Eugenia neolatifolia Sobral & M.A.D.Souza, nom. nov.

Replacement name for Eugenia aubletiana Mattos (2005: 9) sensu Proença et al. (2020: 50), not Mattos (2005: 9).

Lectotype:— FRENCH GUIANA. Karouany (“ Acarouany ”), 1857, Sagot 272 (sheet P 04722646, designated by Proença et al. 2020: 50, [image]; isolectotypes BR 530407 [image], K 000565050 [image], K 000565051 [not seen], P 05229281 [image], P 05229285 [image], S 052964 [image], U 0132247 [image], U 1440846 [image]).

= Calycorectes latifolius (Aubl.) O. Berg (1861: 701). = Eugenia latifolia Aublet sensu O. Berg (1861: 701), not Aublet (1775: 502). = Calycorectes bergii Sandwith (1932: 212), replacement name for Calycorectes latifolius, not Eugenia bergii Niedenzu (1893: 81).

= Catinga oblongifolia Sagot (1885: 197), nom. illeg., replacement name for Calycorectes grandifolius O. Berg (1855 –1856: 317), not C. latifolius (Aubl.) O.Berg. = Eugenia oblongifolia Niedenzu ex T.Durand & B.D.Jackson (1902: 164), nom. illeg., not Eugenia oblongifolia Duthie (in Hooker 1878: 491).

Trees 3–12 m tall. Twigs glabrous, applanate, internodes 50–60 × 2–3 mm. Leaves glabrous; petioles 7–10 × 1.8–2 mm, semiterete, adaxially sulcate or involute; blades elliptic to oblong, 180–220 × 50–85 mm, 2–4 times longer than wide, concolorous or slightly discolorous when dry, dull green, brown or dark brown adaxially and lighter abaxially (sometimes abaxially darker, as in e.g. Hamada 93 and Sang LBB-16244); glandular dots 15–20/mm², smaller than 0.1 mm in diameter, scarcely visible; base cuneate; apex acuminate in 7–11 mm, occasionally with an apiculum to 1 mm; midvein impressed adaxially and strongly raised abaxially; lateral veins 10–15 at each side, plane or slightly sulcate adaxially and raised abaxially, leaving the midvein at angles 60–70°; secondary lateral veins and higher order venation usually perceptible; intramarginal veins usually three, the inner one formed by the loops of the lateral veins, 7–15 mm distant from the margin, the middle one 2–5 and the outer one 1–1.5 mm respectively, the margin itself plane or slightly revolute. Inflorescences axillary or ramiflorous, fasciculiform or shortly racemiform, the axis 2–10 × 2 mm, with 2–8 flowers, covered with simple appressed or ascending grey or brown trichomes 0.1–0.2 mm; bracts triangular, to 1 × 1 mm, scarcely pilose as the axes; pedicels 7–17 × 1–1.5 mm, pilose as the axes; bracteoles narrowly triangular, to 2 × 1 mm, occasionally persisting after anthesis, with scattered trichomes as the axes; flower buds uniformly pilose or with the indumentum more dense on the ovary, to 10 × 7–8 mm, with the calyx completely fused with an apiculum to 1 mm; flowers with ovary 3–4 mm, externally with 8 visible longitudinal ridges; calyx tearing in four lobes, these irregular in size and shape, mostly triangular or elliptic, 4–8 × 2–7 mm, glabrous adaxially; calyx tube present, in the specimen examined not tearing along with the lobes and strongly reflexed and enveloping the apex of the ovary after anthesis; petals four, glabrous, elliptic, to 12 × 6 mm; stamens with filaments to 10 mm, the anthers oblong, to 1 × 0.4 mm; staminal ring 4–5 mm in diameter, sometimes subquadrate, with scattered grey trichomes to 0.1 mm, not tearing in the anthesis; style glabrous, to 10 mm, the stigma punctiform; ovary with two internally glabrous locules and 18–20 ovules per locule. Fruits globose, subglobose or oblate, 20–50 mm × 22–40 mm, with simple grey or brown trichomes to 0.2 mm, yellow when ripe (according to Sagot; see below), smooth or longitudinally ridged, crowned by the remnants of the calyx; seeds one or two per fruit, not examined.

Distribution, habitat and phenology:— Eugenia neolatifolia is a small tree from primary forests presently known from Brazil, French Guiana, Guiana and Suriname. Specimens from Colombia identified under the nonexistent name “ Eugenia oblongifolia Sagot ” (e.g. Pittier 1595, [US, image], Romero 1078 [US, image], and Smith 435 [BR, K, US, images]) indeed represent E. moritziana Karsten (1848 –1849: 18), a species not related to E. neolatifolia. Eight collections from Bagua, department of Amazonas, Peru, kept at herbarium MO and identified as Calycorectes bergii (Días S. 4158, Jaramillo 236, Jaramillo 743, Rodríguez 890, Rojas 519, Vázquez 18691, Vázquez 20950 and Vázquez 21550) could represent a notable extension of the species’ distribution area; unfortunately such specimens were not located by MO staff (Boyle, pers. comm.). Flowers and fruits were collected year-round. Although vegetative and floral characters are quite similar when the examined specimens of E. neolatifolia are compared, the wide variation in fruit morphology is somewhat unusual. Sagot (1885: 197) describes the fruit of Catinga oblongifolia as about the size of a plum, yellow when ripe, smooth, with glandular epicarp and bearing one seed (not all these informations are given in his publication, but a sheet in Sagot’s handwriting attached to specimen P 05229283 bears additional information on the morphology of flowers and fruits). Collections of E. neolatifolia bear fruits that are either smooth (e.g. Andel 5044 and Hamada 93), slightly costate (e.g. Benoist 962, Bordenave 887 and Sastre 6457), and strongly longitudinally costate (e.g. Foresta 549, specimen at US; Hoffman 6005 and 6315, specimens at NY; and Molino 1209).

Vernacular names:—Goyavier sauvage (according to collection Wachenheim 32), irakopi (Richard 18), iwã u (Molino et al. 2022: 598), koi jaran (Hoffman 6315), mïroko enu (Hoffman 6005), twau, wtlãu (both from Grenand 2843), wila u (Molino et al. 2022: 598).

Specimens examined:— BRAZIL. Amapá: Amapari, Serra do Navio, rio Amapari, 23 November 1954, fl., Cowan 38549 (IAN!, INPA!); ibidem, 25 November 1954, fl., Cowan 38605 (NY, image); ibidem, 4 March 2006, fr., Hamada 93 (HUFSJ!, INPA!). Pará: Rio Trombetas, local Porteira, 5 December 1973, fl., Pena 434 (IAN!). FRENCH GUIANA. Camopi: Saut Tampac, riv. Camopi, 22 January 1992, fl., Grenand 2843 (P, image). Cayenne: Cayenne, without date, Leblond s. n. (P 00297777, image). Kourou: Kourou, without date, Richard 18 (P 05229274, P 05229288, images). Mana: Crique Simon, basse crique Courouaïe, 9 January 1970, fl., Oldeman B –2653 (P, image); Karouany, 1857, Sagot s.n. (P 05229286, image—with notes by Sagot on the synonymy of Calycorectes and Catinga, written “ Catinga oblongifolia ”). Roura: Riviére Comté, 13 December 1968, fl., Oldeman B –1998 (P, image); ibidem, 14 December 1968, fl., Oldeman B –2021 (K, P, images). Saint-Élie: Saint-Élie, 21 October 1978, fr., Sastre 6457 (P, image); 22 October 1980, fl., Prévost 1073 (P, image); 1 January 1984, fr., Foresta 549 (P, U, US, images); 4 September 1984, fl., fr., Foresta 619 (U, P, images); 2 December 1984, fr., Molino 1209 (P, image); 5 February 1986, fr., Prévost 2121 (ASU, K, images); 17 February 1986, fr., Riéra 1140 (ASU, image); 15 November 1998, fl., Poncy 1157 (P, image); 16 July 2002, fl., Prévost 4542 (F, image); Saut Nacibo, 25 March 1994, fr., Bordenave 887 (P, image); Saut Stephanie, bassin du Sinnamary, 20 November 1989, fl., Hoff 6022 (P, image); 9 January 1992, fl., Hoff 7283 (P, US, images); Saut Takari-Tanté, bassin du Sinnamary, 19 November 1989, fl., Hoff 5958 (ASU, P, images); 3 December 1994, fl., Prévost 3080 (ASU, P, U, images). Saint-Georges de l’Oaypock: Armontabo, 17 May 1985, fr., Sabatier 1092 (ASU, image). Saint Laurent du Maroni: Charvein, 11 January 1914, fl., Benoist 585 (P, image); Godebert, November 1919, fl., Wachenheim 32 (F, P, US, images); Île Portal, without date, Sagot s.n. (U 0155062, US 00619193, images); Saint Jean, 8 July 1914, fr., Benoist 962 (P, image); 8 July 1914, fl., Benoist 1016 (P, image). Without department: Riviére Maroni, 1861, fl., Mélinon 147 (P, image); ibidem, 1861, fl., Mélinon 154 (P, image); without date, fl., Mélinon 18 (F, image). Without locality, without date: Sagot s.n. (P 05229283, image—with notes by Sagot on flower and fruit morphology). Without locality, without date, Sagot s.n. (P 05229317, image). GUIANA. Potaro Siparuni: Iwokrama, 19 March 1997, fr., Clarke 4121 (US, image). Without department: Essequibo river, Morabali Creek, 5 November 1929, fl., Sandwith 558 (P, US, images); Essequibo river, Kamuni Creek, 21 April 1944, Maguire 22925 (U, US, images). SURINAME. Brokopondo: Brownsberg, 4 May 1917, fr., Bureau v.h. Boschwezen (“B.W.”) 4303 (F, image); ibidem, 4 May 1917, fl., Bureau v.h. Boschwezen (“B.W.”) 6352 (BR, image); 8 November 1973, fl., Vreden 13665 (U, image); 16 February 1977, fr., Lindeman 17 (U, image); 21 October 1977, fl., Sang LBB-16224 (ASU, image); 24 January 1999, fr., Poncy 1174 (P, image); 20 November 2003, fl., Andel 4369 (U, image); 20 November 2003, fl., Andel 4493 (U, image); 23 November 2003, fl., fr., Andel 4505 (U, image). Marowijne: Albina, 11 March 2006, fr., Andel 5044 (U, image). Sipaliwini: Kwamalasamutu, 5 March 2005, fr., Hoffman 6005 (L, NY, US, images); 10 June 2005, fr., Hoffman 6315 (L, NY, US, images).

3.2. Nomenclature

Berg (1861: 701) described Calycorectes latifolius based on a collection made by Louis Claude Richard in French Guiana labeled as Eugenia latifolia. Since the flowers in Richard’s specimen opened by tearing of the calyx lobes, Berg viewed this plant as a species of Calycorectes rather than Eugenia. So, he took up the Aubletian name and dubbed the combination Calycorectes latifolius.

Berg (1855 –1856, 1861) did not examined Aublet’s specimens (see comments by Amshoff 1950: 16 and McVaugh 1969: 222), having to rely in the descriptions and plates of the Histoire des plantes de la Guiane Françoise (Aublet 1775), which led him to a misinterpretation of Eugenia latifolia. In his revision of American Myrtaceae (Berg 1855 – 1856: 224) he took this name to a specimen collected by Heinrich Wullschlaegel in Suriname. Later, in a supplement to this revision (Berg 1861), he interpreted differently Eugenia latifolia, regarding it as conspecific with the Richard collection and without making any later mention to the Wullschlaegel specimen (which nearly one century later would be named Eugenia wullschlaegeliana Amshoff —see Amshoff 1948: 20).

Sagot (1885), in his treatment of the plants of French Guiana, interpreted Calycorectes as congeneric to the earlier Catinga Aublet (1775: 511) due to the irregular opening of the calyx in both genera (Sagot 1885: 196–197; see also handwritten notes annexed to specimen P 05229286). His name Catinga oblongifolia is followed by the statement “ Calycorectes grandifolius Berg revis.”, that is, a revised concept of the Bergian species with a replacement name for it (not a new species), rendering Catinga oblongifolia illegitimate due to the inclusion of the legitimate C. grandifolius as a synonym—see Article 52.1 (Turland et al. 2018: 93): “A name (...) is illegitimate and is to be rejected (...) if the taxon to which it was applied, as circumscribed by its author, definitely included the type of a name that ought to have been adopted, under the rules”). Notwithstanding the mention to Calycorectes grandifolius, Sagot’s short description is more suggestive of C. latifolius than of C. grandifolius. It is possible that the reference to C. grandifolius was a misspelling, since the specimens collected by Sagot and labeled by him as Catinga oblongifolia match the Richard collection identified by Berg as Calycorectes latifolius, and one specimen of Sagot 272 (P 04722646) was identified in Berg’s handwriting as “ Calycorectes latifolius β. longifolius ” (as far as we could check he neither published this combination nor cited Sagot’s specimen).

Sandwith (1932: 212) coined Calycorectes bergii as a replacement name to Calycorectes latifolius, recognizing that Eugenia latifolia and C. latifolius are distinct species: in his words, “By the International Rules of Nomenclature, the combination C. latifolius Berg must be applied to Aublet’s species; the above name [i.e., Calycorectes bergii] is accordingly proposed for Berg’s plant” (Sandwith 1932: 213). Differently from Berg, Sandwith had access to Aublet’s types, and pointed the differences between Eugenia latifolia and Calycorectes latifolius: “Aublet’s species is, however, a very different plant, as is proved by his specimens in Herb. Mus. Brit. [i.e., herbarium BM], which agree well with his figures and description. It suggests a true Eugenia rather than a Calycorectes, and differs from C. latifolius Berg in the shape and venation of the leaves, the thin black glabrescent pedicels and hypanthium, and the very small sepals which are only 2 mm long” (Sandwith 1932: 212–213). After citing the specimens Sandwith 558, Sagot 272, Wachenheim 32 and Mélinon s.n., he also suggests a confusion between the names Calycorectes latifolius and C. grandifolius: “The last two collections were distributed from Paris under the name C. grandifolius Berg, which, to judge for the description, is at once distinguished by the venation of the leaves. The above material [i.e. the four specimens cited] agrees very well with Berg’s description of his C. latifolius which he based on Eugenia latifolia Aubl. ” (Sandwith 1932: 212). It is interesting to note that Sandwith does not mention Catinga oblongifolia in his paper.

Although recognizing Calycorectes bergii as distinct from Eugenia latifolia, Sandwith (1932: 212), stated C. bergii as “ nom. nov. ”, i.e., a replacement name. Article 48.1 (Turland et al. 2018: 90), states that “when an author who adopts a name refers to an apparent basionym or replaced synonym but explicitly excludes its type, the name of a new taxon is considered to have been published that must be attributed solely to that author”. As Sandwith excluded Eugenia latifolia from his definition of Calycorectes bergii, consequently excluding its type, he defined a new taxon, although not designating a type for his name, what has been done nearly 90 years later by Proença et al. (2020: 50), who elected as lectotype one of the specimens cited by him, Sagot 272.

Amshoff (1942: 156) recognized Calycorectes and Catinga as distinct genera because of the supposedly different number of petals—six in Calycorectes and four in Catinga. Coherent to this concept, she recovers Catinga oblongifolia and takes this name as valid against Calycorectes bergii, also stressing the distinction between this species and Eugenia latifolia. In her words: “ Calycorectes latifolius (Aubl.) Berg (...) (quoad descript. tantum, non quoad nomen, non Eugenia latifolia Aubl. [i.e., ‘ Calycorectes latifolius as far as the description, not as far as the name, and not Eugenia latifolia ’])”. The distinction between Calycorectes and Catinga persists in Amshoff’s treatment of the Myrtaceae from Suriname (Amshoff 1951), but then she takes Catinga oblongifolia as a synonym of Catinga moschata Aublet (1775: 511; = Eugenia moschata (Aubl.) Niedenzu ex T.Durand & B.D.Jackson, 1902: 164).

McVaugh (1969: 224) succinctly cites Catinga oblongifolia as a synonym of Calycorectes grandifolius as far as the name is concerned, excluding its description (“quoad nom., non quoad descr.”, i.e.: “as far as the name, not as far as the description”), without any comment on the distinct nature of Sagot’s description if compared to C. grandifolius.

Calycorectes is nowadays considered as a synonym of Eugenia (see Mazine et al. 2020 and Giaretta et al. 2021), since the type species, C. grandifolius (see Amshoff 1942: 156; replacement name: Eugenia neograndifolia Mattos, 2005: 10), is nested in an inclusive Eugenia (Giaretta et al. 2021: 136), more specifically in section Umbellatae O. Berg (1855 –1856: 204). Mattos (2005) presented a nomenclatural synopsis of Calycorectes, including several new combinations and new names. In this paper he proposes Eugenia aubletiana (2005: 9) as a replacement name for Eugenia latifolia, Catinga moschata, Calycorectes latifolius, Calycorectes bergii and Calycorectes oblongifolius Sagot (nonexistent name, but an obvious misspelling of Catinga oblongifolia, since Sagot’s paper is cited after the name). Eugenia aubletiana is illegitimate under Art. 52.1 (Turland et al. 2018: 93; see full citation above), since Mattos included among the synonyms of E. aubletiana the older and valid E. latifolia and this name should have been taken for his diverse assemblage of species.

Eugenia aubletiana was accepted by Mazine et al. (2020) in their list of Brazilian species of Eugenia as a valid name for Calycorectes bergii, as well as by Molino et al. (2022: 598) in their checklist of tree species from French Guiana. Proença et al. (2020) also recognized Eugenia aubletiana, interpreting it as the valid name for C. bergii although at the same time recognizing that Eugenia latifolia was the correct name to be used when considering the species listed by Mattos (2005), a procedure not supported by the Code (Art. 52.1, Turland et al. 2018: 93). In the same paper, Proença et al. proposed a lectotype for E. aubletiana.

Consequently, Calycorectes bergii is a species proposed as new by Sandwith (1932: 212) and lectotypified by Proença et al. (2020: 50). As the name Calycorectes bergii cannot be used under Eugenia due to the existence of an earlier homonym, and the name considered as valid by Proença et al. (2020: 50) is illegitimate, it is necessary a replacement name for this widely distributed Amazonian species.

3.3. Key to the species cited

The present key is intended to briefly distinguish the species cited here. Eugenia wurdackii (McVaugh) Mattos (2005: 10; basionym: Calycorectes wurdackii McVaugh, 1969: 225) is included in the key because, among the species formerly described under Calycorectes, is the most morphologically close to Eugenia neolatifolia.

1. Flower buds with well defined sepals ................................................................................................................................................2

–. Flower buds with sepals visibly fused along most of its extent .........................................................................................................4

2. Sepals to 2 mm .......................................................................................................................................................... Eugenia latifolia

–. Sepals longer than 2 mm ....................................................................................................................................................................3

3. Blades essentially glabrous..................................................................................................................................... Eugenia moschata

–. Blades at least abaxially uniformly covered by grey indumentum ...................................................................... Eugenia moritziana

4. Intramarginal vein less than 2 mm from the margin, sometimes indistinguishable from it ........................... Eugenia neograndifolia

–. Intramarginal vein clearly distinct from the margin, 2–15 mm from it..............................................................................................5

5. Blades 200–220 × 60–85 mm, mostly glabrous; petioles to 8 mm, 3–4% of the blades’ length; intramarginal veins usually three, the inner one 5–12 mm, the middle and outer intramarginal veins 3–4 mm and 1–1.5 mm from the margin respectively, the proportion of the blade outside the inner vein 25–30% of the width at its widest point....................................................... Eugenia neolatifolia

–. Blades 140–170 × 50–70 mm, at least abaxially covered by minute trichomes; petioles 10–14 mm, 7–8% of the blades’ length; intramarginal veins usually one perceptible, 2–5 mm from the margin, occasionally a second vein less than 1 mm from the margin, the proportion of the blade outside the inner vein 8–15% of the width at its widest point................................... Eugenia wurdackii