Euphorbia malevola subsp. bechuanica Leach (1964: 6) Euphorbia limpopoana L.C. Leach ex Carter (2000: 960) A review of the Euphorbia schinzii-complex (Euphorbiaceae) in southern Africa Bruyns, Peter V. Klak, Cornelia Hanáček, Pavel Phytotaxa 2020 2020-03-19 436 3 201 221 5H994 (Leach) Bruyns , Bruyns 2020 , Bruyns 2020 [180,803,554,581] Magnoliopsida Euphorbiaceae Euphorbia Plantae Malpighiales 13 214 Tracheophyta subSpecies schinzii comb. nov. bechuanica      Euphorbia malevolasubsp. bechuanica Leach (1964: 6). Type:— BOTSWANA. Halfway between Palapye and Francistown, July 1937, Obermeyer( holotypePRE!, isotypesK!, PRE!).     Euphorbia limpopoanaL.C. Leach ex Carter (2000: 960). Type:— ZIMBABWE. Fulton’s Drift, 25.5 kmNNW of Beitbridge, January 1963, Leach 11582a( holotypeSRGH).   Distribution & Habitat:— Subsp. bechuanicais widely distributed in Botswanaand Zimbabweand in South Africafrom west and north of the Soutpansberg, but is also recorded in the lowveld south-east of Tzaneen ( Fig. 6).   Discussion:—Plants belonging to subsp. bechuanicawere shown in White et al. (1941: figs 839−841) as a ‘variable form of  E. schinziiPax’ and discussed on page 749. They were first recognized taxonomically as  E. malevolasubsp. bechuanica Leach (1964: 6).  Carter (2000)characterized  E. limpopoanaby ‘lack of rhizomatous branches and its more robust habit, with thicker branches and noticeably much stronger spinescence’. However, even among the specimens she cited, there is wide variation in such features as size of the plant, thickness of branches and length of spines. This variability was indicated previously by White et al. (1941), by referring to it as a ‘variable form of  E. schinziiPax’. The spine-shields may be prominent and they sometimes continue to just above the axillary bud below, but more typically they continue to about halfway down between successive spine-complexes. Hargreaves (1994: 153)showed specimens of  E. limpopoana(from NE Botswana) with a tuberous roostock, while this is absent in plants from the eastern side of its distribution near Hoedspruit (e.g., Bruyns 12101(BOL)), in which the branches also have a paler streak down the grooves between the angles that is absent in Botswanan material. Flowering specimens have been collected between May and November ( Carter 2000), but the main flowering period is June−August.   Euphorbia limpopoanais not always easily separated from what was traditionally known as  E. schinzii. In collections such as Bruyns 12079, 12085, 12091, 12291(all at BOL) from the western part of the former Transvaal and near the border with Botswana, some plants have robust branches with strong spines and some of these robust branches are rhizomatous, while yet others have much less robust branches. For example, from Ramotswa on the Botswanan-South African border, Carter & Leach (2001)cited a specimen under  E. schinziiand Hargreaves (1994: 163)showed two excavated plants with tuberous rootstocks and rhizomatous branches no more than 4 cmlong above the ground, which he referred to as  E. schinzii. Other plants in the area have branches to 20 cmlong above the ground and 20 mmthick, with spines to 17 mmlong, with the same rhizomatous tendency (e.g., Bruyns 12079, 12291,as in Fig. 7). Robust specimens in these populations resemble  E. limpopoanaand only their rhizomatous tendency suggests what was traditionally called  E. schinzii. On the other hand, the daintier plants among them clearly belong to what was traditionally called  E. schinzii. As a consequence, rather than an abrupt break between  E. limpopoanaand what was traditionally called  E. schinzii, there appears to be a gradual change from typical  E. limpopoana( Fig. 7A) to populations where some plants are robust and separable only by their occasional rhizomatous branches from  E. limpopoana(e.g., Fig. 7 B, C) to populations (mainly much closer to Pretoria and at higher altitudes) where the plants are much less robust and are more frequently rhizomatous. Consequently,  E. limpopoanais now placed under  E. schinziias a subspecies, at which rank the epithet ‘ bechuanica’ has precedence.  Selected additional specimens examined:— BOTSWANA. Boboneng (2128CD), 680 m, Bruyns 12404(M). Serowe (2226BC), Leach & Bayliss 12514(NBG). Mosolotsane (2226CD), 1220 m, Bruyns 12306(M). 5 kmnorth of Baines Drift-Pont Drift road (2228BD), 620 m, Bruyns 12409(NBG). Ikongwe, 20 kmsouth of Shoshong (2324AB), 1070 m, Bruyns 12304(UPS). 13 kmnorth of Molepolole (2425BA), 1150 m, Bruyns 12295(BOL). 35 kmfrom Lobatsetowards Gaborone(2425CD), Hansen 3149(PRE). Kanye (2425CD), Hartley 1033(J). MOÇAMBIQUE. 5 kmfrom Massingir (2331DD),  Correia& Marques 967(WAG). Massingir (2331DD), Schäfer 7174(BR, K, WAG). 14 kmfrom Massangena towards Madindere,  Correia& Marques 3165(WAG). SOUTH AFRICA. 25.5 kmsouth of Messina (2229DB), Dyer 3971(BM, K). 64 kmnorth of Louis Trichardt (2229DB), Dyer 3973(BM, K). Halfway between Wyllie’s Poort and Waterpoort (2229DC), Galpin(BOL). Langdraai (2229DC), Buitendag 1079(NBG). Near Sand River, Soutpansberg (2229DC), 1700’, Gillett 3868(BOL). 20 milesnorth of Louis Trichardt (2229DD), Leach 11679(K, NBG). 32–35 kmnorth of Louis Trichardt (2229DD), Dyer 3972(BM, K). Grootgeluk (2229DD), Buitendag 1076(NBG). North of Masekwapoort (2229DD),  Vander Schijff 4439(W).