What is being published on floristics? an overview of floristic studies carried out in the Espinhaço Range

ABSTRACT The campo rupestre in the Espinhaço Range is subject of many floristic studies due to its high rates of species diversity and endemism. These studies are made up of critical, easily accessible data about local richness and biodiversity, which are essential for subsidizing new studies in the fields of conservation, ecology, and systematics. This study describes the results of a survey of floristic studies carried out in the Espinhaço Range and provides a checklist of the plant species listed in them. We found that 294 floristic studies were conducted in the Espinhaço Range, but they were focused only on a few areas. Likewise, we detected sampling gaps in some areas. The Floras studied provide information on 3,191 plants or 6,3% Brazilian flora, of which 2,879 are angiosperms, 247 bryophytes, and 65 lycophytes and ferns. Finally, we observed a lack of standardization regarding the habits reported for angiosperms shrubs and subshrubs which is more problematic for Espinhaço’s flora because of its grasslands or shrublands vegetation. Thereby, we reaffirm the importance and need to encourage new floristic studies in the Espinhaço Range as a means to train new taxonomists and to provide new studies and projects.


Introduction
Campo rupestre, a rocky phytophysiognomy inserted in the Cerrado biome, has 5,344 species of angiosperms distributed in 124 families, 78 species of bryophytes distributed in 17 families, and 77 species of lycophytes and ferns distributed in 16 families (Flora do Brasil 2020).This vegetation harbors high rates of species endemism proportionately to the total number of species (BFG 2015), with ca.1,748 endemic species (Colli- Silva et al. 2019a), of which at least 1,123 are endemic to the campo rupestre of Espinhaço Range (Rapini et al. 2021).This impressive diversity of species represents c. 15% of all Brazilian angiosperms and they are distributed in less than 1% of the Brazilian territory (Silveira et al. 2016;Vasconcelos et al. 2020).The high rates of diversity and endemism in campo rupestre may be the result of of old strains, with low diversification rates of diversification which survived due to climatic stability (OCBIL's theory) or, alternatively, the result of high diversification rates caused by climatic fluctuations in the Pleistocene (Vasconcelos et al. 2020).A third explanation is the "escape to radiate" model, which combines the fragmentation of the campo rupestre along the Espinhaço Range with high extinction rates and adaptive radiation events (Rapini et al. 2021).However, Rapini et al. (2021) pointed out that the diversification and evolution of this flora cannot be explained fully by a single mechanism.
The high rates of species richness and endemism in the Espinhaço Range's campo rupestre generates a unique vegetation that has attracted much interest from botanists and naturalists throughout history, such as von Martius, St. Hilaire, Spix, and Pohl (Lopes et al. 2011).Consequently, several floristic studies and projects have been elaborated to the Espinhaço Range region, including the Flora da Serra do Cipó (Giulietti et al. 1987), Flórula de Mucugê (Harley & Simmons 1986), Flora dos Picos das Almas (Stannard 1995), and Flora de Grão Mogol (Pirani et al. 2003).Furthermore, Floras are essential tools in the professional training of new botanists (Rapini et al. 2008), and other fields of biology, such as systematics and ecology, make extensive use of these studies (Funk 2006;Rapini et al. 2008).
Despite the importance of classical taxonomic studies, i.e., Floras and other botanical works, these have faced difficulty to be published in high-impact journals (Crisci et al. 2020;Zeppelini et al. 2021).Here we provide a detailed compilation of the floristic studies carried out in the Espinhaço Range and aim to demonstrate the importance of these studies and why they still need to be encouraged.Additionally, we provide a checklist of the plant species listed in the floristic studies we have surveyed.Thus, we aim to demonstrate that such studies carry valuable information for understanding vegetation and, therefore, must be continuously encouraged.

Bibliometric research
We searched two databases: Scopus and SciELO.The bibliometric survey was conducted by searching at Scopus' and SciELO's "topic" field (i.e., title, abstract or keyword) with terms relating to both floristic diversity and Espinhaço Range location.The complete set of keywords used in the survey is available in Table S1.We performed searches with and without the term "espinhaço", as not all studies in the Espinhaço Range area have "espinhaço" in the title, abstract or keyword.The search performed without the "espinhaço" term was limited to journals that could publish this kind of work; notwithstanding, the search returned a list of floristic papers from both Minas Gerais and Bahia states, reason why we had to screen these results to keep only studies from Espinhaço Range.The selection of the Espinhaço Range floristic studies was carried out in two stages: i) reading both the title and abstract of each article, and ii) examining every article to confirm the adequacy of the definition adopted.We combined the results from the two databases and deleted duplicated data.
Finally, we also searched the collection of the Boletim de Botânica da Universidade de São Paulo.This journal has a longrecognized relevance for publishing floristic studies, but the journal's content is not currently available in other databases.This search was carried out directly on the journal's website (http://www.revistas.usp.br/bolbot/issue/archive),checking all articles available, including accepted ones published until 2/01/2022.We double-checked all searches.To ensure the accessibility of all data cited here, we conducted our research only in scientific journals, not investigating Ph.D. theses, master's dissertations, and books.

Checklist
After screening the floristic articles, we used the species addressed in them to compile a checklist.The species names follow Flora do Brasil 2020 and to check it, we used the R package "Flora" (Carvalho 2017; R Core Team 2020).Species whose names are not present in Flora do Brasil 2020 were checked manually using the IPNI (2020), Tropicos (2020), andThe Plant List (2013).The taxa identified as affinis (aff.), species incerta (sp.), confer (cf.), and the taxa indicated as new species, were treated according to the original floristic publication.
In addition, for angiosperms we provided a figure detailing the habit of each species, based on the information supplied by the authors in the original floristic study.When authors did not indicate the habit of a species, we treated the case as "missing data".For Monocots, an exception was adopted: when the habit of a species was not indicated, the species was classified as either palm (for all Palmae) or herb (for all other Monocots).When the authors of the floristic study indicate more than one habit for a species, for example "shrub or subshrub" we count all habits in the analysis.Also we treat all lianas as vines, because when the floristic studies were published, there was no standardization.

Map elaboration
The distribution map of floristic studies by geological region in the Espinhaço Range was made in QGIS version 3.12.1 (QGIS Development Team 2018) by superimposing the coordinates of the floristic studies on the shapefiles of Espinhaço Range.

Survey of floristic studies
The search performed in SciELO returned 59 articles using "espinhaço" as a keyword and 350 in Scopus databases.The search performed without the "espinhaço" keyword returned 575 articles from SciELO and 2124 from Scopus.
The screening step kept 294 floristic studies carried out in the Espinhaço Range (Table S2).In terms of number of publications, the three main journals were Boletim de Botânica (253 articles), Rodriguésia (17), and Hoehnea (10).Publication of floristic studies was not limited to scientific journals (ex.Harley & Simmons 1986), but these studies were more accessible.
The distribution of articles published over time is shown in Figure 1.The first publication on the topic dates from 1987; since then, there was constant production until the late 20th century.Publications returned to a significant amount in 2003, when the first volume of Flora de Grão-Mogol was published (Pirani et al. 2003).After 2003 the maximum number of studies published in a single year occurred in 2004 (46), and the minimum, in 2018 (4).However, in 2021 there was only a single article, but it possibly reflects the Covid-19 pandemic and how its restrictions affect herbarium and field works.
The projects Flora da Serra do Cipó (Giulietti et al. 1987) and Flora de Grão-Mogol (Pirani et al. 2003), both published in Boletim de Botânica, are iconic examples of floristics studies in the Espinhaço Range.These projects have had continuity since then due to the effort and commitment of researchers, graduate, and undergraduate students, to maintain the projects running in the long-term.Other Floras and journals typify this continuity, for example, Flora das Serras de Carajas published in Rodriguésia (Viana et al. 2016).Flora do Brasil is a shining example of a Flora -979 taxonomists participated in the project, which monographed the world›s most diverse flora (Flora do Brasil 2020).We regard Flora do Brasil as a fine example for future floristic studies.Indeed, we predict a key role of this project in encouraging new floristic projects in Brazil, as we understand that the project does not overlap or avoid new floristic studies.
The families with more floristic studies in Espinhaço Range were Asteraceae, Myrtaceae, Orchidaceae, Vochysiaceae, and Xyridaceae, all with four publications each.However, when we looked at infra family taxonomic categories, this scenario changes.In this context, the families with more studies were Asteraceae (11) and Fabaceae (10).Asteraceae's studies include four conducted at the family level (Hind 2003;Moura & Roque 2014;Roque et al. 2016;Staudt et al. 2017) Families such as Apocynaceae, Bromeliaceae, Eriocaulaceae, Malvaceae, and Melastomataceae, had few studies at the family level.These groups of plants are very diverse in the campo rupestre of Espinhaço Range, with high rates of endemism (Colli-Silva et al. 2019a) what could encourage the elaboration of floras.Despite that, these taxa were treated in floristic studies in the Espinhaço Range due to the publication of Floras with other taxonomic categories (Table 1).We believe that this can also be related to the fact that these families show a high diversity in the Espinhaço Range and present a complex taxonomic history, which may make it difficult to complete a monography at once.Some situations than can corroborate this is 1) Apocynaceae studies in Grão Mogol's flora are divided in Apocynaceae without Asclepiadoideae (Oliveira & Pirani 2003) and Apocynaceae only Asclepiadoideae (Rapini et al. 2003); for example 2) Eriocaulaceae's studies in Serra do Cipó were treated by genera or infrageneric categorical Paepalanthus subg.Xeractis (Hensold 1998) and Paepalanthus sect.Actinocephalus (Sano 1998), and Syngonanthus (Parra 1998), the same situation occured at Paepalanthus in the Parque Estadual do Biribiri (Andrino et al. 2015); 3) Malvaceae's studies in Serra do Cipó are also divided as subfamilies (Colli-Silva et al. 2019b;Yoshikawa et al. 2019;Yoshikawa et al. 2020).
Our results showed that floristic studies use taxonomic categories to delimit the plants in a study.Also, its suggests the presence of a specialist in these taxonomic categories in most floristic studies from the Espinhaço Range.Floras promote essential tools for the specialist's professional development (Rapini et al. 2008), and the formation of specialists may promote more flora in the future.That highlights the importance of floras for a continued formation of specialists and therefore its continued necessity.
The knowledge on botanical diversity in the Espinhaço Range keeps growing.For example, between 2005 and 2014, 12 new plant species were described yearly, on average (Fernandes et al. 2018).In the last ten years, 20 species of Eriocaulaceae were described (Echternacht et al. 2011;Echternacht & Sano 2012;Trovó et al. 2013;Echternacht 2014;Echternacht & Trovó 2015;Andrino et al. 2016;Costa et al. 2016;Giulietti et al. 2016;Pereira et al. 2016;De Oliveira Chagas et al. 2019;Andrino et al. 2020;Echternacht et al. 2021), seven of which are endangered, 12 critically endangered, and one is data deficient, according to authors.The endemic and microendemic species of the campo rupestre are extremely vulnerable to extinction (Rapini et al. 2021), a fact that reinforces the need of continuous conservation efforts targeting this flora.It is predicted that areas of campo rupestre suitable for conservation will be lost to an extent of up to 82% over the next 50 years if climate change is not reversed (Fernandes et al. 2018).Other factors could be cited as a threathed for the biodiversity of the campo rupestre, such as urban growth, disorderly mining, forestry industry, lack of governance and animal farming (Fernandes et al. 2020).
Another significant result of our research is the confirmation of the overlap between floristic studies and protected areas.The floristic studies mapped were carried out in or close too national and state parks, or other categories of conservation areas.This result attests the importance of protected areas for floristic research in the Espinhaço Range.In fact, to promote and facilitate research is one of the core objectives of creating a protected area (Brasil 2000).In addition, surveys are an important ally for conservation areas, as they can provide important data to decision makers and also present more information about the biodiversity in protected areas, and their conservation.These data can shed light on the biodiversity that is not being watched, and on how to act for its conservation.Thus, floristic studies are relevant for the conservation of the Espinhaço Range as these studies map the local and regional flora, discover, and describe new species, and help Acta Botanica Brasilica, 2023, 37: e20220267  to identify and monitor priority areas for conservation.The relationship between Parque Nacional da Serra do Cipó and the project Flora da Serra do Cipó (Giulietti et al. 1987) are a shining example for that.

The Espinhaço Range
Espinhaço Range can be recognized in two portions, based on the geopolitical Brazilian states in which they occur: Espinhaço Mineiro and Espinhaço Baiano, this division used to be adopted in botanical studies in the past.(Giulietti et al. 1987;Lohmann & Pirani 1996;Pirani et al. 2003;Rapini et al. 2008;Versieux et al. 2008).However, it is also possible recognize four portion in Espinhaço Range, based on local geology: Espinhaço Meridional, Serra do Cabral, Espinhaço Septentrional, and Chapada Diamantina (Chemale Jr et al. 2011).Nevertheless, we found that some studies carried out in Espinhaço Baiano do not mention Espinhaço Range in the "topic field"; these studies were made in Chapada Diamantina (see Torres et al. 2003;Pataro et al. 2017) and Espinhaço Setentrional (see Azevedo et al. 2015;Ferreira et al 2019;Santos et al. 2020).
In the geopolitical classification, the Iron Quadrangle, a unique Pre-Cambrian structure and one of the richest mineral-bearings regions in the world (Almeida et al. 2011) is commonly viewed as part of the Espinhaço Range (Rapini et al. 2008;Almeida et al. 2014;Silveira et al. 2016).The UNESCO (2005) accepted this delimitation in the proposal of Espinhaço Range as a Biosphere Reserve.The inclusion of the Iron Quadrangle into the Espinhaço Range limits is opportune as it intends to provide greater protection to the area (Gontijo 2008), as well as to other campo rupestre areas (Fernandes et al. 2018), given that the region faces strong mining pressure.
The spatial distribution of floristic studies in the Espinhaço Range is presented in Figure 2. In total, Floras are distributed in 13 points: five in Espinhaço Meridional + Serra do Cabral, three in Espinhaço meridional, and five in Chapada Diamantina.There are four unplaced local in the map, because they are regions and not only a locality: Central Portion of Espinhaço Range (Andrino & Costa 2013), Chapada Diamantina (Torres et al. 2003;Oliveira et al. 2003;Ribeiro et al. 2005;Ferreira et al. 2012;Pimenta et al. 2019), Diamantina Plateau (Shimizu & Yamamoto 2012), and Espinhaço Range (Lohmann & Pirani 1996;Melo 2000;Rapini 2001;Lima & Pirani 2003;Araujo et al. 2005;Oliveira et al. 2010;Rapini 2010).In some cases, the locality indicated in the article's title differ from that wherein the floristic study took place (ex: Staudt et al. 2017); we advise against that.
Our results revealed a concentration of floristic studies in Espinhaço Mineiro -260 works, and only 26 in Espinhaço Baiano.Chapada Diamantina province has 381 endemic species (sensu Colli-Silva et al 2019a), but had only 18 floristic studies.Therefore, it is essential to develop additional floristic studies about this vegetation.The area does not have any study project published in article or monograph formats, such as Flora da Serra do Cipó, in Espinhaço Meridional, and Flora de Grão-Mogol, in Espinhaço Setentrional of Minas Gerais.The two major floristic studies in Chapada Diamantina are Flórula de Mucugê (Harley & Simmons 1986) and Flora dos Picos das Almas (Stannard 1995), both of which are published in book format and did not enter our survey.Additionally, we identified five regions with sampling gaps regarding floristic studies: the areas 1) between Serra do Cipó and Biribiri State Park; 2) between Grão-Mogol and Caetité municipalities; 3) between Serras de Licínio de Almeida and the municipality of Mucugê, 4) west of Chapada Diamantina, and 5) north of Chapada Diamantina.Some of these regions had already been identified as areas of low sampling effort for Asteraceae (Campos et al. 2019).These areas need news studies and collecting efforts because these gaps could reflect a lack of knowledge about the geographical distribution of species -Wallacean Shortfall and it can help to promote the discovery of new species-Linnean Shortfall (Hortal et al. 2015) and floristic studies are useful to face these shortfall.Table 1 shows the relationships between sampling points and the number of floristics studies.

Checklist
Table S3 presents the checklist of plant names compiled from the floristic studies of the Espinhaço Range.We assembled 3,191 plant species belonging to 214 families and 949 genera; of these, 2,879 are angiosperms, 247 bryophytes, and 75 lycophytes and ferns.We highlight that all bryophytes are from Flora da Serra do Cipó (Yano & Peralta 2011) and Flora do Grão-Mogol (Yano & Peralta 2009), whilst all Lycophytes and Ferns are from one of three localities: Serra do Cipó, Grão-Mogol, or Chapada Diamantina (Windisch & Prado 1990;Prado 1992;Windisch 1992;Prado & Windisch 1996;Prado 1997;Prado & Labiak 2003;Ferreira et al. 2012).The Flora do Brasil (2020) indicates 49,992 plant species in Brazil; therefore, at least 6,3% of the Brazilian flora was sampled in floristic studies carried out in the Espinhaço Range.This fact highlights the importance of floristic studies in high biodiversity areas.
These species are just a part of the total richness of Espinhaço Range, as the Serra do Cipó checklist has 3,229 Acta Botanica Brasilica, 2023, 37: e20220267 species (Pirani et al. 2015) and the checklist for Catolés, in the Chapada Diamantina, has 1,710 species (Zappi et al. 2003).Both the Serra do Cipó and Chapada Diamantina areas are within the Espinhaço Range, therefore, the species richness in these mountains is higher than that reported herein (i.e., 3,191 species).Consequently, additional floristic studies in this region are much needed to improve data for all plant species of Espinhaço Range and suppress the gaps reported in "Espinhaço Range" topic.
Furthermore, these genera cluster the highest number of species to their respective families in Espinhaço Range, as is the case of Mikania (Campos et al. 2019).Therefore, these taxa should receive special attention for conservation actions in this area.4B: Habits "Shrub" and "Subshrub" pooled in a single class, "Shrub/Subshrub".

Grasses and shrubs
Acta Botanica Brasilica, 2023, 37: e20220267 using the information provided in the studies.Figure 4B presents the shrub and subshrub habits combined into a single category.We found that, in some instances, there was no clear-cutting separation supporting these habits.We believe this happens because there is no single definition that can standardize how taxonomists use these terms.Another situation detected by us is that some authors did not attribute habit to Monocot species (e.g., Longhi-Wagner & Todeschini 2004;Coffani-Nunes et al. 2010;Sano et al. 2010).We believe that the differential secondary growth in Monocots, compared to Eudicots, explain why these authors avoided defining these plants as herbs, shrubs, and subshrubs.For example, some species of Actinocephalus (Eriocaulaceae) have secondary growth (e.g., Actinocephalus rigidus (Bong.)Sano), a growth pattern that occurs due to activity of a secondary thickening meristem; this feature is likely to occur in other species of Actinocephalus as well (Oriani et al. 2008).This species does not have its habit attributed in floristic studies, though (see Sano et al. 2010).
In these cases, we recommend using the definitions of shrub and subshrub present in Beentje (2010)."Shrub: selfsupporting woody plant branching at or near the ground or with several stems from the base.Subshrub: Small shrub with partially herbaceous stems".The definition for herbs in this study, i.e., "plant without a persistent woody stem above ground", is similar to that of Harris & Harris (2001).However, these definitions still do not suit the cases of Monocots that lack cambia but show another secondary growth.Given this scenario, we advocate that future studies should review terminology and propose clearer terms to define the habits of these plants.Sperotto et al. (2020), for example, promoted a similar discussion about liana terminology and suggested a standardization.
Figure 4 highlights the impressive diversity of shrubs/ subshrubs and herbs in the Espinhaço Range.This is probably because most floristic studies were carried out in campo rupestre vegetation.Mucina (2018) discussed whether this vegetation should be classified as shrub or herbaceous, and the differences between the classification of this vegetation throughout history.He concluded that the most appropriate definition would be a "mosaic of grasslands and patches of scrub, with occasional scattered low trees".Furthermore, the campo rupestre vegetation changes along the Espinhaço's spatial landscape: in Chapada Diamantina the vegetation is predominantly shrub, whereas in the South the vegetation becomes predominantly herbaceous (Colli-Silva et al. 2019a).This also suggests that most of the plants in our checklist are from the campo rupestre.

Conclusion
This study highlights the published floristic studies in the Espinhaço Range.Although numerous floristic studies were found, they are concentrated in few locations.It is necessary to direct new sampling efforts to neglected areas, including some regions in the Northern Espinhaço and northwestern of the Chapada Diamantina.The Iron Quadrangle, a region with many floristic studies, is considered part of the Espinhaço Range by many authors.Floristic studies should be seen as fundamental to know and conserve the flora of Espinhaço Range and its campo rupestre.This floristic study of Espinhaço Range comprised a total of 3,191 plant species, or 6.3% of the Brazilian flora, highlighting the remarkable diversity of campo rupestre.Finally, this study also reveals a standardization problem regarding the habits of angiosperms and indicates the issues that can be revisited in future studies to bring clearer definitions about plant habits.Heywood (2001) discussed the relevance and prospects of floristic studies in this century.Here, we intend to highlight some of Heywood's relevant thoughts and unanswered questions that remain current today: 1) The Floras will become bibliographic records of plant diversity that no longer exist.2) Is it possible to attract young scientists to taxonomy?3) Is it possible to persuade decision-makers to subsidize floristic studies in support of biological and conservation studies?
After more than 20 years these thoughts can be easily applied to the reality of the Espinhaço Range, in which: 1) Deforestation, mining, and uncontrolled fire can alter the flora, so that the current flora may represent a record of what once existed; 2) How can young scientists be encouraged to work with taxonomy, if this type of study is hardly accepted in impactful journals and, at the same time, the current research funding system overvalues high-impact publications to grant scholarships and financial resources?; 3) How can we convince decision-makers to support floristic studies if the academy itself has been negligent with this type of research?
Florist studies are still necessary for Espinhaço Range and other regions in Brazil.To do so, there must exist continuous effort on the part of researchers, graduate, and undergraduate students, and it is our job to encourage advisors and boost funding agencies to support these studies.

Supplementary Material
The following online material is available for this article: Table S1 -Keywords used for the bibliometric survey in Scopus and SciELO.Table S2 -Publications selected from Boletim de Botânica and the Scopus and SciELO databases.Table S3 -Checklist compiled from the data present in floristic studies in the Espinhaço Range.*Species known only from typus.It was collected in this region but was not considered in the study.

Figure 1 .
Figure 1.Temporal distribution of floristic studies carried out in the Espinhaço Range, according to the SciELO and Scopus databases; Boletim de Botânica; Scopus, SciELO and Boletim de Botânica.

Figure
Figure 4A presents a histogram of the habits of the species listed in floristic studies in the Espinhaço Range,

Table 1 .
List of floristic studies in the Espinhaço Range and the number of studies carried out in each locality according to the articles.*Onlyfor Espinhaço Mineiro.