A new species, Ormosia neillii (Fabaceae), is described and illustrated from the tepui-like formations of the Cordillera del Cóndor Region in south-eastern Ecuador. Morphological similarities with other species of Ormosia are examined and discussed. Based on IUCN guidelines, a preliminary conservation status of Endangered (EN) is recommended.

Endemism, Fabaceae, Papilionoideae, tepui-like formation, white sands

Ormosia Jackson is a papilionoid legume genus (Fabaceae, Papilionoideae) with tropical and subtropical distribution, comprising more than 150 species from treelets and shrubs to canopy trees (Rudd 1965; Cardoso and de Queiroz. 2010; Cardoso et al. 2012; Cardoso et al. 2014; Torke et al. 2022). The geographical distribution of Ormosia constitutes a clear example of an Asian-American tropical disjunction pattern and recent molecular evidence suggests a Palaeotropical origin in Asia (Torke et al. 2022). Historically, Ormosia has been classified within the polyphyletic tribe Sophoreae, but molecular evidence suggests that Ormosia, together with the genera Clathrotropis Harms, Panurea Spruce ex Benth. and Spirotropis Tul., can be grouped in the so called “Ormosieae clade” (Cardoso et al. 2013, 2017). More recent phylogenetic studies suggest that Clathrotropis, Panurea and Spirotropis share a recent common ancestor and are referred to as the Clathrotropisoid clade (Torke et al. 2022), which does not include Ormosia (Lavin et al. 2005; Cardoso et al. 2013; Torke et al. 2022). Ormosia is characterised the following features: imbricate calyx lobes; ten free stamens; an incurved style; a terminal or oblique, often bilobed stigma; and seeds with hard testa, which are frequently red, black or bicoloured red and black (Rudd 1965; Cardoso and de Queiroz 2010). In the most comprehensive taxonomic revision of Neotropical Ormosia, Rudd (1965) defined the following three sections based mostly on the morphology of seed characters: sect. Ormosia, sect. Macrocarpae Ducke and sect. Unicolores Amsh. Based on Rudd’s classification, sect. Unicolores is characterised by the unicoloured seeds (e.g. red, black or yellowish, but sometimes spots of black). In Ecuador, eight species of Ormosia have been recorded, occurring in Amazonian lowlands below 500 m a.s.l., Andean foothills up to 1000 m a.s.l. and in the Chocó Biogeographic Region in the north-western Andes of Ecuador below 500 m a.s.l. (Ulloa-Ulloa et al. 2021. Even though Ecuadorian Amazonia has been botanically well explored throughout the last 35 years, there are some regions that remain relatively unexplored or biodiversity “darkspots” with several undocumented and undescribed tree species (see Ondo et al. (2024)). The Cordillera del Cóndor, despite recent large-scale floristic inventories and botanical exploration, remains poorly known.

The Cordillera del Cóndor, a sub-Andean range situated between the Andes and the Amazon, spans about 1.1 million hectares and stretches 150 km along the Ecuador-Peru border (Neill 2005). Known as a biodiversity hotspot of endemism for both plants and animals, the northern part of this region shares floristic similarities with the Guiana Shield (Neill 2005; Ulloa Ulloa and Neill 2006). Recent discoveries on the Cordillera’s sandstone plateaus have revealed several new species (Ulloa Ulloa et al. 2012; Prance 2013; Guevara-Andino and Fernández-Alonso 2018; Fernández-Fernández et al. 2020; Clark and Neill 2023) and Incadendron K. Wurdack & Farfan, a recently described monotypic genus (Wurdack and Farfan-Ríos 2017). Our large-scale inventories on the white sand forests have also led to the identification of a previously unknown species of Ormosia with a remarkable dark purple to black corolla that is only known from the pre-montane white sand forests of this region.

We have performed floristic inventories and botanical collections in the Cordillera del Cóndor since 2017. During one of our floristic inventories, we collected specimens that represent an undescribed species in the genus Ormosia. We describe this new species, based on an analysis of morphological characters from material deposited in the following Herbaria: Herbario Nacional del Ecuador (QCNE), Herbario Amazónico del Ecuador (ECUAMZ), Marie Selby Botanical Gardens (SEL) and the Missouri Botanical Garden (MO). We also compared the new species with images of type specimens deposited in JStor Plants (https://plants.jstor.org) and reviewed voucher specimens in the virtual herbaria of the Field Museum (F), the New York Botanical Garden (NY), the Herbario Nacional de Colombia (COL) and the Herbário do Instituto Nacional de Pesquisas da Amazônia (INPA; herbarium abbreviations follow Thiers (2025)). In this work, we used the terminology and classification from Rudd (1965). Earlier works have demonstrated that a single diagnostic character, the position of the calyx in relation to the bud, is the only consistent diagnostic character for the circumscription of taxa to subgenus (Rudd 1965; Cardoso and de Queiroz 2010).

The AOO and EOO for the preliminary IUCN assessment were determined using the software package conR in the R statistical software (Dauby et al. 2017). To test the effects of deforestation, we used the most updated data on deforestation for Ecuador from the online platform Mapa Interactivo (Ministerio del Ambiente, Agua y Transición Ecológica del Ecuador 2024). We used ecosystem layers for the Cordillera del Cóndor Region and deforestation maps for this area between 1990 and 2022. Habitat reduction was then estimated combining deforestation on ecosystems with the Extent of Occurrence (EOO) and Area of Occupancy (AOO) using the clip tool in the ArcGis software (ESRI 2011). The clip tool overlays a range size map (EOO and AOO) with a boundary layer or layers, in this case, the deforestation scenarios from 1990 to 2022 and the ecosystem maps. Then, we defined the species’ potential habitat to estimate habitat loss as the area outside the combined boundary of ecosystems and deforestation maps.

Following the classification proposed by Rudd (1965) and based on our description, Ormosia neillii is recognised in the sect. Ormosia because the sericeous pubescence on the abaxial surface of leaflets; secondary veins straight and almost parallel to the mid-vein; more than nine pairs of secondary veins; the dark-purple to black corolla and the glabrescent dehiscent fruit with ligneous or subligneous valves (Figs 3, 4D–F). The mostly unicoloured light red or dark red seeds of this species is a noteworthy characteristic (Fig. 4E–F). However, the above-mentioned characters present in this species are morphologically more similar to sect. Ormosia than to members of sect. Unicolores. The new species most closely resembles O. cuatrecasasii and O. discolor. Nonetheless, O. neillii is readily distinguished from O. cuatrecasasii by smaller leaves (5.5–14.5 cm long vs. 7–24 cm long), fewer secondary veins [10–12 vs. 11–14(–16)], longer calyx tube (10–15 mm long vs. 6–7 mm long), smaller fruits (3.5–6 cm long vs. 5–10 cm long), unicoloured seeds (vs. bicoloured red and black seeds), shorter hilum (1–1.5 mm long vs. 3 mm long) and fewer seeds per pod (1–3 vs. 1–6). A summary of these characters is provided in Table 1. The distribution of O. neillii in the eastern Andes of southern Ecuador is geographically isolated from O. cuatrecasasii in the Chocó Biogeographic Region in the western Andes of northern Ecuador and southern Colombia (Fig. 6)

Ormosia neillii is also distinguished from O. discolor, a morphologically similar species that inhabits terra firme forests in Central Amazonia. However, O. neillii can be readily differentiated from O. discolor in having larger fruits (3.5–6 × 2–3 cm long vs. 2–5 × 1.5–2 cm long) with strongly cuspidate apex (vs. acute to acuminate apex), larger flowers (15–25 mm long vs. 6–8 mm long) and longer calyx tube (10–15 mm long vs. 4–6 mm long). It also differs from O. discolor in having leaflets with fewer secondary veins (10–12 vs. 15–20) and fruits glabrescent in maturity (vs. fruits covered by minutely fulvo to ferruginous-velutinous pubescence).

In a recent molecular phylogenetic study, Torke et al. (2022) suggested the monophyly of the Nobilisoid clade may be congruent with Rudd’s (1965) classification corresponding to series Nobiles Rudd. The Nobilisoid clade is defined by variation in the colouring pattern of the seeds (Torke et al. 2022) and includes the following species: Ormosia nobilis, O. macrophylla, O. krugii and O. santaremensis. This variation includes entirely red to bicoloured or entirely black seeds from single individuals or even the same pod (Rudd 1965). However, O. neillii shows a consistent pattern of monochromatic seed colour from light to dark red. In addition, the series Nobilis Rudd includes all the species morphologically similar to O. neillii and described in this study, but not all the species in this series were included in the molecular phylogenetic analysis of Torke et al. (2022). In the same study, the authors suggest that O. discolor is a divergent lineage within this clade and would constitute its sister taxon (Torke et al. 2022). Thus, there is uncertainty about the most likely placement of O. neillii in the Ormosia phylogeny, specifically within the Nobilisoid clade proposed by Torke et al. (2022).

It is also interesting to note that O. cuatrecasasii, the species that most closely resembles the new species, inhabits mostly humid forests of the Chocó Region in Colombia and no records from this species have been found in the eastern flanks of the Andes. Thus, despite the morphological similarities for both species, these forests have been geographically isolated for at least the last 10 Mya, which is the estimated divergence time for the Nobilisoid clade (Torke et al. 2022). However, a phylogenetic study including O. neillii is necessary to evaluate its sister-group relationship.

We are grateful to the Herbaria ECUAMZ and QCNE for their support and permission to access material deposited in these institutions. We are specially indebted to Domingos Cardoso for his careful reading and insightful comments during the review process of this manuscript. We are also indebted to María José Endara from the Ecology and Evolution in the Tropics-EETrop research group at Universidad de las Américas for her support during fieldwork and for commenting on earlier versions of the manuscript. We acknowledge the Ministerio del Ambiente, Agua y Transición Ecológica for granting collection and research permits to Universidad de las Américas under Genetic Resources Access Agreement number MAATE-DBI-CM-2021-0187. This work is part of the Chemical Diversity of Amazonian and Andean forests in Ecuador project and has been supported by Universidad de las Américas Research Grant FGE.JGA.21.04 to JEGA. The 2017, 2018 and 2019 field expeditions were supported by the Lawrenceville School’s Harkness Travel Program in collaboration with David A. Neill and the Universidad Estatal Amazónica. Special thanks to the Asociación de Trabajadores Autónomos San Miguel de las Orquídeas (ATASMO) for providing access to their Reserve. We are especially grateful for logistical support from our local guides, Sengundo Velez, Ivan Merino, Laura Merino, Maria Tillaguango and Luis Jimenez. We are also grateful to Clara Guillermina León García and Carlos Humberto Gálvez Guamán from Cabañas Yankuam (Yankuam Lodge) for hosting the Lawrenceville School throughout three consecutive years of field courses. Finally, we thank Efrén Merino-Santi for the illustration of the new species.