Adiantum (Pteridaceae) in Brazil: Key to the species and illustrations

Abstract: Adiantum is a Pantropical genus of ferns, monophyletic, and has about 225 species. It can be recognized by the indusia with veins, bearing sporangia directly on the underside of its reflexed tissue. About 110 species occur in the Neotropical region and 65 of them are reported to Brazil. Among them, 64 are native species and occur mainly in the Amazon Rainforest, Caatinga, Central Brazilian Savanna, and in the Brazilian Atlantic Rainforest. Among the species that occur in Brazil, 16 are endemic and they represent 25% of the total. This paper presents a key to identify the 63 native species of Adiantum in Brazil, excluding the hybrid one. Illustrations of some morphological features, as well as of all species are also presented to help in the process of identification.


Introduction
Adiantum L. is monophyletic, with ca. 225 species, and is the second biggest genus of the family Pteridaceae (PPG I 2016), behind only Pteris L. with 250 spp. Adiantum has a Pantropical distribution, occurring mainly in forests as terrestrial plants. About 110 species are Neotropical.
Recently, Adiantum has been a target of several investigations, including also molecular data to access the evolution of the group and its relationship, and some Brazilian species were sampled for these works. For example, Hirai et al. (2016) and Hirai & Prado (2019) investigated the group of A. raddianum C. Presl and concluded that the group of A. poiretii Wikstr. is the most closely related and they can be separated morphologically by the shape of the indusium (reniform in A. raddianum group and oblong in the A. poiretii group). This important result is also strongly supported by the molecular data indicating that the species of A. poiretii group differ from the A. raddianum group by a unique deletion of 66 nucleotides, at positions 288-353 in the chlN gene alignment. Huiet  Simple leaves (very rare in Adiantum) occur in three distinct clades (davidii, philippense, and peruvianum). Most taxa have leaves that are more than once-pinnate and only a few of these (in the formosum and pedatum clades) exhibit the distinct pseudopedate form. The data about distribution for each studied species show that most species (75%) are restricted to only one of six major biogeographical regions in the World. In this study (Huiet et al. 2018), 48 of the sampled species (ca. 30% of the total) are endemic to South America. This study showed that the previous infra-generic classification for Adiantum failed to compose natural group of species based only on morphological aspect because there are several homoplastic characters.
According to the Flora of Brazil online 2020 (Prado & Hirai, in preparation), there are 65 species occurring in Brazil, 64 of them are native, 16 species are endemic (25% of the total), including one that has a hybrid origin (A. xmoranii J. Prado). Adiantum peruvianum Klotzsch is largely known as cultivated.
The genus itself can be easily identified by the indusia with veins, bearing sporangia directly on the underside of its reflexed tissue (Prado et al. 2017). But the identification of its species is a challenge, especially to the non-expert on ferns.
The main objective of the current work is to present a key for the identification of the recognized species in Brazil (except for the hybrid species), accompanied by illustrations for all taxa, which can be a useful tool for experts and students of ferns.

Material and Methods
All species included in the present key have been recognized in the treatment of the genus Adiantum for Flora of Brazil online 2020 (Prado & Hirai, in preparation) (http://floradobrasil.jbrj.gov.br/reflora/ listaBrasil/PrincipalUC/PrincipalUC.do).
Most of this species appeared previously cited for Brazil since the previous papers involving the genus published by Prado  The key is based on the external morphology of the plants and there is no need to use any additional feature of the anatomical part of the organisms for their identification.
To observe the hairs and scales accurately and thus identify species of Adiantum with the key provided below, it is necessary to have a dissecting microscope with at least 30x and a strong light source. In some species, the hairs become modified as glands, such as in A. dawsonii Lellinger & J. Prado or are protected by the revolute indusia and in both cases their visualization is difficult.
Most of the morphological terms used in the key follow Lellinger (2002). And author abbreviations of scientific names follow Pichi Sermolli (1996).
The endemic species of Adiantum to Brazil appear marked with an asterisk in the key.
Some photos to illustrate this work were provided by the following persons: Michel Boudrie (MB, Adiantum adiantoides (J. Sm.) C. Chr., Moran (RM, A. macrophyllum Sw., A. villosum), and the others were taken by the present authors. Additionally, drawings are also presented to illustrate the species.

Results
Adiantum can be recognized by terete, blackish to castaneous stipes, rachises, and costae, and sporangia borne on the false indusium (i.e., not on the laminar surface below it); also characterized by scales borne at rhizome apices and stipe bases; laminae monomorphic (sterile and fertile laminae similar in morphology), pinnate (rarely undivided) to more divided, sometimes forked or pedate; veins free or rarely anastomosing without included free veinlets; linear epidermal idioblasts (false veins) present or not between the true veins; sori formed on the recurved laminar margins (false indusia), on the veins, paraphyses (i.e., hairs among the sporangia) absent (Prado et al. 2017a).
The main characters used here to recognize the species group are the pattern of venation regular anastomosing veins (forming regular areoles without free veinlets included), irregularly anastomosing (not forming regular areoles and also lacking free veinlets included), or veins free; the veins ending into teeth or ending between teeth at sterile margins of the pinnae/pinnules; rhizome very long-creeping (i.e., cord-like), long-creeping (nodose or not nodose), and short-creeping (generally nodose); indument of the rachises, pinnae, and pinnules (glabrous or with hairs or scales, or both); pattern of the frond division (varying from pinnate to 2-5-pinnate), form of the pinnae/pinnules and indusia. Another important aspect to recognize species is the fact that the pinnae/pinnules/segments are articulate or not (i.e., continuous). In the articulate species, the dark color of the stalk stops at the base of the pinnae/pinnules/segments and, when they are continuous, the color is ending into their laminar tissue. There are some species that present fronds forked one to several times and they are easy to be recognized by this feature.
To facilitate the use of the key, illustrations of some morphological features ( Figure 1) and for all species (Figures 2-13) are presented after the key. Some of these illustrations are not from Brazilian specimens but show the morphology of the treated species.

Discussion
As already commented here, 16 species of Adiantum in Brazil are endemic and have restricted area of occurrence compared to the species widely distributed. Among those with very narrow distribution are: Adiantum diphyllum, Figure 2A Adiantum adiantoides, Figure 2C, D, is not restrict to Brazil, but has its occurrence only in the Brazilian Amazon Forest of the states of Pará, Amazonas, and Amapá (Sundue & Prado 2006).
Among the other species recently described and that are not restricted to Brazil are: Adiantum dawsonii, Figure 10A, B and A. cinnamomeum, Figure 12C, D, Lellinger & Prado (2001), A. giganteum, Figure 10H, Prado (2001), A. windischii, Figure 10I, Prado (2005), and A. nodosum, Figure 10J The key and data here presented are only the first steps to approach the diversity of Adiantum in Brazil. Based on our experience with this interesting and difficult group of plants, certainly, there are more species to be described. During our investigations, we observed some specimens that could be hybrids between known species, but describing these gatherings as distinct taxa is too premature. More fieldwork is necessary to locate more individuals and better study their biology.