Polygala appendiculata Vellozo (1829: 292)

Pastore, José Floriano B., 2013, A review of Vellozo’s names for Polygalaceae in his Flora Fluminensis, Phytotaxa 108 (1), pp. 41-48 : 45

9. Polygala appendiculata Vellozo (1829: 292) . Lectotype (designated here): — BRAZIL. Rio de Janeiro,

" Habitat campis ad littora maris apricis. Offendi prope regium Molendinum farinam e mandihoca conficiendi a

militibus consumendum. ", 1783–1790. Iconograph: Vellozo (1831: tab. 66).

Epitype (designated here): — BRAZIL. Rio de Janeiro, Arraial do Cabo formerly part of the city of Cabo Frio , Between Araruama Lagoon and the main road to Cabo Frio , Arraial do Cabo , 16 July 1953, F . Segadas Vianna et al. 885 ( RB!) .

Polygala leptocaulis Torrey & Gray (1838: 130) . Lectotype (designated by Rankin-Rodriguez 2003): — UNITED STATES OF AMERICA. Without Locality, no date, T.Drummond 28 (NY!, iso-lectotypes FI!, P!).

Polygala paludosa Saint-Hilaire & Moquin-Tandon (1828: 45) var. longispicata Saint-Hilaire & Moquin-Tandon (1828: 45) [as "var. ε longè-spicata "]. Lectotype (designated here): — BRAZIL. Santa Catarina, Laguna, "prope pagalum Porto da Passagem, haud longe ab urbe Laguna" ( Saint-Hilaire 1829), May 1820, A.Saint-Hilaire Cat. C2 1795 (P 733565!, iso-lectotype P 733564!).

Taxonomic notes:— This species has a wide distribution, occurring from the southern United States to Uruguay, is very frequently collected, and has a very troubled nomenclatural history. First published by Saint- Hilaire and Moquin-Tandon (1828) as a variety of P. paludosa , it was revised 143 years later by Wurdack & Smith (1971) and renamed P. leptocaulis Torr. & A.Gray. Wurdack gave priority to the name P. leptocaulis over P. paludosa , arguing that the type variety of P. paludosa is a later synonym of P. tenuis De Candolle (1824: 329) , whereas only P. paludosa var. longispicata is a taxonomic synonym of P. leptocaulis . Since 1974 the name P. leptocaulis has been in wide use until it was changed again to P.tenella Willdenow (1803: 878) by Bernardi (2000), who placed P. gracilis Kunth (and many other names) in synonymy with P. tenella . However, P. tenella , considered a synonym of P. paniculata for almost two centuries prior to Bernardi’s treatment (2000), was synonymized again under P. paniculata by Rankin-Rodriguez (2001), with P. leptocaulis and P. gracilis again considered different species ( Rankin-Rodriguez 2003).

I now find enough evidence to make yet another update on its nomenclature. The species known at present as P. leptocaulis clearly corresponds to the original plate of Polygala appendiculata , as described by Vellozo (1829), which is the oldest known name for this species. Although Vellozo’s plate of P. appendiculata could be seen as somewhat similar to P. gracilis , the latter species has never been collected in Rio de Janeiro and seems to be endemic to the cerrado areas (a specific type of vegetation). Furthermore, Vellozo’s original description notes that P. appendiculata occurs in disturbed areas near the Atlantic, in a vegetation type called restinga, where P. leptocaulis is frequently found.

Nevertheless, P. appendiculata is traditionally placed in synonymy of P. paniculata ( Bennett 1874) . In fact, these two species should not be easily confused, considering the differences in their habitat and, particularly, differences in their inflorescences. All of the above considerations allow for a confident linking of P. appendiculata of the Vellozo plate with P. leptocaulis .

I chose as epitype material from the restinga area of Arraial do Cabo, which is one of the most likely original localities for P. appendiculata . The coastal region of Arraial do Cabo was near one of the military bases in the 18th century and was well known for its traditional cassava festival.