Typification of names in Pachira aquatica Aubl. (Malvaceae, Bombacoideae) with a new combination and new status from the Brazilian Amazon forest

A new combination and a new status, Pachira manausensis , is here proposed. After making field observations and analyzing type and herbarium specimens, we realized that P. aquatica var. manausensis should be treated as a species due to many morphological and distribution differences compared to the typical variety. We provide detailed descriptions and a distribution map, table, key, and plate to distinguish both species using morphological characteristics. Typifications and nomenclatural notes for related names were performed

According to recent phylogenetic studies, Pachira is not monophyletic (Duarte et al. 2011;Carvalho-Sobrinho et al. 2016) because it emerged in two lineages: the Amazonian Pachira clade, with ca.45 species (of which 11 were confirmed to be in this clade); and the extra-Amazonian Pachira clade, with ca.five species (sensu Carvalho-Sobrinho et al. 2016).The latter clade emerged as segregated from the Amazonian Pachira clade and a sister clade of Eriotheca species.
Both clades of Pachira are strongly supported in an analysis by Carvalho-Sobrinho et al. (2016), who mentioned some morphological traits as possible synapomorphies of the clades.However, these were not confirmed so additional phylogenetic and morphological studies about the genus

Materials and methods
The study involved a literature search and analysis of herbarium specimens at B, BM, CEN, CR, F, G, HAL, HUEFS, HUMC, IBGE, INPA, K, LINN, M, MBM, MEX, MG, NY, P, PAMG, R, RB, S, SP, TEPB, U, UFMG, US, and UFMT (acronyms according to Thiers 2020, continuously updated) to write the descriptions of the species and obtain geographic distribution data.We also observed individuals cultivated in the arboretum at the Rio de Janeiro Botanical Garden (JBRJ) to analyze the species in vivo and obtain floral color data.The common names were obtained in the specimens labels.
The map was made with the software ArcGIS for Desktop v. 10. 1 (ESRI 2013), data from labels of the analyzed material, and a shape file of the ecoregions proposed by Olson et al. (2001).
The typifications and nomenclatural notes follow the recommendations of the International Code of Nomenclature for Algae, Fungi, and Plants (ICN) (Turland et al. 2018) and interpretations of McNeill (2014).The conservation status of the species was determined according to the categories and criteria of the IUCN (2019), which provides guidelines for the protection of priority species.To obtain the area of occupancy (AAO) and extent of occurrence (EOO) values, the GeoCAT tool was used (Bachman et al. 2011).
Distribution and Habitat -Pachira aquatica occurs in Central and South America.It inhabits flooded areas in Belize, Brazil, Costa Rica, Mexico, French Guyana, Guyana, and Trinidad and Tobago.In Brazil, P. aquatica is more frequently found in Amazon forest (Amapá, Amazonas, Maranhão, Mato Grosso, Pará, and Tocantins states) but small populations also occur in Cerrado vegetation (the Distrito Federal and the states of Bahia and Piauí) (Fig. 2).This species is used as an ornamental in several countries because of its small size and attractive flowers.Brasilica, 2022, 36: e2021abb0117 Common names -Castanheiro-do-Maranhão, mamorana, munguba, mandrana.

Acta Botanica
Phenology -Flowers seen from February-June and August-December.Fruits seen from February-April and in August and October.
Conservation Status -Least Concern (LC).This species is widely distributed in tropical rain forests and dry forests.In Brazil, P. aquatica occurs in three biomes, Amazon forest, Caatinga and Cerrado.There are some gaps between the Mexican and Brazilian populations (Fig. 2).The low number of specimens from humid forests is because it is difficult to work in this vegetation.It is possible that with a larger sampling of herbaria, the number of known populations would increase.However, even based on the current sampling data, the EOO surpasses 20,000 km 2 and the AOO surpasses 5,000 km 2 .In addition, two populations are in protected areas (Área de Proteção Ambiental Ilha do Combu, Reserva Florestal do Savacan).
Comments -Pachira aquatica is morphologically similar to P. manausensis.To distinguish these species, see the notes under P. manausensis.
In his Flore des Antilles, Tussac (1827) described P. grandiflora and differentiated it from P. aquatica by the apical position of the flower and absence of nectaries on the receptacle.However, both characteristics used by Tussac are variable and a single specimen can exhibit this variation.The material collected by Tussac is mostly unknown.The few known specimens are at the herbaria P and G (Stafleu & Cowan 1983).After searching several herbaria (including P and G), we found no specimens related to the protologue.Therefore, in accordance with Art.9.3 of the ICN, we select an illustration in the protologue as the lectotype of this name.Thus, we here select plate 3 as the lectotype.The choice of plate 3 (flowering branch) over plate 4 (fruiting branch) is because floral characteristics are more important in the circumscription of the name, as noted by Tussac.
Schlechtendal & Chamisso (1831) used the collections of Schiede and Deppe to describe Carolinea macrocarpa, which were from Mexico and archived at the B and HAL herbaria (Stafleu & Cowan 1983).The collection at B was destroyed and the only remaining information about it is Macbride's photo of the collection (negative no.F9537) (see Grimé & Plowman 1986).We found duplicates of the collection at BM and HAL.Therefore, the specimen HAL0128028 is here designated as the lectotype, since it is the most complete duplicate.
Distribution and Habitat -This species is endemic to Amazonas State (Fig. 2), in Brazil, and occurs in flooded areas and terra firme and capoeira vegetation.
Common names -Munguba, mungubarana.Phenology -Flowers seen in January, February, April and from October-December.The supplemental material (Video S1 in supplementary material) shows the anthesis of the flowers, which occurs at 7 to 7:30 pm.Fruits seen in February, April, May, July, August and December.
Conservation B2ab (i,ii,iii).The species has an AOO smaller than 500 km 2 and a distribution restricted to Amazonas State in the municipalities of Barcelos, Janauarí, Manaus and São Gabriel da Cachoeira.To date, no populations are known to occur in protected areas and only five populations are known in nature.In addition, the arboretum at the Rio de Janeiro Botanical Garden (RBv756) has the only known cultivated individual (see comments in the Discussion).
However, we observed that the morphological boundaries between the two species are consistent throughout their populations and propose the new combination and status of Pachira aquatica var.manausensis to Pachira manausensis.
In the protologue of P. aquatica var.manausensis, Robyns indicates that the holotype is at RB and there are isotypes at four other herbaria (i.e., G, K, S and U).Currently, in RB there are four specimens of the collection indicated by Robyns.Thus, according to Art. 8.1 of ICN, we here select a single specimen as the lectotype.We chose the most complete specimen, with a leaf and flower.Duplicates were also found in other herbaria (i.e., MG, NY and SP), in addition to what Robyns cited, which are listed above as isolectotypes.The isotypes indicated at K by Robyns were not found.

Discussion
For P. manausensis, we highlighted in the IUCN assessment that there are no known protected populations and that there is a single cultivated individual protected ex situ (RBv756) at the Rio de Janeiro Botanical Garden (JBRJ).
At the turn of the 19th century, Adolpho Ducke was the main provider of Amazonian specimens to the in vivo and dry collections at the JBRJ and RB herbarium (Ducke 1930b;Egler 1963;Bediaga & Drummond 2007), as well as one of the most important Brazilian naturalists and researchers of the Amazonian flora (Egler 1963;Miranda 1999;Florez & Scheiner 2016).He obtained the material RB21002 (the type material, barcode RB00537260) in Igarapé de Cachoeira Grande (municipality of Manaus, Amazonas State), which was cited as Bombax rigidifolium Ducke in his list of Amazonian plants (Ducke 1930a).
In 1963, Robyns synonymized B. rigidifolium under P. aquatica.In addition, he published P. aquatica var.manausensis using Ducke's collection as the type and referred to the original name of this material as an "Auct.non" (misapplied name, see recommendation 50D1, ICN).
For more information, see the synonym list under P. manausensis.
Therefore, we hypothesized that Ducke probably collected the specimen RBv756 from the same locality or the same population of the type of P. manausensis, when he went on an expedition in Amazon forest in Brazil, and then planted it at JBRJ.To confirm this hypothesis, we consult the list of in vivo specimens from JBRJ with the information that Ducke brought the individual RBv756 from the state of Amazonas and identified it as B. rigidifolium.However, a description about the municipality was lacking so we could not confirm if the specimen is from Manaus.Nevertheless, for some species we know that Ducke used acclimatization to successfully cultivate plants at JBRJ (Ducke 1930c).For this, he kept a good portion of his collections alive in the nurseries at the Emílio Goeldi Museum (Pará) until they were ready for transportation and planting at JBRJ (Ducke 1930b;Florez & Scheiner 2016).This would explain the time span of approximately 10 years between collecting RB21002 (7 December 1927) and planting the individual RBv756 (31 August 1937) in Rio de Janeiro.There is also the possibility that Ducke collected and tried to acclimate this species more than once.
Finally, since the publication by Robyns (1963), no other work has completely revised Pachira.The varieties of P. aquatica proposed by Robyns can be easily interpreted as species due to the differences that we observed.Possibly, Robyns did not observe both taxa in situ, lacked information about the color of structures, or did not have access to enough collections, since many collections we studied were

Table 1 .
Morphological differences between Pachira aquatica and Pachira manausensis.madeafter 1963.Therefore, everything indicates that, with more information, Robyns would have described P. aquatica var.manausensis at the species level.