreviated.
Abbreviations  Csuzdi, 1996 Taxonomic status and geographical distribution of taxa currently accounted to this family are under continuous study, discussion andevaluation, with various generic 1995, 1996;Jamieson et al. 2002;James 2005;Blakemore 2005Blakemore , 2006bBlakemore , 2007Blakemore , 2008bPop et al. 2005; and many others). In this paper, a taxonomic endorsement is partly based on the family rank and its capacity proposed by Csuzdi (1995Csuzdi ( , 1996Csuzdi ( , 2000, with the inclusion of the Benhamiinae. Acanthodrilid taxa are known from the warmer parts of southern and eastern parts of North America and the southern and eastern regions of South America; on the African continent they occur in western, central and eastern regions, and in RSA; they are also known from New Zealand, New Caledonia and parts of Australia. In South Africa 146 indigenous species occur, probably preor Gondwanan and related to South American taxa (Pickford 1937;Plisko 2007, In RSA recorded by Michaelsen (1899aMichaelsen ( , b, 1908Michaelsen ( , 1913a, Pickford (1937), Sciacchitano (1960), Ljungström (1972), and Zicsi & Pajor (1992) from a range of localities: Soutpansberg, small kloof near Louis Trichardt [= Makhado] (LP); near Lesotho border (FS); Kranzkop (KZN); Port Elizabeth (EC); 'Kafferlandet' (which is probably Cape Peninsula), Hout Bay and Mossel Bay (WC); Little Namaland and Komaggas [Namaqualand] (NC). Ljungström (1972) mentioned a sample collected in Pietermaritzburg (KZN), housed in ZMUH. In the NMSAD recorded from garden soil in St Lucia (KZN). Pickford (1937) found this species in native forest (Soutpansberg) and in a number of other sites in various provinces, and predicted its broad distribution, suggesting common occurrence in South Africa. However, during the last decades only one specimen was added to the NMSA collection.
Dichogaster (Diplothecodrilus) (Rosa, 1896: 510) Described from Padang, Sumatra. Known from tropical and subtropical areas in West Africa, South America, India, Burma, Pakistan, and some Indonesian Islands. In RSA recorded by Zicsi (1998) from three distant sites, all in KZN: Durban, Stamford Hill; Pietermaritzburg, Scottsville; and Umfolozi Game Reserve. Occurs in small populations in litter and between plant roots.
Dichogaster (Diplothecodrilus) (Beddard, 1893: 683) Type locality undesignated, described from potting soil in Kew Gardens, UK. Origin probably in West Africa. Known from Africa, South and North America, India, Burma, Australia, and various oceanic islands. In RSA, Horn et al. (2007) found it in natural habitats in the northern part of the Drakensberg (LP). In KZN, Ljungström (1972) repeated after Michaelsen (1913b) from two sites in Pietermaritzburg (Scottsville and Chapel St.); Zicsi (1998)  Central Africa (Stephenson 1930;Sims 1987). The genera and species are mostly endemic, not spreading out of their habitats. A few species, however, intentionally or accidentally transported by man to various parts of the world have been able to adapt to new conditions. has a worldwide distribution. Primarily imported to Information that , transported from Durban to Kew Botanic Gardens (apparently together with ), is endemic to KZN, as reported by Beddard (1893) and Michaelsen (1913b), has been queried for many years (Michaelsen 1912(Michaelsen , 1914Blakemore 2002Blakemore , 2006a Reynolds & Cook 1976: 65). Michaelsen (1913b) repeated after Beddard that the species is endemic in Durban, although in his following paper (Michaelsen 1914) expressed uncertainty on its endemicity, supposing its earlier introduction. The species redescription by Zicsi (1997a) on the material collected in Margate (KZN) or cultivated land, it is supposedly also introduced.
Genus Nemertodrilus Michaelsen, 1890Michaelsen, 1912 Described from Bloemfontein, FS, RSA. Reported from the type locality and Kummernais in Namibia. Known also from the banks of the Orange River (FS). Stephenson (1930: 870) stated this species to be a "peregrine". Michaelsen (1914) supposed a limited distribution, suggesting natural transportation to RSA by rivers. However, it is not know if this species was introduced from other part of Africa or, being indigenous to central western-or eastern Africa, its occurrence in southern Africa is habitual. To date, 100 years after the species was described, and after its second reporting by Michaelsen (1914), it has not been collected in any other part of RSA. Two other species of this genus, recorded below, were found distantly to the type locality, in the north-eastern part of RSA. It is also not clear if these two species are indigenous or are introduced from neighbouring Mozambique, where congeners occur (Michaelsen 1890b). Zicsi & Reinecke, 1992: 151 Described from Marumbeni, KNP, LP, RSA. Known only from the type locality and a few nearby sites, all in KNP: Marumbeni, Eileen, Sndwini, Mafagalamba and between Masitonto and Incomati rivers running in the park. It is not clear if this species is indigenous to RSA, or introduced with planted bushes or trees from neighbouring areas (compare notes for ). Zicsi & Reinecke, 1992: 153 Described from Shabarumbe, KNP, LP, RSA. Known from the type locality and other 19 localities in the central part of KNP, between the Letaba and Olifants rivers is the second of this genus found in north-eastern RSA. Both occur not far apart, and to date have not been collected elsewhere in RSA or in Africa. However, another species of this genus, griseus, is known from the banks of the Zambesi River in Mozambique (Michaelsen 1890b). It is possible that kruegeri and are indigenous to South African soil, extending the range of the eudrilids' occurrence to the south. However, their introduction and spread in the adapted areas are not excluded.

Subfamily Eudrilinae Claus, 1880
Genus Eudrilus Perrier, 1871 (Kinberg, 1867: 98) Described from St Helena Island (South Atlantic). One of the commonest species in West African soils, which are probably its original home. With its high potential for adaptation to new environmental conditions, it is intentionally transported and commercially used in composting, protein production, and agriculture. Viljoen and Reinecke imported this species to RSA from Germany (the stock originated in West Africa) for a variety of experimental studies (Reinecke & Viljoen 1988Reinecke et al. 1992;; and many others). After laboratory experimental use, this species is presently used in vermicomposting production in RSA farms. Two samples from Potchefstroom and Durban compost heaps are in the NMSA collection.
Family Glossoscolecidae Michaelsen, 1900 Indigenous South African Microchaetidae and South American Glossoscolecidae, having a known Gondwanan origin (Mischis 2004;Plisko 2009), are commonly accepted as closely related (Michaelsen 1900;Stephenson 1930;Brinkhurst & Jamieson 1971). To date, 136 microchaetid species recorded in South Africa are indigenous with high endemicity and very low ability to adapt to other than original habitats. Similar observations apply to the majority of the glossoscolecids occurring in South American biotopes. However, one glossoscolecid species, has been reported from numerous countries, being transported purposely or passively by humans, and adapting easily to variable environments where climatic conditions allow.

Genus
Schmarda, 1861 ( ) corethrurus (Müller, 1857: 113) Type locality possibly in Amazonian Brazil. Considered a cosmopolitan species. Showing high potential for crop improvement (Fragoso et al. 1999) is commonly used in agriculture practices. In RSA it is recorded from numerous sites in subtropical northeastern areas of the country, with a selection of coastal zone and hinterland sites along the Indian Ocean (Plisko 2001). Known from protected areas of KNP (Reynolds & Reinecke 1976;Zicsi & Reinecke 1992), and other sites in LP, MP, KZN, EC and WC (Michaelsen 1913b;Ljungström 1972). In inland areas (GP, NW), it has been noted only under or in plant pots, or in greenhouses. The records in the NMSAD increase the number of sites in natural habitats in protected areas: Bluff Nat. Res.; Eastern Shores (Greater St Lucia Wetland Park); Kosi Bay at the edge of Lake Zilonde; Krantzkloof Nat. Res.; Langepa Natural Heritage Site; Mbumbazi Nat. Res.; Ngoye Forest Reserve; Oribi Gorge Nat. Res.; and Skyline Nat. Res. (KZN), and in the northern Drakensberg at New Agatha and Soutpansberg Ratombo (LP). Found with microchaetids: tuberosa (Plisko, 1998) at the border of sugar cane plantation (Plisko 1998); in Dlinza Forest Nat. Res. with (Plisko, 1997), (Plisko, 1992); in the area of Port Shepstone, in polluted sites with (Michaelsen, 1913) and G. tuberosa (Plisko, 1998) (Plisko 2003. Often occurs with megascolecid sp. However, it must be said that the material accredited to this species in various papers and in material housed in the NMSA, may include more than only one glossoscolecid taxon. Moreno (2004) has recognized more species in various collections earlier accredited to corethrurus, applying more characters for the species identi-P. corethrurus is reported outside the Neotropics, a new revision of the P. corethrurus collection gathered in the NMSA and recorded in its database may disclose more glossoscolecid species.
The recognition of Lumbricus terrestris by worm taxonomy; however, worldwide dispersal of lumbcicids accredited mainly to human activities for many years confused the Palaearctic origin of the Lumbricidae. In later years many of the already described species, being collected in distant parts of the world, were described as synonyms. In addition, numerous lumbricid taxa reproducing parthenogenetically in variable morphs, suggested new taxonomic accreditation and were described under various synonyms. The recently published checklist of internationally recognized lumbricid species includes nearly 1000 nominal names, of which 700 species names in 39 genera, are accepted as valid, although many of them are still awaiting a new evaluation (Blakemore 2004a).
In RSA 18 lumbricid species occur, attributed to eight genera: Allolobophoridella (2); Aporrectodea (4); (3); (2); (2); (1); Lumbricus (2); and (2). Most of the species are synantropic, known from many parts of the world. To this country they were introduced at various times in many ways (Ljungström 1972). These species, reproducing in variable parthenogenetic or polyploid morphs, were recorded under variable synonyms, which now makes correct misleading species distributions, as happened with Kinberg, 1867, described apparently from Durban. Michaelsen (1899b) after revision of the original L. terrestris L., however, data referring to the type locality as 'Pt. Natal' (what may be understood as Durban in KZN) was uncertain. Michaelsen (1899b: 421) expressed his doubt in the words "… , and this declaration allows us to take L.
= L. terrestris out of the record of the species occurring in RSA. Ljungström (1972) expressed a similar opinion. Thus the report on the occurrence of L. terrestris in South Africa given by Gates (1972: 119) is not correct. No L. terrestris were collected in South Africa during consecutive earthworm collections undertaken in the past decades by a number of researchers (Michaelsen, Pickford, Ljungström, Reinecke, Reynolds, Pajor, Zicsi, Plisko, members of Ezemvelo Wildlife KZN, members of the University of KwaZulu-Natal (Pietermaritzburg), and others). Experimental introduction of mature individuals of this species into gardens and grasslands near Durban and Pietermaritzburg did not bring the expected results. It seems that L. terrestris does not reproduce under South African ecological and climatic conditions. However, once in a while some information as terrestris (Mallett et al. 1987) probably belong to the Megascolecidae, giving at the cimens have similar dimensions and a violet colouring, both features characteristic of some megascolecids occurring in this country, and also known from the literature for the Lumbricus species.
Genus Allolobophoridella Two species and accredited to this genus are morphologically much alike, undoubtedly differing in the histology of their muscles. Unfortunately, no histology was conducted on the material collected in under various synonyms.
Although its occurrence is noted as less common than trapezoides, its wider distri- et al. 2007, 2008).
Aporrectodea trapezoides (Dugès, 1828: 289) (s. str.) Type locality is Montpellier, France. Described worldwide under numerous synonyms. Its occurrence is usually similar to s. str. Both occur habitually together, although in various ratios. In RSA, reported from numerous sites. For years , and placed under synonymy. Reported by Michaelsen (1913a, b) from Knysna rain forest (EC) and Umgeni Falls (KZN); by Ljungström (1972) from Pietersburg (LP), Ermelo (MP), Potchefstroom (NW), Heidelberg and Roodepoort (GP), Groot Brak River Staasie and Uitenhage Kerkstraat (EC) and Cape Town (WC). Jamieson (1967) reported it from the Krom River, a tributary of the Erste River and north of Stellenbosch. Reinecke and Ljungström (1969) collected it in the area of Potchefstroom with (Michaelsen, 1899). Visser and Reinecke (1977) record it from Moi River irrigation area near Po- Dlamini et al. (2001). Collected with the acanthodrilid Parachilota timothyi Plisko, 2008, other lumbricids: Aporrectodea rosea and , and various megascolecids. The NMSAD records the common occurrence of this subspecies at over , was placed by Michaelsen (1899b) in the genus Allolobophora a) it was indicated as a synonym of trapezoides. (Ude, 1885: 136) Described from garden soil in Göttingen, Germany. Noted from many parts of the world. In RSA reported twice, from garden soil in Roodeport (GP) and Heidelberg (WC). Collected together with and by Reinecke and Ryke (1969).

Genus
Eisen, 1873 (Michaelsen, 1903: 130) Described from Sassari, Northwest Sardinia. Known from a few European countries also from USA and Chile. In RSA, recorded in NMSAD: two clitellate specimens found in litter of the indigenous part of the Newlands Forest partly covered by pine plantation, at the Cape Peninsula (WC). Collected with and Lumbricus rubellus, near , and indigenous juvenile acanthodrilids. (Michaelsen, 1890a: 15) Described from Hamburg, Germany. Known from many parts of the world, although not commonly. Transported with various plants; recorded in manure, composting, and rich organic soil, also in greenhouses. Reproducing in various morphs. Polymorphism synonym var. collected from garden soil in Port Elizabeth (Michaelsen 1899a) (EC), and from Cape Town (Michaelsen 1913a) (WC). (Savigny, 1826: 183) Described from Paris, France. Occurring widely over the world under variable climatic conditions, in forested areas and plantations, also in golf course greens and grasslands. Broadly known from high altitudes, and cold climatic areas. In RSA recorded mostly from Drakensberg sites at variable altitudes, often reaching 1700 m; common in forested Drakensberg area in LP; 29 records in NMSAD refer to protected areas: Karkloof Nat. Res., Qudeni State Forest, Qudeni Forest Reserve, Mount Currie Nat. Res., Monks Cowl, Kamberg, Cavern, Ndema Grotto (KZN); occurs commonly in indigenous forests, plantations, and grasslands (EC, WC), possibly introduced with planted foreign trees, as suggested by the owner of Benvie Farm, where it was found with the lumbricids , and rubidus rubidus, and the megascolecid sp. (Rosa, 1886: 674) Widely distributed, with high morphological variability. During recent decades used for vermicomposting practices. Introduced to RSA for experimental study, with a possible application in compost production (Viljoen et al. 1991;Muyima et al. 1994). It is not known if these experiments have been completed, or if the species has spread into neighbouring soils in RSA.

Genus
Omodeo 1956 (Savigny, 1826: 182) (s. l.) includes D. r. rubidus (Savigny, 1826) and (Eisen, 1873). Both subspecies reproduce parthenogenetically in variable morphs and have been described under numerous synonymic names. These parthenogenetic and polyploidy forms are often separated into subspecies, but rubidus" s. l. Following the observations of Csuzdi & Zicsi (2003), and Blakemore (2004a), two subspecies are recognised in this paper. (Savigny, 1826: 182) Described from Paris, France. Distributed widely all over the world. In RSA reported from numbering localities. Large populations were observed by Horn et al. (2007) in Limpopo forests, and in the Drakensberg North mountain range at altitudes over KZN protected areas: Pongola Bush Nat. Res., Itala Game Reserve, Ngele Forest, Mapelane Nat. Res., Karkloof Nat. Res., Karkloof Benvie farm, Newcastle Incandu Forest, Vryheid Hill Nat. Res., the central part of the Drakensberg at Cathedral Peak, Rainbow Gorge Forest and Doreen Falls, Oribi Gorge, Good Hope Farm (Mondi), Impendle Nat. Res., Gxalingenwa Nat. Res., and Swartkop Forest. Common in arable soils in KZN, FS, GP, EC and WC. Occurs abundantly in natural and cultivated soil, from sea level to high altitudes of the Drakensberg mountains. Under favourable SA climate this species is active throughout year, appearing in large populations. Being such a great consumer of organic matter (Csuzdi & Zicsi 2003) it may compete for food with native invertebrates, and possibly should be controlled in RSA.
Genus Malm, 1877(Bouché, 1972: 381) Selected from batches of buted all over the world, usually occurring with fetida in vermicomposting heaps, composts, and decaying organic matter. Being externally and anatomically similar to fetida, and usually occurring together with it, is often not recognized and separated, but is recorded as fetida. It is possible that in composting hips and vermiculture farms in the RSA this species also occurs with fetida. A sample imported from France containing fetida and logy and reproductive abilities by Reinecke et al. (1991). The material housed and recorded in the NMSAD under E. fetida may include . DNA study is necessary to separate these two species, and this has not yet been applied to the material in the NMSA collection. Easton (1983) placed it in synonymy with E. fetida. Blakemore (2004a) supported (Savigny, 1826: 182) Described from Paris, France. Widely distributed over the world for variable reasons. In RSA known from composting hips, vermicomposting farms, nurseries, around barns and from rich decaying organic matter. Reported by Ljungström (1972) from Pietersburg (LP); Heidelberg and Roodeport (GP); and Cape Town, Cape Flats and Bergvliet (WC). Zicsi (1998) reported its occurrence in Durban, Stamford Hill (KZN). In the NMSAD recorded at 27 sites in GP, NW, KZN, EC and WC. Synantropic. Found once with . This species reproduces abundantly under favourable conditions, being the most miculture and protein production, and as an indicator in polluted habitats. Undergoing worldwide variable experimental study, fetida also aroused interest in this country. Various experiments were and are being conducted by vermiculture producers at universities and agriculture institutions, and the members of the Earthworm Interest Group in South Africa, with an extended literature on the use of this taxon. From the number of published notes, only a few are cited: Venter & Reinecke (1987); Reinecke et al. (1990Reinecke et al. ( , 1991; Reinecke et al. (1992); Reinecke & Alberts (1994); Maleri et al. (2008).
Genus (Michaelsen, 1900) (Savigny, 1826: 184) Described from Paris, France. Amphibiosus/limicolous, with a high preference for damp habitat; widespread worldwide. In RSA recorded from banks of natural inland watercourses, man-made lakes and damp, decomposing organic matter. Found at Pretoria Zoological Garden (Michaelsen 1913c) and the Jukskey River near Bryanston (Jamieson 1967), both sites in GP. Michaelsen (1913a), Pickford (1937), Jamieson (1967) and Ljungström (1972) (Savigny, 1826: 181) Described from Paris, France. Distributed in many parts of the world. In RSA found only at one site in Eagle Ridge Forest Resort, near Stutterheim (EC). Collected from garden soil (6 clitellate, 10 juvenile) on the property of a German family who planted In Europe it is known from isolated sites and usually occurs in small populations (Plisko 1973;Csuzdi & Zicsi 2003). Gates (1972) repeated, after Karppinen and Nurminen (1964) that in Finland, this species was introduced to greenhouses and spread within a short time to neighbouring gardens. Recently Kuu and Ivask (2010) recorded rapid spread of large populations of this species in northern and western Estonia. (Örley, 1881: 584) Widely distributed over the world. In RSA reported under numerous synonyms, from a Cape Town garden (WC), a Potchefstroom plant pot (NW), Sabie (MP), and the Limpopo forests (LP). In NMSAD, recorded from over 200 sites in all RSA provinces, from natural and cultivated biotopes. Reproduces in variable parthenogenetic and polyploid forms. Under favourable climatic conditions occurs abundantly in moist soils, dispersing widely in natural and cultivated land. Occasionally found together with the introduced lumbricids D. rubidus, D. octaedra and , and with the indigenous microchaetids Plisko, 1997 andPlisko, 2002, and with the acanthodrilid Parachilota timothyi Plisko, 2008.
Family Megascolecidae Rosa, 1891 Taxa accounted presently to Megascolecidae are continuously under study, discussion and evaluation, with various genera and families being accredited. Many species commonly distributed across the world have been frequently described as synonyms in various genera. The most common species were assembled in species groups assigned under the earliest erected species names (Sims & Easton 1972;Easton 1982). Blakemore (2004b), in a revision of pheretimoid earthworms, eliminated numerous synonyms and relocated some of the commonest species into different genera. In the present paper, for all cosmopolitan pheretimoid species recorded in the literature under various synonyms, comprehensive synonymies are not included. A large part of the material collected during last twenty years and housed in NMSA is recorded in the NMSAD under the broad name "Pheretima-group", waiting for accreditation to valid taxa. Species collected in RSA and recorded in the present paper are accommodated in the genera  (Blakemore 2004b). In RSA Dlamini (2002) reported it from the Eshowe area (KZN). Recorded in the NMSAD from greens of Pretoria Country Club and Hartbeespoort Dam (GP). Twenty-three records refer to KZN culand grassland at Cedara Agriculture College. Ten records refer to natural habitats in the Doreen Clark Nat. Res., in Queen Elizabeth Park at its border with the Victoria Country Club, and at Ngome Forest (all KZN). (Kinberg, 1867: 102) Described from Oahu, Hawaii (sub cortice arborum). Known from tropical and temperate regions, transported by man. In RSA known from numerous natural, often protected, areas and cultivated sites. Reported from number of localities in MP, GP, NW, FS, KZN, EC and WC by Zicsi & Pajor (1992), Zicsi & Reinecke (1992) and Zicsi (1998). Common in decomposed litter in indigenous forests, and plantations. Occurs in small populations. Found with other introduced megascolecids, and the microchaetid (Plisko, 1993). Blakemore (2004b), Sims (1987) and some other authors place in synonymy with corticis. In the present paper is accepted as a valid species (see below). (Baird, 1869: 40) Described from Plas Machynleth, North Wales. Broadly distributed by man all over the world. In RSA reported by Michaelsen (1913a, c), Ljungström (1972), Dlamini (2002) and Horn et al. (2007) from nearly 50 natural and cultivated habitats in LP, GP, NW, FS, KZN, EC, and WC. Occurs in variable morphs, in large populations. Reproduces parthenogenetically in morphs often recorded under synonyms. In the material collected during the last 20 years commonly observed variations may include: variably sized spermathecae and variably shaped diverticula; prostate glands single or paired, juvenile, rudimentary, usually variously reduced in size, or lacking. Male sterility was greenhouses and composting heaps (unpubl. notes). Gates (1972) also recorded male sterility and various reductions of prostatic glands in Burmese individuals. Sims and Easton (1972) designated this species name for a species-group, inserting a number of related species. Blakemore (2004b) synonymized it with corticis. However, the material accredited in this paper to this species differs from corticis by highly developed male sterility. (Kinberg, 1867:102) Described from Rio de Janeiro (' '). The indigenous range of this species is probably in South-East Asia; presently widely distributed and known from warm temperate zones, and numerous greenhouses, from all continents and many islands. In RSA recorded by Zicsi and Reinecke (1992) from Nelspruit and Sabie (MP), and Zicsi (1998) from Tsitsikamma National Park (WC), collected together with D. r. rubidus. (Rosa, 1891: 396) Described from Hawaii. The original species range is believed to be China (Gates 1972); currently distributed all over the world. In RSA broadly distributed in LP, MP GP, NW, KZN, and EC, occurring in moderate populations in natural biotopes and Dlamini (2002). In the NMSAD recorded from 71 sites. Sims and Easton (1972) designated the species name for a species-group assembling species with related features. Easton (1982) and others found this species to be a junior synonym of gracilis; however, it is considered valid in this paper. (Horst, 1893: 66) Described from Tijbodas, Java. Distributed worldwide in tropical and subtropical zones. In RSA occurs in numerous sites in indigenous forests with exotic plantations (LP). Occurs commonly in natural habitats in protected areas, and in grasslands, farms (KZN). Recorded in large populations by Ljungström (1972), Zicsi (1998) and Dlamini (2002) from numerous sites in KZN. Horn et al. (2007) recorded it from LP; Zicsi and Reinecke (1992) reported it from Sabie and Nelspruit (MP). Once reported from the Stutterheim area (Zicsi 1998). In the NMSAD recorded 27 times from various (Beddard, 1892: 166) Described probably from Pengang, (?) Malaysia. Known worldwide from tropical and collected by Zicsi and Reinecke (1992) from the riverbank in Potchefstroom (NW), and the Sabie area (MP). Recorded in the NMSAD from Pretoria city parks and greens (GP). (Grube, 1879: 554) Described from Rodriguez. Known from many parts of the world, from variable habitats; distributed by man. In RSA collected from over 100 sites in natural and cultivated  from Skukuza (LP), Sabie area, Nelspruit, Graskop (MP) and various sites in KZN. Zicsi (1998) and Dlamini (2002) collected it from a number of localities in KZN. Horn et al. (2007) reported it from Drakensberg North, New Agatha (LP).
Genus Metaphire Sims & Easton, 1972(Kinberg, 1867 Described from Sausolita Bay, San Francisco, California, USA. Known from many parts of the world from variable habitats. In RSA recorded only from experimental plots at Cedara Agriculture College and their close neighbourhood (KZN). Possibly this species has been introduced with seedlings of maize. (Perrier, 1872: 122) Described from 'Cape of Good Hope. Kapland' [RSA]. Redescribed by Horst (1883) as . Its occurrence in South Africa was noted by Beddard (1895b) and Michaelsen (1900Michaelsen ( , 1910bMichaelsen ( , 1913aMichaelsen ( , b, 1922, although its endemicity was questioned. and described under various synonyms. The numerous repetitions under various generic accreditations indicate the species' appearance in the Cape Peninsula. However, the species' derivation in this country may not be accepted, because the origin of pheretimoids does not extend to South Africa. Some of the revisions performed on original and synonymic material suggested an optional taxonomic evaluation. Michaelsen (1922) judged both the species and as valid, although indicating noted disaccords. Sims and Easton (1972) and Blakemore (2008c) place both species in the genus Metaphire.

Genus
Perrier , 1872, Perrier, 1872 The species is sometimes assigned to the acanthodrilids, but after a recent molecular et al. 2002). Described from the Himalayas (Gates 1972). Common in Asia; distributed by man around the world. Extensively used in experimental and applied practices in the breakdown of organic wastes, vermicomposting, protein production and soil toxicology, and as indicator of soil quality. A batch of specimens imported to RSA for experimental study (Hallatt et al. 1990(Hallatt et al. , 1992Reinecke et al. 1992;Reinecke & Alberts 1994;Reinecke & Pieters 1997, 1998 initiated the growth of small populations in compost heaps near Potchefstroom (NW). Also found at a vermicomposting farm in Camperdown (KZN).

Genus
Perrier, 1874 (Grube, 1855: 127) Described from Villa Franca on the French Riviera. Widely distributed along seasand-and mud-dwelling species. Euryhaline, easily dispersed by drifting on debris, or transported by various ways. Described in the megascolecoidea, but in recent years accredited by various authors to the acanthodrilids. Recently after revision it was accepted as megascolecoid, and was placed in the genus recommended by Beddard (1891) with Blakemore (2008a) suggesting its placement in the restored family Pontodrilidae RSA recorded by Michaelsen (1913b) from the coastal zone at Kosi Bay in Zululand (KZN) under the synonym P.
Family Ocnerodrilidae Beddard, 1891 Genus Eukerria Michaelsen, 1935(Beddard, 1895a Originating probably from South America; spread with substantial human involvement widely over warm and temperate zones. Recorded throughout the southern hemisphere from various regions. Noted also from Burma, India and Australia. In RSA known from numerous natural littoral habitats, man-made water catchments, and variable moist biotopes elsewhere, where collections have been undertaken. Michaelsen (1913c, b) reported it from Pretoria Zoological Gardens (GP) and Howick, near Umgeni Falls (KZN). Jamieson (1967) collected it from numerous sites along rivers in FS, NW, MP, KZN, EC and WC. Ljungström (1972) found it in variable biotopes, mainly on river banks and in irrigated areas, in NW, FS, EC and KZN. Visser and Reinecke (1977) reported it from the Mooi River area, near Potchefstroom (NW). Zicsi (1998) recorded it from a number of sites in MP, NW, KZN and EC. This species may be aquatic or littoral, seldom terrestrial, usually occurring between roots of aquatic vegetation. Twenty-two records in the NMSAD refer to samples collected in littoral biotopes.

Genus
Eisen, 1900 (Beddard, 1893: 259) Type locality uncertain. Distribution and species adaptation observed in many parts of the world. In RSA reported only by Zicsi (1998) from a few sites in the Durban and Pietermaritzburg areas (KZN).

Genus
Eisen, 1878 Two ocnerodrilid species, and , produce variable morphs and belong to one taxon. In this paper, they are given separately. Beddard, 1893: 703 Beddard, 1893 and Beddard, 1893 have been described from Kew Gardens (UK) on material apparently received from Durban, KZN, in a box containing various plants. Although both types were initially deposited in the British Museum, the present housing of the type material is unknown, leaving its origin in uncertainty (Blakemore 2006a;Reynolds & Cook 1976: 65). Michaelsen's (1913b) statement that this species is "endemic for Natal" cannot be accepted as this species, before being transported to UK, was probably introduced to South Africa, and Michaelsen (1914) was also of this opinion. To date, this species' distribution is not certain, possibly being the southern part of Africa. Taxonomic evaluation of material known from RSA is not clear. Described in by Beddard (1893); by Michaelsen (1913b) acknowledged under ( ), accepting Beddard accreditation to ( ) when reporting it from garden soil in Chapel Street, Pietermaritzburg. Zicsi and Pajor (1992) record it under drilus for the material collected from garden soil near the Cathedral Peak hotel in Drakensberg. Zicsi (1998) placed it in when recording it from Ashburton (near Pietermaritzburg) grassland, at the edge of neighbouring garden. Reynolds and Cook (1976), confusing the generic names and Ilyodrilus, placed it in the Blakemore (2006a) may be a junior synonym of Eisen, 1878.
Eisen, 1878: 10 Typus amissus (Reynolds & Cook 1976: 146). Type locality unknown. Widely distributed in tropical and temperate zones by human activity (Righi 1984). Zicsi (1996Zicsi ( , 1997b reported it from the eastern part of Africa. This species, with great variability in the male genital system and with reduction of spermathecae creating variable morphs, has been accounted to a variety of species. A not clearly elucidated generic relationship between and tion, adding synonyms to taxa accounted to both genera. Blakemore (2006a) is of the opinion that is a junior synonym of . Csuzdi (pers. comm.) supposes that is a thecate form of , although has not been recorded from South America and possibly also not outside Africa.
In Zicsi (1998) found this species in the same province, in the Fouriesburg area on the bank of Vaal River near Parys (NW). Ljungström (1972), following information on his extended collecting of earthworms in this country. However, during many years of earthworm collection in South Africa by numerous members of Ezemvelo Wildlife KZN and by Plisko, this species was rarely recorded. All material reported by Zicsi from RSA is kept in the HNHM.
Genus Pygmaeodrilus Michaelsen, 1890Michaelsen, 1910a Described from Bothaville, FS, RSA. The type depository is unknown; however, it might be housed in ZMUH. Known from Namibia (Grootfontain) and a few sites in RSA, where recorded only from the type locality and few sites in the same province: at Vaal Dam catchments at Kalkspruit, Grootvlei coal mine, and the Sabie River, Lover Sabie Rest Camp in KNP (LP) (Jamieson 1967). One of the smallest megadriles, being 25 to 40 mm in length and 1.5-1.6 mm in diameter. Its indigenous status is not and to southern African continent is transported naturally by rivers.

DISCUSSION
Fifty introduced earthworm species were selected from published sources and drawn from the NMSA database. The majority of species are well known as broadly distributed all over the world. A few taxa described from South Africa at various times, sometimes with an indication of their endemic status, need reappraisal. Metaphire and are recorded with uncertainty on their taxonomic and distributional status. The eudrilins , N. kruegeri and , and the ocnerodrilid , might be native to RSA or indigenous to neighbouring territories, extending the geographical distribution of the Eudrilidae and the Ocnerodrilidae to the southern part of the African continent. Their inclusion in introduced taxa might be temporary.
The collection of in Margate (KZN) from garden soil, and its redescription by Zicsi (1997a reports from Tanzania, where it was also collected in garden soil or cultivated land, suggest its possible introduction from a not yet established origin. Most of the taxa not native to RSA soils belong to widely dispersed lumbricids, megascolecids and glossoscolecids, which have been introduced variably, mostly by human activity, although natural transportation might play a possible role. Some of them are now well adapted to South African soils and climatic conditions, and have developed into large communities. Two lumbricids, s. l. and A. rosea, have become common in various habitats, occurring in numerous sites in all taxa, which were partly or fully exterminated from these habitats by unfavourable agriculture or industrial practices. and play a large role in generating compost in gardens, and in commercial vermicomposting practices, with and possible adaptation to composting practices. It was observed that the lumbricids rubidus s. l. and lacteum, the megascolecids , , , , , the glossoscolecid , and two species of the Benhamiinae, chogaster (Diplothecodrilus) bolaui and , occur abundantly in natural habitats, often in protected areas. These species are great eaters of organic matter and might be competitive consumers of food needed for native invertebrates. These species also show a tendency to spread and to colonize neighbouring biotopes. Because no appropriate protection against deliberate or occasional earthworm introduction is yet applied in RSA, concern is felt about infestation of soils by further introduction of these, or other, competitive species. A unit, falling under government control, should be established to regulate the transportation of earthworm samples between countries, or nationally for vermicomposting, sugar cane cultivation, experimental study, or any other reason.