Alpine Botany

https://doi.org/10.1007/s00035-023-00298-1

ORIGINAL ARTICLE

Implementing spatial analyses to measure angiosperm biodiversity

from the high‑altitude grasslands of the Atlantic forest
Igor M. Kessous1,2 · Leandro Freitas1

Received: 9 May 2023 / Accepted: 23 June 2023

© Swiss Botanical Society 2023

Abstract
Campos de altitude are high-altitude grasslands found in the mountain ranges of southeastern and southern Brazil, which are
characterized by high species richness and endemism. Because of the difficulty in delimiting this vegetation type, measur-
ing biodiversity patterns is challenging. Here, we compared the application of two methods using spatial data to estimate
angiosperm diversity in campos de altitude: (1) filtering occurrence data by elevation, canopy height, location and keywords
and; (2) the same as the first, however, adding a filter of “campos de altitude” in the vegetation type of the Flora e Funga do
Brasil database. Also, we discuss conservation status, plant collections, endemism, vegetation data and similarity among
14 sites harboring campos de altitude. Our two resulting lists indicated between 1087 and 2398 angiosperm species and
infraspecific taxa in campos de altitude, mostly belonging to Asteraceae and Poaceae and endemic to Brazil. Extrapolations
of species richness suggest a potential number of up to 4000 species. Of the taxa assessed for conservation status, 53–65%
are threatened or near threatened. The flora of campos de altitude is more similar on closely located mountains rather than
on mountains with similar geological characteristics and origin. We provide an editable list online destined to seek help
from taxonomists to generate a more accurate species list, to support advances in knowledge on this unique tropical montane
ecosystem.

Keywords Beta-diversity · Biogeography · Conservation · Grassland · Mountain · Similarity analyses

Introduction Campos de altitude (Fig. 1) are montane ecosystems con-

sisting of unique high-altitude grasslands that occur in the
Montane floras have a high degree of endemism and diver- highest and coldest regions of Atlantic Forest in southeastern
sity (Chen et al. 2009; Merckx et al. 2015). As on islands, and southern Brazil (Eiten 1983; Veloso et al. 1991; Safford
the isolation of flora on mountaintops leads to similar pat- 1999a, b). While being surrounded by a dense forest matrix
terns of dispersal, speciation and extinction (Merckx et al. and under the influence of similar climatic conditions,
2015; Rahbek et al. 2019). Depending on the context and on campos de altitude differ by having shallow, acidic, and
the organism, mountains can function as bridges or barriers extremely nutrient-poor soils (Safford 1999a, b), favoring
for species flow (Perrigo et al. 2020). Covering around 25% their grasslands condition. Typically, these grasslands occur
of the planet’s terrestrial area, some estimates suggest that at or above ca.1800–2000 m a.s.l., especially in Caparaó,
mountain regions harbor more than 85% of the terrestrial Serra dos Órgãos, and Itatiaia national parks (Safford 1999a,
species; this pattern is especially observed in the Neotropics, b). However, the proximity to the ocean and latitude strongly
such in the Andes and the mountains of the Atlantic Forest influence this elevational threshold; in some cases, this
(Rahbek et al. 2019). formation starts at ca.1500 m a.s.l. (Vasconcelos 2011) or
even at ca.1300 m in southernmost Brazil (Paraná and Santa
* Igor M. Kessous Catarina states) (Martinelli and Orleans e Bragança 1996;
[email protected] Struminski 1996; Mocochinski and Scheer 2008; ICMBIO
2021). Elevation and soil type are known to influence the
1
Jardim Botânico do Rio de Janeiro, Rua Pacheco Leão, 915, phylogenetic relationships and distribution of species (Neri
Rio de Janeiro 22460‑030, Brazil
et al. 2017; Gastauer et al. 2020). Thus, delimiting campos
2
Gothenburg Global Biodiversity Centre, Carl Skottsbergs de altitude based solely on elevation is complex, since these
Gata 22B, 41319 Gothenburg, Sweden

13
Vol.:(0123456789)
 Alpine Botany

Fig. 1  Campos de altitude and their flora. a Landscape view of in Parque Nacional da Serra dos Órgãos; e Barbacenia irwiniana
Parque Estadual dos Três Picos (PETP); b Grassland in Parque L.B.S.m. in Parque Nacional da Serra dos Órgãos. Photos: Igor M.
Nacional do Itatiaia (PNIT); c Vriesea itatiaiae Wawra in Parque Kessous
Nacional do Itatiaia; d Zygopetalum maculatum (Kunth) Garay

grasslands usually form mosaics with the adjacent cloud for- in the Atlantic Forest usually on igneous or metamorphic
ests, with different micro-climates and biotas. rocks, in the mountaintops of the Serra do Mar and Serra
In Brazil, campos de altitude are not the only type of da Mantiqueira (Safford 1999a). Also, the Serra Geral, at
natural mountaintop grasslands, a common misconception in the southernmost part of the Serra do Mar and extending
the literature and databases. Campos de altitude occur only into the interior of Brazil and Paraguay (Frank et al. 2009),

13
Alpine Botany

harbors grasslands that are recognized as campos de altitude of the campos de altitude, and estimate the extrapolated spe-
(ICMBIO 2021). Despite originating from the same tectonic cies richness, under the hypotheses that (1) angiosperm taxa
event in the Paleocene (Almeida and Carneiro 1998), the occurring in campos de altitude are underestimated in previ-
Serra do Mar and the Serra da Mantiqueira exhibit signifi- ous inventories; (2) campos de altitude of the Serra da Man-
cant structural, geographical, and biological differences, tiqueira and Serra do Mar are internally more similar than
making them distinct “vegetation islands”. The Serra da to each other; and (3) the application of “filtering” methods
Bocaina is located between these two mountain ranges and to spatial data and habitat information can be used to esti-
is also considered part of the Serra do Mar. Recent climatic mate the biodiversity in campos de altitude. Based on the
events have favored the emergence of a unique biota in the information collected, we estimated the number of taxa per
region (Behling et al. 2007). family, threatened taxa, life forms, habitat, and endemism,
Other open montane vegetation types occur in Brazil, and provide explanations for species distribution patterns.
mainly in the Cerrado (campos rupestres); and in the Ama-
zon (tepuis). In contrast with campos de altitude, these vege-
tation types occur on sandstone or quarzitic rocks (Alves and Methods
Kolbek 2010; Silveira et al. 2020). Many plant collections
from these other grasslands include the term “campo de alti- To construct lists of angiosperm taxa (specific and infraspe-
tude” on their labels; however, these plant communities are cific) occurring in campos de altitude and surroundings, we
biogeographically different. The flora of the campos rupes- downloaded from Global Biodiversity Information Facility
tres is more closely related to that of the tepuis, while cam- (GBIF) 9,101,129 records (GBIF.org -09 February 2022-
pos de altitude are more related to the Andes (Safford 1999b, GBIF Occurrence Download https://​doi.​org/​10.​15468/​dl.​
2007; Vasconcelos 2011; Guedes et al. 2020). Because of e9c9cf) of all preserved specimens of Magnoliopsida (Eud-
their similarities to Andean Páramos, campos de altitude icots + Basal angiosperms) and Liliopsida (Monocots) from
came to be known as Brazilian Páramos, although with a all of Brazil. We excluded ambiguities and all records not
greater influence of seasonality (Safford 1999a, 2007). Nev- identified at least to species level.
ertheless, the term campos de altitude was already widely
used in the literature and in the national biogeographic clas- Backbone list
sification and is therefore more commonly used today to
refer to this vegetation. To obtain a backbone list (BBL), we divided the search into
The flora of the campos de altitude includes ancient plant two steps: first, we obtained a list of taxa with georefer-
lineages (Silveira et al. 2020) and has high species richness enced data (decimalLatitude and decimalLongitude) further
and endemism (Safford 1999a; Neri et al. 2017). Advances in filtered by elevation and canopy height; second, we obtained
taxonomy and associated technology over the past 20 years a list of taxa filtered by keywords to also encompass the
have dramatically increased the number of known species in occurrences without geographical coordinates.
Brazil. Previous work focusing on the diversity of campos de In the first step, we removed duplicates and cleaned
altitude suggested that they contain fewer than 1000 vascular biased records using the R (R Core Team 2020) package
plant species (Safford 2007), encompassing 100 endemics CoordinateCleaner (Zizka et al. 2019), flagging and remov-
(Alves and Kolbek 2010). However, in these studies, the ing “capitals”, “centroids”, “equal”, “gbif”, “institutions”,
scope of the concept of campos de altitude was not always “zeros” and “countries”. We then downloaded two raster
clear, and most of them included only those high-altitude files containing elevation information (Earth Resources
grasslands typical of the northern Serra do Mar and Serra Observation and Science—EROS (2018), with 30 m reso-
da Mantiqueira. With the recent expansion of open-source lution:  https://w ​ ww.u​ sgs.g​ ov/c​ enter​ s/e​ ros/s​ cienc​ e/u​ sgs-e​ ros-​
data repositories, the use of spatial data has enabled us to archi​ve-​digit​al-​eleva​tion-​shutt​le-​radar-​topog​raphy-​missi​on-​
reach a better understanding of species distribution patterns, srtm?q​ t-s​ cienc​ e_c​ enter_o​ bject​ s=0#q​ t-s​ cienc​ e_c​ enter_o​ bjec​
habitats, and conservation. ts ) and canopy height, with 1 km resolution: and extracted
Here, we compared two methods using spatial data to the information for each point using the R package raster
estimate the diversity of angiosperms in campos de altitude. 3.5–29 (Hijmans et al. 2022). There are no official maps
We provide two lists of species and infraspecific taxa, the specifically containing the sites of campos de altitude (Vas-
first based on sensu lato searching (filtering occurrence data concelos 2016), thus we specified elevation filters accord-
by elevation, canopy height, location and keywords) and the ing to the literature for each site where they occur (Table 1,
other based on sensu stricto searching (the same as the first, Fig. 2) or when lacking information, using Google Earth®.
however adding a filter of “campos de altitude” in the veg- In addition, we applied a < 5 m filter for canopy height, since
etation type of the Flora e Funga do Brasil). Furthermore, campos de altitude are covered predominantly by grasses,
we aimed to provide an editable list of angiosperm species shrubs, and forbs, with only a few scattered small trees.

13
 Alpine Botany

Table 1  Protected areas that include the campos de altitude used in the present study and their minimum elevation threshold
Mountain range Protected area (acronym) Min. elevation References Reference coordinates

Serra ­Gerala Parque Nacional de São Joaquim 1300 m ICMBIO (2021) – 28.513, – 49.528
(PNSJ)
Serra do Mar (S) Parque Estadual Pico Paraná (PEPP) 1300 m Mocochinski and Scheer (2008) – 25.242, – 48.855
Serra do Mar (S) Serra de Araçatuba (ARAC) 1300 m Mocochinski and Scheer (2008), Silva – 25.943, – 48.996
et al. (2010)
Serra do Mar (C) Parque Nacional da Serra da Bocaina 1500 m Freitas (2002) – 22.815, – 44.657
(PNSB)
Serra do Mar (N) Área de Proteção Ambiental de 1500 m IBAMA (2007) – 22.411, – 43.216
Petrópolis (APAP)
Serra do Mar (N) Parque Estadual do Desengano (PEDS) 1500 m Google Earth – 21.893, – 41.906
Serra do Mar (N) Parque Nacional da Serra dos Órgãos 1800 m Cronemberger and Viveiros-de-Castro – 22.496, – 43.087
(PNSO) (2007)
Serra do Mar (N) Parque Estadual dos Três Picos (PETP) 1800 m Safford (1999a)b, Google Earth – 22.400, – 42.799
Serra da Mantiqueira Parque Estadual de Campos do Jordão 1800 m Safford (1999a), Fundação Florestal – 22.713, – 45.444
(PECJ) (2015)
Serra da Mantiqueira APA Mantiqueira: Parque Estadual da 1800 m Safford (1999a)b, Google Earth – 22.059, – 44.681
Serra do Papagaio (PESP)
Serra da Mantiqueira APA Mantiqueira: Pedra da Mina 1800 m Safford (1999a)b, Google Earth – 22.428, – 44.843
(PDMI)
Serra da Mantiqueira APA Mantiqueira: Pico dos Marins 1800 m Safford (1999a)b, Google Earth – 22.592, – 45.121
(PIMA)
Serra da Mantiqueira Parque Nacional do Itatiaia (PNIT) 2000 m Safford (1999a), Alves et al. (2016) – 22.372, – 44.633
Serra da Mantiqueira Parque Nacional do Caparaó (PNCP) 2000 m Safford (1999a), Rodrigues et al. – 20.457, – 41.787
(2009)
Serra da Mantiqueira Parque Estadual da Serra do Brigadeiro 1500 m Safford (1999a)b, Google Earth – 20.665, – 42.417
(PESB)c

Minimum elevation (approximate values) was used as a filter along with canopy height < 5 m in the analyses. We considered only the sites that
contained at least 50 taxa in SLL. APA Mantiqueira was divided into three different groups, due to the geographic delimitation of each of the
mountains
APA Mantiqueira Área de Proteção Ambiental da Serra da Mantiqueira, C Central, N Northern, S Southern
a
Boundary of Serra Geral and Serra do Mar
b
No specific information, estimated by latitude according to Safford (1999a) and manually by Google Earth
c
Outgroup site

We then downloaded the map of the Brazilian protected Final lists

areas (MMA, http://m ​ apas.m
​ ma.g​ ov.b​ r/i​ 3geo/d​ atado​ wnloa​ d.​
htm ), rasterized using the functions shapefile and rasterize To evaluate incorrect occurrences and obtain data on veg-
of the R package raster (Hijmans et al. 2022), and extracted etation type, conservation status, endemism, life form, fam-
the information from each coordinate point to determine the ily, and habitat, we merged our BBL data with previously
distribution of each taxon in the different campos de altitude. published data (Moreira et al. 2020; Carrijo et al. 2018),
We excluded the records corresponding to other Brazilian and unpublished lists obtained by our research group (in
mountaintop vegetation types, such as the grasslands of cen- PNIT and PNCP), and extracted the information from Flora
tral and northern Brazil. e Funga do Brasil (http://fl ​ orad​ obras​ il.j​ brj.g​ ov.b​ r/r​ eflor​ a/l​ ista​
For the keyword search, we used the raw list downloaded Brasil/). For this, we performed the function get.taxa of the
from GBIF, including records with and without coordi- R package flora 0.3.5 (Carvalho 2021). The Flora e Funga do
nate data, and performed the search in the locality column: Brasil database is based on information from about 1000 tax-
“campo de altitude” and “campos de altitude”, ignoring onomists who collaborated in the description of the Brazil-
capitals. Again, we filtered the search output according to ian Flora for more than a decade (Brazil Flora Group 2021).
the elevation obtained by the GBIF database, keeping only For the synonymized taxa not recognized in this R pack-
the records according to Table 1, or those that had no eleva- age, we performed searches for the accepted name of each
tion information. We removed all records not corresponding separately, using the platforms World Flora Online (http://​
to campos de altitude and duplicates. www.w ​ orldfl
​ orao​ nline.o​ rg/) and Tropicos (https://w ​ ww.t​ ropi​

13
Alpine Botany

50°W 40°W

20°S

PNCP

PESB

PEDS

PNSO PETP
PESP

PNIT
PECJ
PIMA PNSB APAP

PEPP PDMI

Elevation (m)
ARAC
3000

2000
PNSJ
30°S
1000

Fig. 2  Map with the campos de altitude analyzed here. Map color references the elevation raster. Acronyms are explained in Table 1

cos.​org/), and resubmitted them for analysis in the package (function wilcox.test) performed in R to verify differences
flora. We excluded all taxa considered to not occur in Brazil. in the number of taxa. Here, we considered only campos
The final output, covering all taxa and localities, is referred de altitude located at or above at least 1300 m and with
to here as the sensu lato list (SLL), representing species more than 50 taxa in SLL (except by the outgroup), after a
and infraspecific taxa that occur in campos de altitude and detailed literature search for each specific case. We did not
possibly in surrounding vegetation types. Because of meth- include sites that are doubtful as to their vegetation or “high
odological issues of resolution of the rasters and in order to altitude rupestrian complexes” (see Benites 2002; Garcia
include border species, we considered the use of “campos 2003; Garcia and Pirani 2005; Vasconcelos 2011). An edit-
de altitude and surroundings” to refer to SLL. Preliminary able SSL-based angiosperm taxa list can be found in https://​
sensu lato lists were made for all localities (Table 1) sepa- doi.​org/​10.​5281/​zenodo.​81143​53.
rately and then merged.
To obtain our sensu stricto list (SSL), we applied a “cam- Cluster, network, dissimilarity analyses,
pos de altitude” filter to the vegetation type column in our conservation status, and collections
preliminary sensu lato lists, representing only those taxa
that occur in campos de altitude according to the Flora e We fully coded the presence and absence of each taxon
Funga do Brasil. Thus, we were able to compare the results by site and performed clustering analyses for both data-
and analyses between the two datasets. The resulting lists sets, using the R package BinMat (van Steenderen 2022).
were compared by the non-parametric Mann–Whitney U test For the hierarchical cluster analyses, we chose the Jaccard

13
 Alpine Botany

distance method, applying the clustering method of Results

WPGMA (Weighted Pair Group Method with Arithmetic
Mean), since we considered that there is an unequal num- Sensu lato list
ber of taxa in the sites/clusters. We performed the func-
tion upgma and applied 10,000 bootstrap replications to Our sensu lato list (SLL) resulted in 2398 species and
calculate the p-value of each cluster. Reliability was meas- infraspecific taxa, 651 genera, and 131 families (Online
ured by approximately unbiased (AU) probability values Appendix 1). The estimated number of angiosperm taxa
and bootstrap (BP). We then performed the dissimilarity in each campos de altitude (and surroundings) is given in
analyses using the Jaccard method, and obtained the num- Table 2. Asteraceae is the best represented family, encom-
ber of taxa for each site, using the functions vegdist and passing 345 taxa, followed by Poaceae (191 taxa) and Orchi-
specnumber, respectively, of the R package vegan 2.6.2 daceae (162 taxa) (Fig. 3a). Most of the taxa are herbs (997
(Oksanen et al. 2022). taxa; shrubs: 311; trees: 184; Fig. 3a), terrestrial (1792 taxa;
We performed a network analysis using the eLasso epiphytes: 108; rupicolous: 61; Fig. 3a), and endemic to Bra-
(least absolute shrinkage and selection operator) method zil (1407 taxa; non-endemic: 926, Fig. 3a). A total of 224
(van Borkulo et al. 2014) for SLL and SSL, using the R taxa of angiosperms are threatened (CR: 44; VU: 45; EN:
package IsingFit 0.3.1 (van Borkulo et al. 2016). This pro- 135; Fig. 3a) and 47 taxa are near threatened. In general,
cedure allowed us to use binary data, using the Extended our results indicated a larger number of taxa in sites with
Bayesian Information Criterion (EBIC; Chen and Chen more collections and higher mean numbers of collections
2008), which helps to decrease false-positive rates (van per taxon (Table 2). Of the four most diverse sites, PNSB
Borkulo et al. 2014). We applied the AND-rule and chose had the fewest collections per taxon (2.4), while PNIT
the higher gamma value (1.0) to avoid unreliable connec- and PNSJ had the most collections per taxon (4.0 and 5.3,
tions, especially false positives (van Borkulo et al. 2014, respectively). PEDS, PESP, PECJ, APAP, and PIMA had
2016). To statistically compare the output networks result- fewer than two collections per taxon.
ing from SLL and SSL, we submitted our binary datasets
to a direct comparison test (network comparison test),
using the NCT function of the R package NetworkCom- Sensu stricto list
parisonTest 2.2.1 (van Borkulo et al. 2019), with the same
parameters as for the network construction. The sensu stricto list (SSL) list resulted in 1087 species
Conservation threat status (CR: Critically Endangered; and infraspecific taxa, 347 genera, and 89 families (Online
VU: Vulnerable; and EN: Endangered) of the taxa were Appendix 2). This list showed similar taxonomic patterns
obtained from the latest Official National List of Endan- to the SLL. The Asteraceae was also the best represented
gered Plant Species (MMA 2022). Information for near family, with 225 taxa, followed by Poaceae (112 taxa) and
threatened (NT) and least concern (LC) taxa was obtained Fabaceae (58 taxa) (Fig. 3b). Also, most of the taxa are herbs
by use of the package flora 0.3.5 (Carvalho 2021). Life (539 taxa; shrubs: 131 taxa; subshrubs:108 taxa; Fig. 3b),
form, habitat, endemism, and the corresponding fami- terrestrial (809 taxa; rupicolous: 37 taxa; epiphyte: 15 taxa;
lies (according to the Flora e Funga do Brasil), in addi- Fig. 3b), and endemic to Brazil (686 taxa; non-endemic:
tion to the conservation status, were plotted using the 384 taxa, Fig. 3b). According to the SSL, 133 taxa of angio-
R package ggplot2 (Wickham 2016). We included only sperms (12%) are threatened (CR: 33 taxa; VU: 19 taxa; EN:
accurate and unambiguous information about habitat and 81 taxa; Fig. 3b), and 20 taxa are near threatened according
life form. After obtaining the final lists, we estimated the to the IUCN criteria. Of the total number of angiosperm
mean number of collections per taxon (Table 2). We plot- taxa, about 223 (20% in the SSL) are endemic to the cam-
ted the graphical networks using the package qgraph 1.9.3 pos de altitude vegetation. The number of taxa did not pre-
(Epskamp et al. 2022). sent significant differences between the two lists (W = 157,
We estimated species-accumulation curves (SACs) p = 0.07).
through the method of “rarefaction” and the extrapolated
species richness for the GBIF species dataset (based on the
SLL vouchers), as well as for each site separately using the R Species accumulation curves and extrapolation
package vegan 2.6.2 (Oksanen et al. 2022). For the extrapo-
lated species richness analyses, we used multiple estimators, The analysis of SACs and extrapolated species richness dem-
as suggested by Brose et al. (2003): Chao, Jacknife (first onstrate a continued increase in the number of species within
and second order, Jack1 and Jack2 respectively) and Boot- campos de altitude, suggesting that the number of species
strap (Boot). All codes and list of vouchers are available on can reach up to 4052 species. (Fig. 4; Table 2). Across the
GitHub (https://​github.​com/​kesso​us/​campos_​de_​altit​ude). entire dataset, the observed richness accounted for mean

13
 Alpine Botany

Table 2  Estimated number of taxa, collections, and extrapolated species richness (based on the SLL vouchers) from each site
Site SLL taxa SSL taxa Total collec- Mean collections GBIF species Chaoc Chao (se)c Jack1c Jack1 (se)c Jack2c Bootc Boot (se)c nc C (%)
number number tion ­numbera per taxon (SLL) ­numberb

PNIT 986 495 3975 4 761 1598.0 120.7 1170.4 74.6 1477.8 929.4 39.3 114 76.2
PNSJ 680 380 3641 5.3 706 1117.3 58.2 1060.2 78.1 1261.0 861.7 46.8 62 63.2
PNCP 595 395 2269 3.8 496 917.2 75.4 742.0 67.9 914.4 600.2 38.9 62 75.0
PNSB 535 231 1306 2.4 566 1110.5 78.3 930.0 190.2 1170.0 719.6 111.5 53 54.5
PNSO 277 160 854 3.1 289 535.2 51.0 462.5 36.5 574.1 362.4 21.1 70 57.3
PEDS 221 101 635 1.8 230 431.5 47.8 361.4 55.5 447.6 286.1 33.3 21 57.9
PDMI 221 123 550 2.6 230 600.2 86.2 387.7 61.9 505.2 294.9 29.7 15 49.4
PESP 209 121 396 1.9 227 562.6 75.6 387.9 63.3 504.9 293.4 32.5 17 47.8
PECJ 202 100 353 1.7 209 476.4 62.1 356.8 39.2 461.8 270.0 20.6 36 51.6
ARAC​ 186 88 621 3.3 190 359.7 43.5 303.1 32.4 377.2 238.1 18.3 30 58.2
PEPP 137 68 474 3.4 143 283.2 40.0 234.5 48.3 294.8 181.7 27.6 27 55.1
APAP 125 65 225 1.8 128 314.7 57.2 216.4 35.7 281.3 164.2 18.4 20 51.2
PIMA 95 64 152 1.7 98 278.0 62.0 170.9 41.4 225.4 127.8 22.2 13 47.4
PETP 50 32 110 2.2 51 90.2 18.2 83.3 11.9 102.0 65.0 6.8 20 58.7
PESB d 41 18 68 1.6 44 213.9 102.1 79.8 10.8 110.0 58.1 4.7 17 35.5
All 2398 1087 15,629 6.5 2374 3694.6 110.0 3422.0 161.6 4052.5 2832.2 99.7 132 68.5

SLL Sensu lato list, SSL Sensu stricto list, Jack1 Jacknife first order, Jack2 Jacknife second order, Boot Bootstrap, se standard error, n number of transects, C approximate comparison of the
average values obtained from extrapolation methods with the SLL values. Sites abbreviations are in Table 1
a
Collections obtained in BBL, identified at least to species level after filtering (elevation + canopy height or keyword search) and cleaned
b
The number of species was determined prior to incorporating species from Moreira et al. (2020) and Carrijo et al. (2018), and before applying name corrections using the R package flora 2.6.2
(Oksanen et al. 2022), as well as manual corrections using World Flora Online, Tropicos, and Flora e Funga do Brasil
c
Analyses made using the GBIF species number
d
Outgroup site

13
 Alpine Botany

2
Terrestrial: 179
T

Terrestrial: 809
1500

1000

: 37
08
750

1
s: 6
yte: 1
a b

olous
Cl

15
Su

500 500

8
lou
im

:1

:7
b

te :
Su

1
Epiph
sh

be

:2

ite

ite
8
ico
bs

Rupic
Cl

y
250
ru

c:
Tr
r:

as

Tre : 43
tic

as
h
He hr

im
Sh

Rup
e
b:

He uati
Epip
rb:
15

ar
ub

ar
e:

a
ru

be

e:
997 07 Sh
18

u
ip

ip
0 0
b: 18 : 14 :1
2
He ru

Aq

Aq
19
m

m
emic
2

3 4 rb: b: 08
11
1 1

He
End 539 1 31 : 68
6
5 5 emic
926 End
ic:
−endem : 384
Non demic
Non−en
2 NA: 65
2 NA: 17
LC: 242 EN: 81
EN: 1
35 4 LC: 7
8
4
47 : 33
NT: 5 CR 20
:4 : A
VU : 44 Po Aste NT 19 Po ste
CR 3 ac rac VU
:
3 ac
ea
rac
ea

Fa chid
ea ea e: e:
O last

Or
e:

ba
e: 22
rc

Cy
11
Me

34 5

: 25

Mela
Fab

25

ce
19

Lamiace
Ericaceae: 31
9

hi

5 2
Cyp
ae: 78
Myrtaceae: 78
Rubiace
e: 6

pe
1

ae
da

iaceae:
ceae
ace

:5
ac
ce

rac
stom
erac

om
acea

8
ae

ea
ae:

eae
Solanace

:1
at

e
Irida

:5
ae: 33
eae :

62
ac

atac
ae: 84
meli

:4
110

Bromel

5
ea

8
e:

eae:
Bro

105

15
3

34
Fig. 3  Number of taxa by: 1- life form; 2- endemism; 3- family; 4- LC least concern, NT near threatened, EN endangered, CR critically
conservation status; 5- habit. a sensu lato list (SLL); b sensu stricto endangered, VU vulnerable. Values of each line are explained on the
list (SSL). Data obtained from the flora package, according to the side of the pink bar
Flora e Funga do Brasil (2022) and MMA 2022. IUCN categories:

a b PNIT
+
+
+
PNSJ
++
++
+
+ PNSB
++++ PNCP
+ +++
+
++
+++
+
++ PEDS
++++
+++++++ PNSO
+
++

2000 ARAC
+ +

+++
+++++ + APAP
+

600
+ + ++
++++
++ PESP
++
+
+ +
PECJ
++ +++
++ PDMI
PIMA
PEPP
Number of species

+
+++ PETP
1500 +
PESB

+
+++
400

1000

++
+

+++
++++++ 200
500
+ ++
+
+ +++
++

+
+ +
+
+
+
+
+

0 0
0 20 40 60 80 100 120 0 20 40 60 80 100
Years Years

Fig. 4  Species accumulation curves (SACs) per years. a SAC of the the SAC estimated by the “random” method; b SAC of each site
entire GBIF species dataset. The blue line represents the SAC esti- separately using the GBIF species dataset. Confidence index is repre-
mated by the “rarefaction” method. The boxplots curve represents sented by the shadows. Acronyms are explained in Table 1

13
Alpine Botany

68.5% (58–83%) of the extrapolated richness (Chao: 64%, in SSL; 2– Serra da Mantiqueira and Serra da Bocaina;
Jack1: 69%, Jack2: 58%; Boot: 83%). The completeness and 3– Southern Serra do Mar + Serra Geral. Importantly,
means varied across the different sites. PNIT and PNCP the Southern and Northern Serra do Mar did not cluster
had the highest means (76.2% and 75%, respectively), while together in any of the analyses; however, the Northern
PESB and PIMA exhibited the lowest means (35.5% and Serra do Mar emerged together with the Serra da Man-
47.4%, respectively). Similar patterns were observed in the tiqueira in SSL, and partially in SLL, excluding PETP.
separate datasets, with the minimum percentage typically Internally to the Southern Serra do Mar + Serra Geral,
around 45–50%, often associated with Jack2, and the highest the localities in Paraná state (PEPP and ARAC) clustered
percentage occurring in Boot, ranging from 75 to 85% (refer with strong support in both analyses. In the Northern
to Table 2 for precise values). Serra do Mar communities, PNSO and PEDS clustered
together in both analyses. In the Serra da Mantiqueira,
Hierarchical relationship and dissimilarity we observed a strongly supported cluster of PNSB, PNIT,
and PNCP, with PNCP and PNIT clustering together. The
The hierarchical cluster dendrograms based on the SSL relationship among PDMI, PESP, PECJ, and PESP was
and SLL data showed very similar topologies and branch inconclusive, with PECJ + (PESP + PDMI) in SLL and
support (Fig. 5; acronyms explained in Table 1). Three PESP + (PECJ + PDMI) in SSL. The dissimilarity analyses
main clusters were formed: 1– Northern Serra do Mar (Table 3) showed congruent results to the cluster analyses,
(PNSO, PEDS, APAP), with PETP nested together only although with some differences. However, analyzing the

a PDMI b
40 79
56 94
67 PESP PECJ
38
91 74
Serra da Mantiqueira +

PECJ PESP
Serra da Bocaina

43 40
75 PNCP 73
100 100
100 100
100 PNIT
33 100 45
67 100 100 75
PNSB
17
78

PIMA
21
66

PEDS
85 56
97 73
Serra do Mar

20 77 PNSO
Northern

80 87
95 98
93
APAP 49 98
85

90 PETP
95

ARAC
+ Serra Geral
Serra do Mar

100 100
Southern

100 100
39 PEPP 27
60 67

PNSJ
BP
AU
PESB

1.00 0.95 0.90 0.85 0.80 0.75

0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.0

Height
Height

Fig. 5  Hierarchical clusters dendrogram generated by the R pack- the approximately unbiased probability values (AU). Colors refer to
age BinMat. a data from the sensu lato list (SLL); b data from the mountain ranges: reddish-brown-Serra da Mantiqueira; green-Serra
sensu stricto list (SSL). The number above the branches refers to the do Mar; blue- Serra Geral. Acronyms are explained in Table 1
bootstrap value (BP) and the number below the branches refers to

13
 Alpine Botany

Table 3  Jaccard dissimilarity analyses in both datasets

PNIT PNSO PNSJ PNCP PECJ PNSB PEPP ARAC​ PETP APAP PESB* PEDS PIMA PESP PDMI

PNIT – 0.830 0.827 0.656 0.842 0.753 0.917 0.920 0.956 0.945 0.986 0.898 0.891 0.844 0.805
PNSO 0.879 – 0.911 0.834 0.889 0.863 0.871 0.878 0.871 0.840 0.989 0.797 0.886 0.898 0.859
PNSJ 0.871 0.935 – 0.850 0.947 0.869 0.931 0.904 0.960 0.977 0.992 0.954 0.970 0.950 0.916
PNCP 0.760 0.862 0.872 – 0.870 0.789 0.926 0.924 0.943 0.910 0.968 0.862 0.907 0.866 0.857
PECJ 0.880 0.899 0.939 0.890 – 0.843 0.872 0.944 0.935 0.956 0.974 0.908 0.837 0.837 0.795
PNSB 0.804 0.883 0.899 0.824 0.861 – 0.905 0.922 0.965 0.928 0.988 0.915 0.919 0.850 0.869
PEPP 0.947 0.908 0.933 0.930 0.896 0.928 – 0.762 0.913 0.961 0.988 0.895 0.935 0.926 0.883
ARAC​ 0.942 0.913 0.912 0.927 0.937 0.925 0.767 – 0.947 0.945 1.000 0.908 0.959 0.944 0.940
PETP 0.968 0.899 0.973 0.950 0.946 0.970 0.931 0.960 – 0.934 0.980 0.892 0.945 0.938 0.931
APAP 0.963 0.883 0.985 0.933 0.972 0.942 0.969 0.974 0.952 – 0.936 0.806 0.934 0.937 0.944
PESB a 0.989 0.974 0.993 0.969 0.975 0.981 0.994 0.996 0.989 0.956 – 0.947 0.975 0.978 0.978
PEDS 0.926 0.853 0.962 0.888 0.926 0.925 0.922 0.923 0.925 0.887 0.956 – 0.900 0.933 0.902
PIMA 0.925 0.906 0.971 0.915 0.871 0.934 0.941 0.955 0.942 0.947 0.970 0.910 – 0.886 0.873
PESP 0.883 0.900 0.953 0.886 0.858 0.883 0.935 0.949 0.956 0.956 0.980 0.943 0.899 – 0.854
PDMI 0.860 0.888 0.934 0.881 0.857 0.903 0.909 0.948 0.954 0.967 0.977 0.922 0.899 0.856 –

Lower matrix refers to SLL and upper matrix refers to SSL. In bold, the lowest dissimilarity between each pair (non-repeating pairs). Acronyms
are explained in Table 1
a
Outgroup site

lowest dissimilarity between each site, we observed sev- et al. 2007; Mocochinski and Scheer 2008; Pompeu et al.
eral of the same pairs in both datasets (Table 3). 2014; Campos-Cordeiro and Neri 2019; Campos et al.
2020). Considering the sites analyzed here, we estimate that
Network analyses the number of angiosperms in campos de altitude is between
our SLL (2398) and SSL (1087). The former number is
The network analyses showed similar results regarding posi- higher since it probably contains species of other vegetation
tive and negative connections in the datasets, although with types. On the other hand, the latter may have discarded some
more connections (p < 0.05) in SLL than in SSL (Fig. 6). records, particularly of grassland plants that occur bordering
The analysis based on the SLL dataset (Fig. 6a) showed a forests. Further advances in developing reliable species lists
large number of positive connections. PNSJ (Serra Geral) of the campos de altitude will depend on collection efforts
showed only negative connections, with PNIT, APAP, and and record validation by specialists on different families.
PEDS (p < 0.05). SSL network also showed a negative con- Therefore, an online and editable list is available at https://​
nection between PNSJ and PNIT, in addition to positive con- doi.​org/​10.​5281/​zenodo.​81143​53.
nections in three groups: Southern Serra do Mar, Northern Previous studies suggested that fewer than 1100 plant
Serra do Mar, and Serra da Mantiqueira + PNSB, with the taxa (Fundação Florestal 2015), including fewer than 1000
latter two positively connected. The Network Comparison vascular plants (Safford 2007) occur in campos de altitude.
Test showed mostly non-significant differences in compar- We demonstrate that it is not possible to estimate accurately
ing the network structures of the SLL and SSL datasets edge the number of angiosperms in campos de altitude solely
weight (per edge: Online Appendix 3; maximum difference through the Flora e Funga do Brasil information on habi-
in edge weights: p = 1) and global strength (p = 0.77). tat occurrence. Our analyses confirmed that number of taxa
were underestimated in previous inventories. According to
Flora e Funga do Brasil database (searched in August 2022),
Discussion an estimated 2353 species (and 179 infraspecific taxa) of
angiosperms are associated to campos de altitude. However,
Output lists this number includes certain taxa that do not occur in cam-
pos de altitude but in other Brazilian ecosystems, such as the
Few studies have aimed to estimate the number of plant Amazon (tepuis), Cerrado (campos rupestres), and Pampas
species in the campos de altitude, and most of them have (campos sulinos). Although some authors have used the term
focused on different collection methods, elevation gradients, campos de altitude for Amazonian montane grasslands (e.g.,
or specific regions or groups (Pereira et al. 2006; Ribeiro Martinelli 2007), we here considered the more recognized

13
Alpine Botany

APAP
PEDS
a PNSJ

PNSO
PETP

PNIT
PIMA

PNCP
PEPP
ARAC PNCP
PDMI PECJ b
PNIT PNSJ
PDMI
PESP
PNSB
PIMA
PECJ
PEPP
PESP

ARAC
PNSB

APAP

PETP
PNSO
PEDS

Fig. 6  Graphical networks among campos de altitude communities parent where p > 0.05 and less transparent toward 0. Only connections
generated by the R packages qgraph and IsingFit. Each line weight with p < 0.05 are shown. a data from the sensu lato list (SLL), b data
of the edges indicates positive (green) and negative (red) connection from the sensu stricto list (SSL). Acronyms are explained in Table 1
strength. Alpha transparency refers to the p-values, which are trans-

use in literature for campos de altitude, it means, only those SLL including more tree and epiphyte families than SSL, as
from the Atlantic Forest. Applying an “Atlantic Forest” filter, expected considering the two methods. On the other hand,
this number reduces to 2170 species (and 163 infraspecific Asteraceae and Poaceae were consistently the most diverse
taxa), although still including more than 1000 taxa that occur families in the campos de altitude (Fig. 3), as observed in
in Rio Grande do Sul state, outside the usual defined geo- previous studies of this vegetation (Safford 1999a; Fundação
graphic range of the campos de altitude (after Martinelli and Florestal 2015; Alves et al. 2016; Flora e Funga do Brasil)
Orleans e Bragança 1996; Safford 1999a), and not included and of the alpine flora as a whole in the Americas (Figueroa
in the present study. et al. 2022).
High-altitude environments around the world are occu-
Family and functional diversity pied by herb- and shrub-rich communities, usually without
tall trees (White et al. 2000; Alves and Kolbek 2010; Blair
Grasslands are regions of utmost importance for human et al. 2014; Dixon et al. 2014; Le Stradic et al. 2015; Mucina
survival, representing more than a third of the Earth’s land 2018; Rada et al. 2019). Here, we found that the number of
cover (Buisson et al. 2022). Despite the term “grasslands”, herbaceous plants is dramatically higher (ca. 3–4 times) than
not only grasses, but especially forbs and shrubs are pre- the number of shrubs, the second most common life form in
sent in these environments (Dixon et al. 2014). Brazil ranks both datasets, which can be explained by the difficult envi-
seventh among countries in total grassland area (ca. 17% ronmental conditions for a tropical vegetation. The campos
of its entire land area; White et al. 2000). Here, we found de altitude environment is subject to wide variations in daily
slightly divergent results between the two datasets, with and seasonal temperature, solar radiation, and precipitation

13
 Alpine Botany

(Safford 1999a, b). In the Andean Páramos, grasses and fires during the dry season. Conversely, natural fires during
forbs are estimated to be more resistant to water deficiency the wet season are more limited and may even be beneficial
and low-temperature injury than woody plants (Rada et al. for some species (Safford 1999a, 2001; Aximoff and Rodri-
2019), which could explain the predominance of these life gues 2011). These results demonstrate the urgency of acting
forms in campos de altitude. to protect the campos de altitude, through specific national
The difference in frequency of life forms between our plans for this vegetation and a greater effort to assess the
two datasets (i.e., the larger number of tree-like life forms in risks to this environment and its species.
SLL) reflects that SLL captured cloud forest species, while
SSL was more closely restricted to species of campos de Similarity among campos de altitude communities
altitude. Orchidaceae and Bromeliaceae are represented in
campos de altitude (Safford 1999a); however, these families Similarity analyses focused on different South American
showed a considerable increase in SLL, with a high number mountain ranges suggest the proximity of the campos de
of epiphytic species. We also observed an increase in the altitude flora to floras in other high-elevation regions in
number of Myrtaceae and Rubiaceae in SLL, which was South America, especially the Andean region (Safford
clearly induced by the surrounding cloud forests that harbors 2007), although the internal relationship between them still
many species of these families (Pompeu et al. 2014). lacks a clear resolution. Our results refuted our expectation
of a higher similarity among floras of the same geological
Conservation, endemism and biogeography formations.
In contrast to our expectation, Southern Serra do Mar
About 59–63% of the angiosperm taxa occurring in campos sites are more similar to the Serra Geral (PNSJ) than to
de altitude are endemic to Brazil, which is similar to the those in the Northern Serra do Mar (APAP, PNSO, PEDS,
proportion for the Atlantic Forest as a whole (ca. 63%; Flora and PETP). One explanation for the pattern observed is
e Funga do Brasil). Also, our analysis showed that ca. 20% the greater distance of the Southern Serra do Mar from the
of the taxa (in SSL) occur only in campos de altitude, in Serra da Mantiqueira + Serra da Bocaina (PEPP to PECJ: ca.
general agreement with previous estimates based on more 400 km) compared to the Northern Serra do Mar (PNSO to
restricted sampling (ca. 21%; Safford 1999a). While their PNIT: ca.150 km), facilitating the flow of species between
environmental conditions have led to high endemism, cam- them. Proximity also seemed to influence the clustering, and
pos de altitude are hypothesized to be gateways for plant networking Northern Serra do Mar.
groups into the Brazilian Shield, especially regarding the The Serra da Bocaina, clustered with the sites of the Serra
Andes-Atlantic Forest pathway, as in the case of certain Bro- da Mantiqueira, also explained by their greater proximity
meliaceae (Givnish et al. 2011) and liverworts (Santos and (40 km from PNIT, although set apart by the lowland Paraiba
Costa 2010), and through the co-occurrence of several other do Sul river valley). On the other hand, proximity does not
groups (Martinelli and Bragança 1996; Safford 1999a). In explain the relationship of the PNIT + PNCP cluster, which
addition to the Andean similarity, campos de altitude share are ca. 350 km apart. One explanation for the massive pres-
taxa with several other disjunct phytogeographic elements ence of shared taxa in PNIT and PNCP probably relates to
(temperate, holarctic, austral-antarctic; see Safford 1999a) the recent Last Glacial Maximum, which led to the vegeta-
and with other South American mountain regions (Safford tion of the campos de altitude extending to lower elevations,
2007; Alves et al. 2007). which reached about 700 m in the Caparaó surroundings,
Montane ecosystems are rich in endemic species, which forming a “corridor”. These regions were subsequently iso-
makes them a priority for conservation (Chen et al. 2009; lated as the Holocene advanced, as supported by palynologi-
Merckx et al. 2015; Assis and Mattos 2016; Rahbek et al. cal and zoological data (Safford 1999a, b; Gonçalves et al.
2019). Most localities harboring campos de altitude are 2007; Meireles and Shepherd 2015).
included in protected areas (see Table 1), and are even Some previous analyses have attempted to investigate
covered by a broader national law (the Brazilian “Lei da the similarity of the campos de altitude to other mountain
Mata Atlântica”). However, over time, campos de altitude ecosystems and surrounding physiognomies (Safford 2007;
are subject to many kinds of human pressure (Martinelli Meireles and Shepherd 2015; Mendonça 2017). Meireles
2007; Ribeiro and Freitas 2010) and are especially vulner- and Shepherd (2015) found that cloud forests at high alti-
able to climate change (Neri et al. 2017). Only ca. 20% of tudes are similar to each other, which seemed to be reflected
the taxa in both lists have been assessed for their conser- in our locality networks, as SLL showed many more connec-
vation (SLL: 513; SSL: 231), of which a high proportion tions than SSL. On the other hand, Mendonça (2017) sug-
presented a threatened or near threatened status (SLL: 53%; gested that the flora of the campos de altitude of the Serra
SSL: 65%). One of the primary factors affecting endemic da Mantiqueira is similar to the floras of the campos rupes-
and endangered campos de altitude species is human-caused tres of south-central Minas Gerais and the campos sulinos

13
Alpine Botany

(here also considered as campos de altitude) of the Serra caution when interpreting the results obtained from SACs
do Mar of Paraná, in contrast to our results. Also, Safford and to consider other factors that may impact estimates of
(2007) suggested that the campos de altitude (sensu Saf- species richness. By employing multiple methodologies,
ford 2007) are floristically similar to the Aparados da Serra we can obtain a more robust estimate of true species rich-
(near PNSJ). Here, PNSJ clustered, with low support, with ness and enhance the reliability of our conclusions.
PEPP + ARAC, and is the only one site located in the Serra Although SACs extrapolate the number of species in
Geral, at the southern limit of the Serra do Mar. our dataset, our statistical analyses were based on a ran-
dom subset of all occurring species. In our study, the sites
Extrapolation of species richness with the largest number of collections were mostly those
and methodological considerations with the most species, being influenced by several aspects
and influencing the analyses performed. PNCP, PNSJ, and
The lack of collections can lead to premature conclusions PNIT have the largest extent of campos de altitude and
in spatial analyses (Farooq et al. 2021). As expected, our have the most collections; however, the sampling bias goes
species accumulation curves (SACs) show an increasing beyond these aspects, especially regarding accessibility.
trend in the number of species in most sites and in the total This factor becomes clear when sites near large cities, uni-
dataset. The areas differ greatly in size and accessibility, versities, and research institutions or with easy access are
which affects sampling sufficiency and collection efforts. better sampled (Meineke and Daru 2021). Some of the
Taxonomic, geographic, and temporal biases (according sites are only accessible by long trails, sometimes requir-
to Meineke and Daru 2021) are often observed in these ing more than a day of walking (e.g., PDMI), while others
datasets. can be reached by car (e.g., PNIT). Our results evidenced
The SACs and extrapolated richness have demonstrated a disproportionate collecting effort among the campos
that the expected number of species can vary depending on de altitude, making it highly necessary for further col-
the specific site. However, it is important to note that our lections to focus on these less collected sites, especially
SAC analyses did not account for the inclusion of imputed PEDS, PESP, PECJ, APAP, and PIMA (mean collections
species from recent reviews of two significant sites (PNIT per taxon < 2). On the other hand, even the most intensely
and PNCP, as detailed in the methods section), which could collected sites did not reach saturation in the number of
have positively influenced the comprehensiveness of our species.
species lists. Methodological bias is always present in statistical
Considering the entire dataset, we found that at least 58% analyses. Here we tried to standardize biases and errors,
of the species were accounted for in the true richness, with since all data were submitted to the same method, without
variations of up to 83% depending on the estimation method further external intervention. However, the collection bias
used. It is crucial to acknowledge that while SACs can pro- is present in spatial data databases such as GBIF (Beck
vide an estimate of expected species numbers, they are not et al. 2014), a frequent problem in analyses of large-scale
always considered reliable sources (Longino et al. 2002; occurrence data. GBIF data come from different collec-
Brose et al. 2003). tions, which have different collection methods and focus,
Moreover, it is noteworthy that the campos de altitude and therefore some groups and sites are better sampled
represent only approximately 350 ­km2 of the Atlantic Rain- than others. Also, this can be problematic when perform-
forest (Safford 1999a), which covers a vast area of about ing SACs, as there may not have clear sampling events. On
200,000 ­km2. At first glance, it might seem unlikely that the other hand, by using automated methods to clean the
this relatively small region encompasses a quarter of the raw matrix, we eliminated some of these biases, although
native species found within this domain. The Flora and it was then necessary to deal with the information loss
Funga do Brasil estimates the Atlantic Rainforest to host (according to Zizka et al. 2020), one of the main issues
approximately 16,000 angiosperm species. Nevertheless, it of filter methods. In traditional checklist methods, some
is plausible that the count of angiosperms in the Atlantic taxonomic groups are better studied than others, and often
Forest is underestimated. Furthermore, due to the notable taxonomic revisions are performed by researchers who are
distinctiveness of the campos de altitude habitat in compari- not specialists in all the families involved, resulting in dis-
son to the forest matrix, it is conceivable that they contribute proportionate effort in the accuracy of species identifica-
a substantial portion, potentially a quarter, of the species tion. Therefore, to reduce the bias of information loss,
within this domain. and assist to reach the total coverage of species, we here
It is important to mention that the shape of the SAC can proposed the creation of an editable list to seek an accurate
be influenced by the spatial distribution of species, which number and the identity of taxa that occur in campos de
can vary depending on the habitat and environmental con- altitude.
ditions (Brose et al. 2003). Hence, it is crucial to exercise

13
 Alpine Botany

Conclusions Declarations
Conflict of interest The authors declare no competing interests.
According to the two methods we tested, between 1087 and
2398 angiosperm taxa are present in the Brazilian campos Consent for publication Both authors have approved the submitted
de altitude, an increase over previous estimates. However, manuscript.
according to our extrapolation analyses, this number could
potentially exceed 4000. Geographical proximity is a bet-
ter predictor of floristic similarity than presence in the
same mountain range. Also, we found that a relatively References
few taxa in campos de altitude have conservation assess-
Almeida FFM, Carneiro CDR (1998) Origem e evolução da Serra do
ments, and of these, ca. 53–65% are threatened or near Mar. Braz J Geol 28:135–150
threatened. In campos de altitude, many collections have Alves RJV, Kolbek J (2010) Can campo rupestre vegetation be flo-
focused on only a few sites, and this should be changed for ristically delimited based on vascular plant genera? Plant Ecol
a more comprehensive knowledge of this unique tropical 207:67–79. https://​doi.​org/​10.​1007/​s11258-​009-​9654-8
Alves RJV, Cardin L, Kropf MS (2007) Angiosperm disjunction
ecosystem. New collections in under-sampled sites should “Campos rupestres –restingas”: a re-evaluation. Acta Bot Bras
be carried out, as only then will we be able to determine 21:675–685. https://​doi.​org/​10.​1590/​S0102-​33062​00700​03000​14
the true diversity of these sites. Alves RG, Zaú AS, Oliveira RD (2016) Flora dos campos de altitude
Overall, our SLL and SSL showed similar patterns, em quatro áreas do Maciço do Itatiaia, nos estados do Rio de
Janeiro e Minas Gerais, Brasil. Pesquisas Botânica 69:109–140
indicating that our “filtering” method based on location, Assis MV, de Mattos EA (2016) Vulnerabilidade da vegetação de cam-
elevation and canopy height may be suitable for estimating pos de altitude às mudanças climáticas. Oecologia Aust 20:24–36.
biodiversity in campos de altitude and in other mountain https://​doi.​org/​10.​4257/​oeco.​2016.​2002.​03
environments. The estimated numbers and analyses are Aximoff I, Rodrigues RDC (2011) Histórico dos incêndios florestais
no Parque Nacional do Itatiaia. Cienc Florest 21:83–92. https://​
mostly proportionally similar between the two datasets. doi.​org/​10.​5902/​19805​09827​50
The differences between the SLL and SSL networks and Beck J, Böller M, Erhardt A, Schwanghart W (2014) Spatial bias in
number of taxa were statistically not significant. the GBIF database and its effect on modeling species’ geographic
Herewith, we make available an online and editable list to distributions. Ecol Inform 19:10–15. https://d​ oi.o​ rg/1​ 0.1​ 016/j.e​ co-
inf.​2013.​11.​002
seek contributions from taxonomists to eventually estimate a Behling H, Dupont L, Safford HD, Wefer G (2007) Late Quaternary
precise number of angiosperm species, to serve as a baseline vegetation and climate dynamics in the Serra da Bocaina, south-
for phytogeographic and conservation analyses. Finally, we eastern Brazil. Quat Int 161:22–31. https://​doi.​org/​10.​1016/j.​
suggest that further conservation and biogeographic studies quaint.​2006.​10.​021
Benites VM (2002) Caracterização dos solos e das substâncias húmicas
explore the real state-of-the-art of campos de altitude spe- em complexos rupestres de altitude. PhD Thesis, Universidade
cies conservation status and drivers of their great diversity. Federal de Viçosa
Blair J, Nippert J, Briggs J (2014) Grassland ecology. In: Monson
Supplementary Information The online version contains supplemen- RK (ed) Ecology and the environment. Springer, New York, pp
tary material available at https://​doi.​org/​10.​5281/​zenodo.​81143​53 and 389–423
https://​doi.​org/​10.​1007/​s00035-​023-​00298-1. Brazil Flora Group (2021) Brazilian Flora 2020: Leveraging the power
of a collaborative scientific network. Taxon 71:178–198. https://​
Acknowledgements We thank the members of the Floral Biology doi.​org/​10.​1002/​tax.​12640
Lab—JBRJ for their valuable discussions on the project and Janet Reid Brose U, Martinez ND, Williams RJ (2003) Estimating species rich-
for the English revision. We gratefully acknowledge the Fundação Car- ness: sensitivity to sample coverage and insensitivity to spatial
los Chagas Filho de Amparo à Pesquisa do Rio de Janeiro—FAPERJ patterns. Ecology 84:2364–2377. https://d​ oi.o​ rg/1​ 0.1​ 890/0​ 2-0​ 558
(204.326/2021; E-26/204.327/2021) for the postdoctoral fellowship Buisson E, Archibald S, Fidelis A, Suding KN (2022) Ancient grass-
and research grant to IMK; Conselho Nacional de Desenvolvimento lands guide ambitious goals in grassland restoration. Science
Científico e Tecnológico—CNPq (PQ:304794/2018-0) and FAPERJ 377:594–598. https://​doi.​org/​10.​1126/​scien​ce.​abo46​05
(CNE: E-26/201.000/2022) for the research grant to LF. Campos PV, Villa PM, Schaefer CEGR, Nunes JA, Porembski S, Neri
AV (2020) Community composition, beta diversity and structure
Author contributions Both authors contributed to the design and con- of high altitude grasslands along an altitudinal gradient in south-
ception of the study. IMK analyzed the data and wrote the paper, with eastern Brazil. Rev Biol Trop 68:977–986. https://​doi.​org/​10.​
comments from LF. 15517/​RBT.​V68I3.​37704
Campos-Cordeiro AA, Neri AV (2019) Spatial patterns along an ele-
Funding This work was funded by the Fundação Carlos Chagas Filho vation gradient in high altitude grasslands. Brazil. Nord J Bot
de Amparo à Pesquisa do Rio de Janeiro, FAPERJ (204.326/2021; 37:e02065. https://​doi.​org/​10.​1111/​njb.​02065
E-26/204.327/2021; E-26/201.000/2022). Carrijo TT, Alves-Araújo AG, Amorim AMA, Barbosa DEF, Barcelos
LB, Baumgratz JF, Bueno VR, Coelho RLG, Costa DP, Couto
Data availability The online version contains supplementary mate- DR, Delgado CN, Dutra VF, Flores TB, Furtado SG, Giacomin
rial available along with this article and at https://​github.​com/​kesso​ LL, Goldenberg R, Gomes M, Gonzaga DR, Guimarães EF,
us/​campos_​de_​altit​ude. Heiden G, Kameyama C, Labiak PHE, Lírio EJ, Lohmann LG,
Matos, FB, Moraes PLR, Meireles LD, Menini-Neto L, Monteiro

13
Alpine Botany

D, Moreira MM, Morim MP, Mota MCA, Oliveira, JRPM, Pastore Bromeliaceae: insights from an eight-locus plastid phylogeny. Am
JFB, Pederneiras LC, Pereira LC, Rapini A, Salimena FRG, Silva J Bot 98:872–895. https://​doi.​org/​10.​3732/​ajb.​10000​59
AV, Silva-Neto SJ, Sobral MEG, Souza MC, Sylvestre LS, Trovó Gonçalves PR, Myers P, Vilela JF, de Oliveira JA (2007) Systemat-
M, Viana PL, Forzza RC (2018) Lista de espécies de plantas ter- ics of species of the genus Akodon (Rodentia: Sigmodontinae)
restres do Parque Nacional do Itatiaia. In: Catálogo de Plantas in southeastern Brazil and implications for the biogeography of
das Unidades de Conservação do Brasil. Jardim Botânico do Rio the campos de altitude. Misc publ—Mus Zool, Univ Mich 197
de Janeiro. Disponível em: https://​catal​ogo-​ucs-​brasil.​jbrj.​gov.​br. Guedes TB, Azevedo JA, Bacon CD, Provete DB, Antonelli A (2020)
Accessed 21 June 2022 Diversity, endemism, and evolutionary history of montane biotas
Carvalho G (2021) Package ‘flora’. https://c​ ran.r-p​ rojec​ t.o​ rg/w
​ eb/p​ acka​ outside the Andean region. In: Rull V, Carnaval AC (eds) Neo-
ges/​flora/​flora.​pdf. Accessed 5 May 2022 tropical diversification: patterns and processes. Springer, Cham,
Chen J, Chen Z (2008) Extended bayesian information criteria for pp 299–328
model selection with large model spaces. Biometrika 95:759–771. Hijmans RJ, Van Etten J, Cheng J, Mattiuzzi M, Sumner M, Greenberg
https://​doi.​org/​10.​1093/​biomet/​asn034 JA, Lamigueiro OP, Bevan A, Racine EB, Shortridge A, Hijmans
Chen IC, Shiu HJ, Benedick S, Holloway JD, Chey VK, Barlow HS, MRJ (2022) Package ‘raster’. https://​cran.r-​proje​ct.​org/​web/​packa​
Hill JK, Thomas CD (2009) Elevation increases in moth assem- ges/​raster/​raster.​pdf. Accessed 15 Aug 2022
blages over 42 years on a tropical mountain. Proc Natl Acad Sci IBAMA (2007) Plano de Manejo da APA de Petrópolis. Resumo Exec-
USA 106:1479–1483. https://​doi.​org/​10.​1073/​pnas.​08093​20106 utivo. https://​petro​polis.​r j.​gov.​br/​sma/​phoca​downl​oad/​Docum​
Cronemberger C, Viveiros-de-Castro EB (2007) Ciência e conservação entos/​Prote​cao_​Ambie​ntal/​resumo_​execu​tivo_​apa.​pdf. Accessed
na Serra dos Órgãos. IBAMA, Brasília 12 March 2022
Dixon AP, Faber-Langendoen D, Josse C, Morrison J, Loucks CJ ICMBIO (2021) Plano de uso público do Parque Nacional de São
(2014) Distribution mapping of world grassland types. J Biogeogr Joaquim. https://​www.​gov.​br/​icmbio/​pt-​br/​assun​tos/​biodi​versi​
41:2003–2019. https://​doi.​org/​10.​1111/​jbi.​12381 dade/u​ nidad​ e-d​ e-c​ onser​ vacao/u​ nidad​ es-d​ e-b​ iomas/m
​ ata-a​ tlant​ ica/​
Eiten G (1983) Classificação da vegetação do Brasil. CNPq/Coordena- lista-​de-​ucs/​parna-​de-​sao-​joaqu​im/​arqui​vos/​plano_​uso_​publi​co_​
cao Editorial, Brasília sao_​joaqu​im.​pdf. Accessed 11 March 2022
Epskamp S, Costantini G, Haslbeck J, Isvoranu A, Cramer AO, Le Stradic S, Silveira FA, Buisson E, Cazelles K, Carvalho V, Fer-
Waldorp LJ, Schmittmann VD., Borsboom D (2022). Package nandes GW (2015) Diversity of germination strategies and seed
‘qgraph’. https://​cran.r-​proje​ct.​org/​web/​packa​ges/​qgraph/​qgraph.​ dormancy in herbaceous species of campo rupestre grasslands.
pdf. Accessed 15 May 2022 Austral Ecol 40:537–546. https://​doi.​org/​10.​1111/​aec.​12221
Farooq H, Azevedo JA, Soares A, Antonelli A, Faurby S (2021) Map- Longino JT, Coddington J, Colwell RK (2002) The ant fauna of a tropi-
ping Africa’s biodiversity: more of the same is just not good cal rain forest: estimating species richness three different ways.
enough. Syst Biol 70:623–633. https://​doi.​org/​10.​1093/​sysbio/​ Ecology 83:689–702. https://​doi.​org/​10.​1890/​0012-​9658(2002)​
syaa0​90 083[0689:​TAFOAT]​2.0.​CO;2
Figueroa HF, Marx HE, de Souza Cortez MB, Grady CJ, Engle-Wrye Martinelli G (2007) Mountain biodiversity in Brazil. Braz J Bot
NJ, Beach J, Stewart A, Folk RA, Soltis DE, Soltis PS, Smith SA 30:587–597. https://​doi.​org/​10.​1590/​S0100-​84042​00700​04000​05
(2022) Contrasting patterns of phylogenetic diversity and alpine Martinelli G, Orleans e Bragança J (1996) Campos de Altitude. Index,
specialization across the alpine flora of the American mountain Rio de Janeiro
range system. Alp Bot 132:107–122. https://​doi.​org/​10.​1007/​ Meineke EK, Daru BH (2021) Bias assessments to expand research
s00035-​021-​00261-y harnessing biological collections. Trends Ecol Evol 36:1071–
Flora e Funga do Brasil (2022) Jardim Botânico do Rio de Janeiro. 1082. https://​doi.​org/​10.​1016/j.​tree.​2021.​08.​003
http://​flora​dobra​sil.​jbrj.​gov.​br/. Accessed 24 June 2022 Meireles LD, Shepherd GJ (2015) Structure and floristic similarities of
Frank HT, Gomes ME, Formoso ML (2009) Review of the areal extent upper montane forests in Serra Fina mountain range, southeast-
and the volume of the Serra Geral formation, Paraná Basin, South ern Brazil. Acta Bot Brasilica 29:58–72. https://​doi.​org/​10.​1590/​
America. Pesqui Em Geocienc 36:49–57. https://​doi.​org/​10.​ 0102-​33062​014ab​b3509
22456/​1807-​9806.​17874 Mendonça JGF (2017) Campos de altitude do Parque Estadual da Serra
Freitas L (2002) Biologia da polinização em campos de altitude no do Papagaio, Minas Gerais, Brasil: Composição florística, fitogeo-
Parque Nacional da Serra da Bocaina. PhD Thesis, Universidade grafia e estrutura da vegetação. Master’s Dissertation, Universi-
de Campinas dade Federal de Juiz de Fora
Fundação Florestal (2015) Plano de Manejo do Parque Estadual de Merckx VS, Hendriks KP, Beentjes KK, Mennes CB, Becking LE, Pei-
Campos do Jordão. São Paulo. https://​smast​r16.​blob.​core.​windo​ jnenburg KT, Afendy A, Arumugam N, de Boer H, Biun A, Buang
ws.​net/​funda​caofl​orest​al/​2017/​02/​Diagn%​C3%​B3sti​co-e-​Plane​ MM (2015) Evolution of endemism on a young tropical mountain.
jamen​to.​pdf. Accessed 20 May 2022 Nature 524:347–350. https://​doi.​org/​10.​1038/​natur​e14949
Garcia RJF (2003) Estudo florístico dos campos alto-montanos e matas MMA (2022) Portaria MMA Nº 148, de 7 de Junho de 2022. https://​
nebulares do Parque Estadual da Serra do Mar-Núcleo Curucutu, www.​icmbio.​gov.​br/​cepsul/​images/​stori​es/​legis​lacao/​Porta​r ia/​
São Paulo, SP, Brasil. PhD Thesis, Universidade de São Paulo 2020/P_​mma_​148_​2022_​altera_​anexos_​P_​mma_​443_​444_​445_​
Garcia RJF, Pirani JR (2005) Análise florística, ecológica e fitogeográ- 2014_​atual​iza_​espec​ies_​ameac​adas_​extin​cao.​pdf. Accessed 02
fica do Núcleo Curucutu, Parque Estadual da Serra do Mar (São Aug 2022
Paulo, SP), com ênfase nos campos junto à crista da Serra do Mar. Mocochinski AY, Scheer MB (2008) Campos de altitude na serra do
Hoehnea 32:1–48 mar paranaense: aspectos florísticos. Floresta 38:625–640. https://​
Gastauer M, Thiele J, Porembski S, Neri AV (2020) How do altitude doi.​org/​10.​5380/​rf.​v38i4.​13158
and soil properties influence the taxonomic and phylogenetic Moreira MM, Carrijo TT, Alves-Araújo AG, Rapini A, Salino A,
structure and diversity of Brazilian paramo vegetation? J Mt Sci Firmino AD, Chagas AP, Versiane AF, Amorim AM, da Silva
17:1045–1057. https://​doi.​org/​10.​1007/​s11629-​019-​5403-1 AV, Tuler AC (2020) A list of land plants of Parque Nacional do
Givnish TJ, Barfuss MH, Van Ee B, Riina R, Schulte K, Horres R, Caparaó, Brazil, highlights the presence of sampling gaps within
Gonsiska PA, Jabaily RS, Crayn DM, Smith JAC, Winter K (2011) this protected area. Biodivers Data J 8:e59664. https://​doi.​org/​10.​
Phylogeny, adaptive radiation, and historical biogeography in 3897/​BDJ.8.​e59664

13
 Alpine Botany

Mucina L (2018) Vegetation of Brazilian campos rupestres on siliceous Santos ND, Costa DP (2010) Altitudinal zonation of liverworts in the
substrates and their global analogues. Flora 238:11–23. https://​ Atlantic Forest, Southeastern Brazil. Bryologist 113:631–645.
doi.​org/​10.​1016/j.​flora.​2017.​06.​007 https://​doi.​org/​10.​1639/​0007-​2745-​113.3.​631
Neri AV, Borges GRA, Meira-Neto JAA, Magnago LFS, Trotter IM, Silva RR, Cervi AC, dos Santos ÉP (2010) Flórula do Morro dos Perdi-
Schaefer CEG, Porembski S (2017) Soil and altitude drive diver- dos, Serra de Araçatuba, Paraná, Brasil: Ericaceae. Acta Biol Par
sity and functioning of Brazilian Páramos (campo de altitude). J 39:87–97. https://​doi.​org/​10.​5380/​abpr.​v39i0.​20256
Plant Ecol 10:771–779. https://​doi.​org/​10.​1093/​jpe/​rtw088 Silveira FA, Dayrell RL, Fiorini CF, Negreiros D, Borba EL (2020)
Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara Diversification in ancient and nutrient-poor Neotropical ecosys-
RB, Simpson GL, Solymos P, Stevens MHH, Wagner H, Oksanen tems: how geological and climatic buffering shaped plant diversity
MJ (2022) Package ‘vegan’. https://​cran.r-​proje​ct.​org/​web/​packa​ in some of the world’s neglected hotspots. In: Rull V, Carnaval AC
ges/​vegan/​vegan.​pdf. Accessed 15 Aug 2022 (eds) Neotropical diversification: patterns and processes. Springer,
Pereira IM, Oliveira-Filho ATD, Botelho SA, Carvalho WAC, Fontes Cham, pp 329–368
MAL, Schiavini I, Silva AFD (2006) Composição florística do Struminski E (1996) Parque Estadual do Pico Marumbi, caracterização
compartimento arbóreo de cinco remanescentes florestais do ambiental e delimitação das áreas de risco. Master’s Dissertation,
maciço do Itatiaia, Minas Gerais e Rio de Janeiro. Rodriguésia Universidade Federal do Paraná
57:103–126. https://​doi.​org/​10.​1590/​2175-​78602​00657​108 van Borkulo CD, Borsboom D, Epskamp S, Blanken TF, Boschloo L,
Perrigo A, Hoorn C, Antonelli A (2020) Why mountains matter for Schoevers RA, Waldorp LJ (2014) A new method for construct-
biodiversity. J Biogeogr 47:315–325. https://​doi.​org/​10.​1111/​jbi.​ ing networks from binary data. Sci Rep 4:5918. https://​doi.​org/​
13731 10.​1038/​srep0​5918
Pompeu PV, Fontes MAL, Santos RMD, Garcia PO, Batista TA, Car- van Borkulo CD, Epskamp S, van Borkulo MC (2016) Package ‘Ising-
valho WAC, Oliveira Filho ATD (2014) Floristic composition and Fit’. https://​cran.r-​proje​ct.​org/​web/​packa​ges/​Ising​Fit/​Ising​Fit.​pdf.
structure of an upper montane cloud forest in the Serra da Man- Accessed 30 June 2022
tiqueira mountain range of Brazil. Acta Bot Brasilica 28:456–464. van Borkulo CD, Boschloo L, Borsboom D, Penninx BWJH, Waldorp
https://​doi.​org/​10.​1590/​0102-​33062​014ab​b3239 LJ, Schoevers RA (2019) Package ‘NetworkComparisonTest’.
R Core Team (2020) R: A language and environment for statistical https://c​ ran.r-p​ rojec​ t.o​ rg/​web/p​ ackag​ es/​Netwo​rkCom​paris​onTest/​
computing. R Foundation for Statistical Computing, Vienna, Aus- Netwo​rkCom​paris​onTest.​pdf. Accessed 7 July 2022
tria. https://​www.​Rproj​ect.​org/ van Steenderen C (2022) Package ‘BinMat’. https://​cran.r-​proje​ct.​org/​
Rada F, Azócar A, García-Núñez C (2019) Plant functional diversity in web/​packa​ges/​BinMat/​BinMat.​pdf. Accessed 10 July 2022
tropical Andean páramos. Plant Ecol Divers 12:539–553. https://​ Vasconcelos MFD (2011) O que são campos rupestres e campos de
doi.​org/​10.​1080/​17550​874.​2019.​16743​96 altitude nos topos de montanha do Leste do Brasil? Braz J Bot
Rahbek C, Borregaard MK, Colwell RK, Dalsgaard BO, Holt BG, 34:241–246. https://​doi.​org/​10.​1590/​S0100-​84042​01100​02000​12
Morueta-Holme N, Nogues-Bravo D, Whittaker RJ, Fjeldså Vasconcelos VV (2016) Mapa de estratificação de altitude para veg-
J (2019) Humboldt’s enigma: what causes global patterns of etação no Brasil. Geosul 31(62):7–18. https://​doi.​org/​10.​5007/​
mountain biodiversity? Science 365:1108–1113. https://​doi.​org/​ 2177-​5230.​2016v​31n62​p7
10.​1126/​scien​ce.​aax014 Veloso HP, Rangel-Filho ALR, Lima JCA (1991) Classificação da
Ribeiro KT, Freitas L (2010) Potential impacts of changes to Brazilian vegetação brasileira, adaptada a um sistema universal. Instituto
forest code in campos rupestres and campos de altitude. Biota Brasileiro de Geografia e Estatística, Rio de Janeiro
Neotrop 10:239–246. https://​doi.​org/​10.​1590/​S1676-​06032​01000​ White RP, Murray S, Rohweder M, Prince SD, Thompson KM (2000)
04000​29 Grassland ecosystems. World Resources Institute, Washington,
Ribeiro KT, Medina BMO, Scarano FR (2007) Species composition DC
and biogeographic relations of the rock outcrop flora on the high Wickham H (2016) ggplot2: elegant graphics for data analysis.
plateau of Itatiaia, SE-Brazil. Braz J Bot 30:623–639. https://​doi.​ Springer-Verlag, New York
org/​10.​1590/​S0100-​84042​00700​04000​08 Zizka A, Silvestro D, Andermann T, Azevedo J, Duarte Ritter C, Edler
Rodrigues MT, Cassimiro J, Pavan D, Curcio FF, Verdade VK, Pel- D, Farooq H, Herdean A, Ariza M, Scharn R, Svantesson S (2019)
legrino KCM (2009) A new genus of microteiid lizard from the CoordinateCleaner: standardized cleaning of occurrence records
Caparaó Mountains, southeastern Brazil, with a discussion of from biological collection databases. Methods Ecol Evol 10:744–
relationships among Gymnophthalminae (Squamata). Am Mus 751. https://​doi.​org/​10.​1111/​2041-​210X.​13152
Novit. https://​doi.​org/​10.​1206/​622.1 Zizka A, Carvalho FA, Calvente A, Baez-Lizarazo MR, Cabral A, Coe-
Safford HDF (1999a) Brazilian Páramos I. An introduction to the lho JFR, Colli-Silva M, Fantinati MR, Fernandes MF, Ferreira-
physical environment and vegetation of the campos de altitude. J Araújo T, Moreira FGL (2020) No one-size-fits-all solution to
Biogeogr 26:693–712. https://​doi.​org/​10.​1046/j.​1365-​2699.​1999.​ clean GBIF. PeerJ 8:e9916. https://​doi.​org/​10.​7717/​peerj.​9916
00313.x
Safford HDF (1999b) Brazilian Páramos II. Macro-and mesoclimate of Publisher's Note Springer Nature remains neutral with regard to
the campos de altitude and affinities with high mountain climates jurisdictional claims in published maps and institutional affiliations.
of the tropical Andes and Costa Rica. J Biogeogr 26:713–737.
https://​doi.​org/​10.​1046/j.​1365-​2699.​1999.​00312.x Springer Nature or its licensor (e.g. a society or other partner) holds
Safford HDF (2001) Brazilian Páramos. III. Patterns and rates of post- exclusive rights to this article under a publishing agreement with the
fire regeneration in the Campos de Altitude. Biotropica 33:282– author(s) or other rightsholder(s); author self-archiving of the accepted
302. https://​doi.​org/​10.​1111/j.​1744-​7429.​2001.​tb001​79.x manuscript version of this article is solely governed by the terms of
Safford HDF (2007) Brazilian Páramos IV. phytogeography of the such publishing agreement and applicable law.
campos de altitude. J Biogeogr 34:1701–1722. https://​doi.​org/​
10.​1111/j.​1365-​2699.​2007.​01732.x

13