Dicorynia paraensis Benth.

Falcão, Marcus José De Azevedo, Torke, Benjamin Marland & Mansano, Vidal De Freitas, 2022, A Taxonomic Revision of the Amazonian Genus Dicorynia (Fabaceae: Dialioideae), Phytotaxa 554 (1), pp. 1-31 : 15-17

Dicorynia paraensis Benth. in Hook. J. Bot. 2(10): 82–83 (1840).

TYPE: — BRAZIL. “Pará” [probably actually the state of Amazonas], 1781–1789, Unknown collector [probably Richard, L.C.M.] P 02142577 (Lectotype: P!; Isolectotype: P!). Designated by R. C. Koeppen (1967).

Medium to large trees, (6–) 8–35 (–40) m tall; trunk 10–100 cm in diameter; buttresses up to 2 m tall. Leaves (10–) 14–42 (–50) cm long, leaflets 5–11 (–13), elliptical, ovate, or oblong, generally acuminate to cuspidate at apex, more rarely acute or caudate, base obtuse, cordate, truncate, or cuneate, the abaxial face glabrous to sparsely pubescent, papillate or not, terminal leaflets (5.5–) 8–18 (–21) × (2–) 3.5–9 (–13) cm, the length 1.4–4 times the width; petiolules (2–) 3–10 (–12) mm long; axillary buds laterally oblong, elliptical, obovate, orbicular or lanceolate, being acuminate, acute, or obtuse at apex, 2–7 × 1–4 (–5) mm, close or apart the leaf attachment point; terminal buds strongly capitate, usually orbicular at apex, more rarely acuminate, 4–8 × 5–10 mm. Inflorescences (9–) 13–32 (–45) × 4–26 cm; sericeous to pilose, rarely tomentose, golden to brown to dark brown; bracts 4–10 × 2–5 mm; pedicels 4–7 mm long. Flower buds 7–11 × 5–7 mm; sepals 7–11 × 4–7 mm; petals 8–16 × 6–13 mm, claws 2–5 mm; filament of longer stamen 6–10 × 1–3 mm, filament of shorter stamen 2–4 × 1–3 mm, anther of longer stamen with 4 sporangia, anther of the shorter stamen with 8–11 sporangia in the median to apical region, both anthers 3–5.5 × 2–3.5 mm; carpel 3–6 (–8) × 1–3 mm, style 3–6 mm long, ovules 2–4. Fruits (3–) 4–8.7 (–9) × 2.5–4.5 × 0.3–0.5 cm, wing (2.5–) 4–7 mm wide; seeds 1.5–2 × 1–1.7 × 0.4 cm. (Figs. 3–4, 6B, D–H; 7–8; 9A–D, H, K, N–O, R).

Distribution, Habitat, Biogeographical Patterns, and Ecology: — Dicorynia paraensis is distributed in the Amazon basin, mainly north of the Amazon River in in the Brazilian states of Amazonas, Pará and, rarely, Amapá. It also occurs in adjacent portions of Venezuela, Colombia, and Peru (Fig. 5). Its presence in Peru and in the Brazilian state of Amapá is first cited here ( Dávila , N. 1406 F; H. Medeiros 3114 RB766186). Its presence in the state of Roraima is not confirmed. The few specimens cited for this state are sterile Papilionoideae (J.G. Soler 96 RB00141726) or misidentified Martiodendron excelsum . Ducke (1948) already mentioned the total absence of the genus in the Branco River (a tributary of the Negro that crosses Roraima). However, due to the state’s proximity to the areas where Dicorynia occurs in the upper Negro River and upper Uatumã and the existence of similar environments in the region, the possible occurrence of the genus in Roraima must not be discarded. Some existing references to the occurrence of the genus in the state of Tocantins are just confusion with the Tonantins River, in the state of Amazonas (Fig. 5).

The species occurs in “várzea” and “igapó” seasonally flooded forests, campinarana forests, and humid “terra firme” ombrophilous forests, on both sandy and clay soils, never far from rivers. It was considered once one of the most common trees along the banks of the Negro River and its larger tributaries ( Ducke 1948, Horn 1948). However, outside this limited region, it is considerably rare. Since the distribution is mainly riparian, the seasonality of floods in these environments is likely important for seed dispersal in at least three of the four varieties of this species (var. ingens, var. paraensis , and var. uaupensis ). The dry fruits contain a very thin seed chamber with air, which allows them to float considerable distances from the parental trees, often washing ashore on riverside beaches. In the same habitats, the close related genera Apuleia and Dialium also exhibit similar partial hydrochoric dispersal via floating dry fruits ( Falcão, 2020b, unpubl. Data, present work). In the case of Dicorynia and Apuleia , the laterally compressed and modestly winged morphology of the fruits also allows for some level of anemochoric dispersal. However, the distances that seeds travel by this mode appear to be quite limited ( Jesel, 2005). However, anemochory is presumably the primary mode of dispersal in the variety macrophylla since it occurs mainly in “terra firme”.

In D. paraensis , the habitat seems to be of great importance in infra-specific delimitation, mainly between varieties macrophylla and paraensis , which almost always occupy, respectively, “terra firme” and “igapó” forests. This also happens in São Gabriel da Cachoeira, in the upper Negro River, the only area where both varieties co-occur and where some level of reproductive isolation between them may still exist. Variety macrophylla can rarely occur in flooded forests, with two observed individuals in “várzea” in the Solimões River, a “white water” river, and three individuals in “igapó” in lower Negro River and Uatumã River, both “black water” rivers. In the lower Amazon River, var. macrophylla possibly occupies a different habitat than var. ingens , whose two known individuals occur in possibly floodable areas on the riverside. However, this should be taken with caution due the small number of known individuals of var. ingens . Var. uaupensis occurs in two different environments depending on the region: on Japurá River, a “white water” river where only this variety occurs , all individuals occur in “várzea” areas. It occurs in “terra firme” forests in the upper Negro River basin, in Uaupés, Içanã, and Guainia Rivers to the south of Venezuela, all “black water” rivers. Finally, in the lower Negro River, from Anavilhanas to Manaus, and in the Uatumã River, both “black water” rivers, the variety is almost always in “igapó” forests. Although these differentiated environmental patterns are quite strong, the environmental conditions that may be leading to this variation in areas of occurrence remain a mystery (Figs. 5; 10).

With the infrageneric circumscription proposed here for D. paraensis , only var. uaupensis and var. paraensis occur in Venezuela, only var. uaupensis occur in Colombia, and only var. macrophylla occur in Peru. In Brazil we have var. macrophylla in Amazonas and Pará, var. ingens only in Pará, var. paraensis only in Amazonas and var. uaupensis in Amazonas and, rarely, in Amapá. Due to some of its morphological characteristics and distribution, the four varieties of D. paraensis form two more closely related groups: one with var. paraensis and var. uaupensis , taxa with smaller leaves and fewer leaflets, being taxa mainly restricted to the Negro, Japurá, and Uatumã River basins, more to the north of the genus distribution. The other group, with var. macrophylla and var. ingens , with larger leaves and more leaflets, occurring mainly in Solimões and Amazon Rivers, to further south of the genus’ distribution range, with a small population of var. macrophylla restricted to the region of São Gabriel da Cachoeira, on the upper Negro River. It is not known whether these two distribution centers communicated across Negro River or the north of the Solimões basin and whether the variety still occurs between these two areas and lacks collections, or whether the two populations are currently isolated (Fig. 5).

In the taxonomic circumscription proposed by Koeppen (1967), most of D. paraensis varieties have an overlapping geographical distribution. With the circumscription presented here, biogeographic and ecological patterns could be proposed. D. paraensis var. ingens occurs in comparatively less humid areas concerning the other varieties (2000 mm). Var. macrophylla , and D. guianensis are in intermediate precipitation regions (2200 and 2800 mm), and var. uaupensis and var. paraensis present a greater amplitude (2200-3700 mm) (Fig. 10). Besides that, the Amazonian rivers seem to play an important role in the possibly hydrochoric dispersion of the fruits in D. paraensis , as the genus is almost absent from the south of the Amazon River, being present along its main tributaries to the north, with rare exceptions in the lower Maués River and the far west of Brazil and far east of Peru in the basin of the Juruá and Solimões Rivers. Also, different varieties are partially contained in basins of different rivers (Fig. 5), which may indicate that riverine barrier processes may interfere in Dicorynia distribution. Due to the dispersal difficulty, the species can have reduced gene flow from side-to-side of wider rivers such as the Amazon. This biogeographic pattern was also observed in other plants and animals. The apparent absence of the species in most of the southern Amazon could also be related to seasonality due the region’s relatively low precipitation in the drier months of the year ( Wallace 1852, Collevati et al. 2009, Nazareno et al. 2017, 2019).

A variety of interactions between D. paraensis and animals have been observed. For example, insect galls frequently occur in fruits of specimens from throughout the distribution of the species (Nascimento, O.C. 216 IAN), and the seeds are often consumed by Coleopteran larvae on fruits still on the trees (Falcão, personal observation). Since the identical fruits of D. guianensis are reported as commonly consumed by parrots ( Jesel 2005), the possible ecological importance of D. paraensis to Amazonian avifauna should also be studied. The flowers of the species are odorous and probably pollinated by bees ( Dulmen et al. 2001). Collectors have noted associations with ants, which access flowers or build nests in the branches (Hervé Galliffet; personal observation, Lima, H.C. 1987 MO).

Etymology: —The epithet refers to the collection of the type in Pará, at the time corresponding to “Grão-Pará”, which encompassed the modern states of Pará and Amazonas. Since the specific locality is unknown, it is not possible to know with certainty from which of these two modern states the type collection originated. However, given the distribution of var. paraensis , it is likely that the collection originated from the Negro River basin in Amazonas.

Phenology: —This species blooms from October to January, with peak flowering usually in November; it rarely blooms from February to April. Fruiting commences in January. Old fruits remain on trees until flowering occurs the following year, this is a recurrent pattern in closely related genera of subfamily Dialioideae .

Uses: —Despite the widespread commercial harvest of D. guianensis in French Guiana and Suriname and apparent similarity between its timber and that of D. paraensis ( Milanez and Mattos Filho 1959, Koeppen 1967), the commercial potential of the timber of the latter species is little developed, with only rare reports of its use in building construction in the Brazilian state of Pará ( Ducke 1948, Koeppen 1967). Fruits, although dry, are rarely mentioned as edible ( Koeppen 1967, R.L. Fróes 21539).

Conservation: —The estimated EOO for D. paraensis of 1,599,196 km is expansive. It is distributed in several protected areas, such as Reserva Florestal Adolpho Ducke, Reserva Biológica de Uatumã, and Parque Nacional de Anavilhanas in Brazil, which implies a conservation status of Least Concern. However, unlike the relatively abundant D. guianensis , this species appears to be rare, except in some areas of the Negro River drainage ( Macbride 1943, Ducke 1948, Horn 1948, present work). Moreover, since D. paraensis is mostly restricted to riverine areas, the Area of Occupancy (AOO) is likely to be only a small fraction of the EOO. The riverside seasonally inundated zone is at relatively high risk from anthropogenic alterations, such as increasing deforestation on several Amazonian regions, even in legally protected areas ( Potapov et al. 2017; Montibeller et al. 2020; INPE, 2020) and altered hydrology due to hydroelectric development, as occurred, for example, with the construction of Balbina dam in Brazilian Uatumã River basin, one of the main areas of occurrence of D. paraensis . Thus, efforts should be made to better understand the impacts of these threats on population trends.

Vernacular Names: —Angélica, Angélica-do-Pará, Ingazeiro, Cedrinho, Cumarurana-vermelha, Cumaruranaroxo, Itauba-rana, Macucu-roxo, Taboarana, Tachi-branco, Tapaiúna and Tapanhaúna. The species appears to be little known, even among natives, lacking popular names in numerous regions. Several of the vernacular names are also used for other genera of Fabaceae , such as Tachi-branco for Tachigali Aubl. , Cumarurana-roxo for Dipteryx Schreb., Taboarana for Acosmium Schott and Leptolobium Vogel , and Itauba-rana for Sweetia Spreng. , Acosmium , and Leptolobium .

Taxonomic Comments: — Koeppen (1967) briefly indicates differences in axillary buds of some varieties of D. paraensis . Those descriptions are subjective and, in some occasions, were not confirmed in the present work. However, the axillary buds, when observed laterally, proved here to be a strong taxonomic character. Under frontal view, they are homogeneous, with a triangular to orbicular shape, but laterally, the buds in var. macrophylla are large and generally oblong/elliptical, often acuminate at apex, this character being practically absent in other varieties. Rare specimens of var. macrophylla have large orbicular buds. In var. paraensis , the buds are always small and narrowly lanceolate to lanceolate. Such buds are not found in other varieties. Buds in var. uaupensis are relatively variable, more frequently narrow obovate, commonly apart the leaf’s attachment point, while in other varieties, the bud is closer to the leaf pulvinus. Less commonly, they are small orbicular to oblong and, in rare cases, large oblong and acuminate, as in var. macrophylla . Finally, var. ingens have small orbicular buds, different from var. macrophylla , on which, when rarely orbicular, they are much larger. (Fig. 6).

Identification key to varieties in Dicorynia paraensis

1. Leaflets generally 5–7, very rarely 9; axillary buds laterally lanceolate, or narrowly obovate, rarely orbicular or oblong; leaves (10–) 14–31 (–40) cm long; Colombia, Venezuela, and Brazil.

2. Leaflets 5(–7), generally 1.5–2 times longer than wide; petiolules (5–) 7–10 (–12) mm long; axillary buds commonly deciduous, laterally lanceolate, being always less than 0.5mm above the leaf’s insertion point, 1.5–3 (–4) × 0.5–2 (–3) mm; occurring almost always in “igapó” forests............................................................................................................. Dicorynia paraensis var. paraensis

2. Leaflets 7, very rarely 9, about 2–3 times longer than wide; petiolules (2–) 4–7 (–9) mm long; axillary buds not deciduous, laterally narrowly obovate and acuminate, less commonly oblong or orbicular, being generally 1–4mm apart from the point of insertion of leaves, (2–) 3–5 (–7) × 1–3 (–4) mm; occurring in “terra firme”, “várzea”, or “igapó” forests .......................................................... ..................................................................................................................................................... Dicorynia paraensis var. uaupensis

1. Leaflets almost always 11, very rarely 9 or 13; axillary buds laterally orbicular, elliptical or oblong; leaves (20–) 26–50 cm long; Peru and Brazil.

3. Larger leaflets ovate to oblong, less than 2 times longer than wide; petiolules 3–4 (–7) mm long, practically hidden by the leaf blade; axillary buds orbicular, obtuse at apex, ca. 3 × 2 mm; Brazil (basin of the Trombetas River in Pará) ..................................... ........................................................................................................................................................... Dicorynia paraensis var. ingens

3. Larger leaflets usually narrowly oblong, about 2.5–3.5 times longer than wide; petiolules (3–) 6–8 (–10) mm long; axillary buds laterally oblong to elliptical, usually acuminate at apex, more rarely obtuse at apex, (4–) 6–7 × 3–4 (–5) mm; Peru, Brazil (Amazonas and Pará)............................................................................................................... Dicorynia paraensis var. macrophylla