RESEARCH ARTICLE
Diversity and composition of herbaceous
angiosperms along gradients of elevation and
forest-use intensity
Jorge Antonio Go´mez-Dı´az1*, Thorsten Kro¨mer2, Holger Kreft3, Gerhard Gerold1, Ce´sar
Isidro Carvajal-Herna´ndez4, Felix Heitkamp1
1 Section of Physical Geography, Georg-August-Universita¨t Go¨ttingen, Go¨ttingen, Germany, 2 Centro de
Investigaciones Tropicales, Universidad Veracruzana, Xalapa, Veracruz, Mexico, 3 Department of
Biodiversity, Macroecology & Biogeography, Georg-August-Universita¨t Go¨ttingen, Go¨ttingen, Germany,
4 Herbario CIB, Instituto de Investigaciones Biolo´gicas, Universidad Veracruzana, Xalapa, Veracruz, Mexico
* jgomezd@gwdg.de
Abstract
Terrestrial herbs are important elements of tropical forests; however, there is a lack of
research on their diversity patterns and how they respond to different intensities of forest-
use. The aim of this study was to analyze the diversity of herbaceous angiosperms along
gradients of elevation (50 m to 3500 m) and forest-use intensity on the eastern slopes of the
Cofre de Perote, Veracruz, Mexico. We recorded the occurrence of all herbaceous angio-
sperm species within 120 plots of 20 m x 20 m each. The plots were located at eight study
locations separated by ~500 m in elevation and within three different habitats that differ in
forest-use intensity: old-growth, degraded, and secondary forest. We analyzed species rich-
ness and floristic composition of herb communities among different elevations and habitats.
Of the 264 plant species recorded, 31 are endemic to Mexico. Both α- and γ-diversity display
a hump-shaped relation to elevation peaking at 2500 m and 3000 m, respectively. The rela-
tive contribution of between-habitat β-diversity to γ-diversity also showed a unimodal hump
whereas within-habitat β-diversity declined with elevation. Forest-use intensity did not affect
α-diversity, but β-diversity was high between old-growth and secondary forests. Overall, γ-
diversity peaked at 2500 m (72 species), driven mainly by high within- and among-habitat β-
diversity. We infer that this belt is highly sensitive to anthropogenic disturbance and forest-
use intensification. At 3100 m, high γ-diversity (50 species) was driven by high α- and within-
habitat β-diversity. There, losing a specific forest area might be compensated if similar
assemblages occur in nearby areas. The high β-diversity and endemism suggest that mixes
of different habitats are needed to sustain high γ-richness of terrestrial herbs along this ele-
vational gradient.
Introduction
The majority of biomes are undergoing rapid changes, especially in the tropics [1]. Conse-
quently, growing human pressure on ecosystems poses a marked threat to global biodiversity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 1 / 17
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPENACCESS
Citation: Go´mez-Dı´az JA, Kro¨mer T, Kreft H, Gerold
G, Carvajal-Herna´ndez CI, Heitkamp F (2017)
Diversity and composition of herbaceous
angiosperms along gradients of elevation and
forest-use intensity. PLoS ONE 12(8): e0182893.
https://doi.org/10.1371/journal.pone.0182893
Editor: RunGuo Zang, Chinese Academy of
Forestry, CHINA
Received: September 16, 2016
Accepted: July 26, 2017
Published: August 8, 2017
Copyright: © 2017 Go´mez-Dı´az et al. This is an
open access article distributed under the terms of
the Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.
Funding: This work was supported by the
Deutsche Forschungsgemeinschaft (HE 6726/4).
Researcher mobility was granted within the project
BIOVERA to TK (CONACyT) and FH (by the DAAD
with funding of the Bundesministerium fu¨r Bildung
und Forschung (BMBF)). JAGD was funded by the
Consejo Nacional de Ciencia y Tecnologı´a (216230/
311672) and the Deutshce Akademischer
[2]. Considering current rates of deforestation and forest degradation [3], undisturbed forests
will become scarce and increasingly fragmented [4]. Deforestation for agriculture, as well as
non-sustainable agrarian and forestry practices, increases the demand for new land threaten-
ing primary forests and associated biodiversity [5]. Forest-use intensity is defined as a forest
disturbance derived from the human action acting at a broader landscape level. The effects of
forest conversion on plant diversity are well known [6]; however, there is a lack of knowledge
on how anthropogenic forest-use intensity affects diversity and composition of plant commu-
nities [7]. Forest degradation may have different effects on biodiversity, depending on the eco-
system, the kind of degradation (temporal and spatial extent, intensity) and the taxa of interest
[6].
A global meta-analysis of trends of forest degradation found an average loss of 18% of spe-
cies richness due to forest use [8]. According to the analysis, land-use change and the increase
of non-native invasive species have the highest negative impact on species richness [8]. How-
ever, some landscapes have experienced increases in plant species richness because invasions
of exotic species are exceeding the loss of native species [9]. The magnitude of anthropogenic
change and high diversity in tropical mountains calls for more studies to identify the patterns
of diversity change due to forest-use intensity along elevational gradients [10].
Little is known about the influence of forest-use intensity on herbaceous angiosperms
despite their importance for tropical diversity and ecosystem function [11]. The few studies
available have mixed results with forest-use intensity reported to have positive, neutral, or neg-
ative effects [12]. Moreover, high numbers of primary forest species and endemic species have
been found in naturally regenerating [13] and secondary forests. Thus, such habitats might
help to conserve endemic species [14]. Forest use may also lead to an increase in species num-
bers due to an alteration of the light regime and a suppression of more competitive herbs [15].
Contrasting findings may reflect different study conditions (biome, ecosystem, taxa of inter-
est), different scales (e.g. landscape, plot) and sampling techniques involved [16]. Therefore, it
is important to carry out more empirical research to quantify the effects of forest-use intensity
on herbs using a robust and replicated study design that accounts for the different components
of herb diversity (α, β, and γ).
Despite the uncertainties about forest-use intensity effects on species richness, changes in
species composition have been reported often with the rarest species found mainly in native
communities [17]. Several studies have shown that with increasing forest-use intensity, local
(α-diversity) and total (γ-diversity) species richness declined linearly, whereas species turnover
between plots increased [18]. The forest-use intensity in the most intensively used plots led to
sparse herb layer and reduced species richness [19].
Studies on latitudinal and elevational gradients show that effects of forest-use intensity on
plant diversity may change depending on the ecosystem or ecozone. In general, α- and γ-diver-
sity of plants decrease with increasing latitude, [20,21], and the highest diversity of plants and
various others taxa is usually concentrated in tropical lowland areas with high and evenly dis-
tributed rainfall [22,23]. Several studies of tropical elevational gradients have shown mid-ele-
vations peaks in species diversity followed by a linear decrease [20] for various plant groups
[24]. Most of the previous studies on elevational gradients of forest herbs focused on selected
herbaceous families [25] and were conducted in near-natural ecosystems. Given the ongoing
degradation of primary forests [6], studies on elevational gradients should be extended to
include habitats that differ in forest-use intensity or anthropogenic influence.
This study aims to assess how herbaceous angiosperm diversity varies along elevational gra-
dients in the tropics and how those patterns are affected by variation in forest use intensity.
We thus focused on patterns of α-, β-, and γ-diversity of herbaceous angiosperms along com-
bined gradients of elevation and forest-use intensity. The study was conducted along the
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 2 / 17
Austauschdienst (91549681). The research of TK
was supported by CONACyT. The funders had no
role in study design, data collection and analysis,
decision to publish, or preparation of the
manuscript.
Competing interests: The authors have declared
that no competing interests exist.
Eastern slopes of the volcano Cofre de Perote in central Veracruz, Mexico, along an elevational
gradient from sea level up to 3500 m, which exhibits a large range of environmental conditions
over a short geographic distance of just c. 80 km. We established plots at eight different loca-
tions (separated by c. 500 m in elevation) and with three different forest-use intensities (old-
growth, degraded, and secondary forest).
Methods
Study area
We established eight sites along an elevational gradient between 30 and 3540 m on the Eastern
slopes of the National Park Cofre de Perote, an extinct volcano of 4282 m elevation in the cen-
tral part of the state of Veracruz, Mexico (Fig 1). This region is located at the junction of the
Trans-Mexican volcanic belt and the Sierra Madre Oriental, a mountainous area between 19˚
25’ 5.7” and 19˚ 36’ 54” N and 94˚ 44’ 43.5” and 97˚ 09’ 36.9” W. The state of Veracruz covers
72420 km2 and has a diverse angiosperm flora (6876 species), which represents about 31% of
the Mexican flora [26]. More than 80% of Veracruz’ primary vegetation has been converted
Fig 1. Map of the Eastern slopes of the Cofre de Perote in Veracruz state, Mexico. Black dots show the study locations. 1 = La Mancha, 2 =
Palmarejo, 3 = Chavarrillo, 4 = Los Capulines, 5 = El Zapotal, 6 = El Encinal, 7 = Los Pescados, and 8 = El Conejo.
https://doi.org/10.1371/journal.pone.0182893.g001
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 3 / 17
and the remaining parts are highly fragmented [27], therefore it is recognised as a priority
region for conservation in Mexico [28].
We placed our study locations at the following elevations above sea level: 30–50 m, 610–670
m, 900–1010 m, 1470–1650 m, 2020–2230 m, 2470–2600 m, 3070–3160 m and 3480–3540 m
(Table 1, Fig 1). To simplify, from now on we will refer to every site as a categorical unit (50,
650, 1000, 1500, 2100, 2500, 3100, 3500 m). We used a Garmin1 GPSMAP 60Cx device to
record information about geographical reference and elevation.
Data collection
The sampling design, as well as frequently used terms are shown in Fig 2. Fieldwork was con-
ducted between February 2012 and January 2014. We sampled the presence/absence of terres-
trial herbaceous angiosperms, which we defined as plants without a persistent aboveground
woody stem or plants with only slightly woody stems, rooted on the forest floor and of short
stature (generally < 1 m); vines were excluded [31]. We recorded all species in 20 m × 20 m
plots, without considering seedlings [32]. We choose a plot size of 400 m2 because for the her-
baceous flora in humid tropical forests this area is regarded to be representative, while small
enough to minimise within-plot variation of abiotic factors [33,34]. We located the plots in the
three different habitats subjected to different degrees of forest-use intensity; old-growth (OG),
degraded (DE), and secondary forest (SE) stands (n = 5 plots for each habitat). These catego-
ries follow Newbold et al. [35] and are defined in Table 2. Each habitat was present at each of
the eight locations resulting in a total number of 120 plots and a sampled area of 48000 m2.
The Secretarı´a de Medio Ambiente y Recursos Naturales (SEMARNAT SGPA/DGVS/2405/
14) issued us a plant collection permit (NOM-059-SEMARNAT-2010), which covered the
whole study area and the collection of protected species mentioned in the Mexican legislation.
We used elevation (m), mean annual temperature (˚C), mean annual precipitation (mm/a)
and habitat (a factor with three levels OG, DE and SE), as well as their interaction as explana-
tory variables. We obtained mean annual temperature (MAT) using data loggers (HOBO PRO
v2) for one year (January-December 2014). We installed a total of 42 data loggers, two in every
forest-use habitat at six elevations (500 m, 1000 m, 1500 m, 2500 m, 3000 m, and 3500 m)
along the entire gradient. Within the plots, we placed data loggers on trees at a height between
two and three meters. We obtained mean annual temperature data for the two elevations (50
m and 2000 m) and data of the mean annual precipitation (MAP) from eight climatological
stations (near to the sampling sites) operating along the elevational gradient during the period
1951–2010 [30].
Table 1. List of the study locations and climatic conditions along the elevational gradient at the Cofre de Perote, central Veracruz, Mexico. Informa-
tion is given on elevational range, vegetation type according to Leopold [29], mean annual temperature (MAT), mean annual precipitation (MAP), days of rain
(DR), and days below 0˚C according to National Meteorological Service of Mexico (data from 1951–2010) [30].
Location Elevation (m) Vegetation type MAT (˚C) MAP (mm) DR DB
La Mancha 30–50 Tropical semi-humid deciduous forest 26 1221 81 0
Palmarejo 610–670 Tropical semi-humid deciduous forest 23 938 86 0
Chavarrillo 900–1010 Tropical Quercus forest 21 1552 123 0
Los Capulines 1470–1650 Humid montane forest 18 1598 145 0
El Zapotal 2020–2230 Humid montane forest 14 3004 199 3
El Encinal 2470–2600 Pinus-Quercus forest 12 1142 100 12
Los Pescados 3070–3160 Pinus forest 10 821 113 14
El Conejo 3480–3540 Abies forest 8 829 112 16
https://doi.org/10.1371/journal.pone.0182893.t001
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 4 / 17
Species identification
We collected at each location, but not in every plot, specimens of all species if possible in qua-
druplicates and deposited at the Mexican herbaria CIIDIR, MEXU, XAL, and XALU. Details
about species identifications, geographical distribution and classification can be found in
Go´mez-Dı´az et al. [36]. The presence/absence data of all the species found at the plots can be
found at the Supporting Information (S1 File).
Data analyses
We calculated for each of the 120 herb communities in the dataset, i.e. replicated plots (5 plots
per habitat x 8 locations = 120), α-diversity, the proportion of endemics and β-diversity [37].
Alpha-diversity is the count of the number of species (plot-based species richness) in each
standardised replicate plot [38]. In order to evaluate the value of endemic species, we measured
the proportion of endemic species per plot. We used the dissimilarity (1-S) variant of the
Fig 2. Schematic representation of the sampling design. We measured α-diversity in plots of 20 m x 20 m
given as a mean of five plots. Five plots represent one habitat. We defined habitat as a homogenous type of
forest-use intensity within one location. A location is representative of an elevational belt and harbors three
different habitats (old-growth OG, degraded DE, and secondary SE). We measured two different β-diversities
based on pairs of plots. Within-habitat β-diversity represents the compositional heterogeneity of a habitat. It is
measured as the 1-Sørensen index based on multiple pairwise comparisons of the five plots within each
habitat of a specific location. Between-habit β-diversity represents the compositional heterogeneity between
different forest-use intensity. Measurement is similar to within-habitat β-diversity, but multiple pairwise
comparisons based on the plots between habitats of a specific location. We defined γ-diversity as the total
number of the local species pool across all 15 plots within a location, i.e. three habitats with five plots each.
https://doi.org/10.1371/journal.pone.0182893.g002
Table 2. Classification of habitats with different forest-use intensities according to the main physiognomic characteristic, the gap fraction in the
canopy, dominance of canopy trees, the percentage of shrubs, and the presence of lianas [35].
Habitat Characteristic Gaps (%) Forest-use intensity Canopy trees Shrub (%) Lianas
Old-growth No obvious forest-use, dominance of mature trees <10 Low High <30 No
Degraded Selective logging, grazing and understory removal 11–25 Medium Low 30–50 Low
Secondary Regrown after clear-cut >25 High very low >50 High
https://doi.org/10.1371/journal.pone.0182893.t002
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 5 / 17
Sørensen index [37] as a measure of β-diversity and we calculated it as:
b  diversity ¼ 1  S ð1Þ
where S is the Sørensen index:
S ¼
2C
Aþ B
ð2Þ
which is a coefficient of association, where a value of 1 shows that a pair of plots under consid-
eration has exactly the same species. The Sørensen index can be adjusted to measure species
turnover (effective or real) [39]. The index 1-S is also a measure of β-diversity because if a coef-
ficient is near to 1, it shows that the units do not share species, and, therefore, they have high
β-diversity [40]. As a measure for landscape-scale β-diversity, we compared the effect of habi-
tats on floristic composition between OG and DE, OG and SE as well as DE and SE at each site
[41]. We calculated plot-by-plot, i.e. each plot of habitat A was compared to five or four (in the
case of within β-diversity) other plots of habitat B. Additionally, we calculated the standard
errors for the β-diversity estimator, letting a statistically rigorous contrast of two or other simi-
larity index values. Standard errors were calculated by the bootstrap process, which needs
resampling the observed data for pairs of samples and re-computing the estimators N times
[41]. We performed these analyses of β-diversity in EstimateS 9.1.0 [42].
We used a set of generalized linear models (GLM) and informatic theoretic approach based
on the bias-corrected Akaike’s information criterion (AICc) to examine the extent to which
herb α-diversity, the proportion of endemics and β-diversity is related to elevation, habitat,
MAT and MAP. These sets of models aimed to describe the pattern of distribution of the
response variables, namely herb α-diversity, the proportion of endemics and β-diversity. We
used elevation, elevation2, habitat, MAT, MAT2, MAP, MAP2 and various two-way and three-
way interactions as explanatory variables in these models. We included polynomial terms of
elevation, MAT and MAP in order to detect potential hump shapes in the relationships
between elevation and the inter-correlated climatic variables with the response variables. We
did not include the data below 500 m in the case of β-diversity due to the lack of species in six
plots and we used “habitat transition” with six levels (OG to OG, OG to DE, OG to SE, DE to
DE, DE to SE and SE to SE) instead of the independent variable “habitat”. We constructed
thirty-one plausible models combining these variables for this set of models [38].
We checked data for normality using the Shapiro-Wilk normality test. We checked the
homogeneity of variance using the Bartlett test. None of the variables followed a normal distri-
bution. Therefore, we chose an error structure for each variable depending on the nature of
the variable. We choose the error structure according to the “descdist” function of the R pack-
age “fitdistrplus” version 1.0–7. We used a negative binomial error structure for models of α-
diversity. We used a beta error structure for models of the proportion of endemics. We used a
log-normal error structure for models of β-diversity. We used maximum likelihood estimation
and AICc values compared to choose the best model for each response variable in each set of
models. We performed all analyses in R 3.2.1 [43]. Finally, we calculated an R2 for each model
[38].
Multiplicative diversity partitioning. We used multiplicative diversity partitioning [44]
and focused on three distinct components of community differentiation (α-diversity), (β-
within habitat) and compositional dissimilarity (β-between habitats) [45,46]. We partitioned
total γ-diversity per location into multiplicative components representing per plot α-diversity,
within habitats β-diversity and among habitats β-diversity. We used the function “multipart”
of the R package “vegan” to partition diversity [45,47]. Through the text, we will refer to the
results of this analysis as relative (relative α-diversity, β-within habitat, etc.)
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 6 / 17
Estimation of species diversity (Hill numbers). Following Chao and Jost [48], we used a
diversity profile estimator of qD (Hill numbers). The diversity estimator of order q in each pro-
file represents the asymptote in the rarefaction and extrapolation curves [49]. We calculated
the profile estimator with the function “diversity” of the R package “SpadeR” [50]. We consid-
ered the cases of q from 0 to 3 with increments of 0.25.
Results and discussion
Alpha diversity
Elevation, elevation2, habitat, MAT, MAP2, interactions between elevation and elevation2, an
interaction between elevation and MAT, as well as an interaction between elevation2 and
MAT modeled best the α-diversity (Table 3). This model explained a large amount of the vari-
ance in species richness (GLM R2 = 0.69). Alpha-diversity followed a hump-shaped pattern,
showing a peak at 2500 m and declines towards the extremes of the gradient that do not change
depending on the habitat (Fig 3, Table 4). Fewer species are found at high MAT and this effect
is greatest at low elevations due to the inclusion of the MAT + elevation interaction.
Hump-shaped diversity patterns have been reported for different groups of vascular plants
[24], such as palms, Acanthaceae, Bromeliaceae and woody plants along tropical mountains
[51–53]. Compared to previous findings, our results show the peak in α-diversity shifted
towards higher, instead of mid-elevations [54]. One explanation for this pattern is that herbs
are adapted to cold climates and had a success in temperate zones due to be annual and the
production of underground structures (e.g. rhizomes and stolons) [55].
Lower elevations in Veracruz are subject to prolonged dry seasons (Table 2), which may
limit α-diversity. With increasing elevation, precipitation increases, while temperatures and
potential evapotranspiration decrease. Species richness of ferns has been reported to be posi-
tively related to humidity [56]. In our study, angiosperm richness peaked between 2500 m and
3100 m, where precipitation and the number of rainy days already decrease. However, the
Pinus-Quercus forests at 2500 m are often subject to fog, whereas in Pinus forests (3100 m)
light transmission to the forest floor is high [57], which likely increases the ground cover of
angiosperms and thus also their diversity.
Surprisingly, forest-use intensity had no significant effect on α-diversity (Fig 3). The lack of
a detectable net-change in α-diversity might indicate that the level of forest-use intensity is still
relatively moderate; however, other life forms (e.g. trees, epiphytes and ferns) might show con-
trasting patterns. It is quite well documented that forest herbs profit from better light condi-
tions in DE or SE. Newbold et al. [35], for example, found that the richness of vascular plant
species can increase by 40% due to the conversion of old-growth forests to secondary
Table 3. Summaries of generalized linear models. The summaries link herb α-diversity, the proportion of endemics and β-diversity to environmental
explanatory variables, along an elevational gradient at central Veracruz, Mexico. We reported the best models, according to the bias-corrected Akaike infor-
mation criterion (AICc). We also give the change in AICc between the best model and the next best and worst. Finally, an R2 measuring variation explained by
the model is given.
Response Model AICc ΔAICc (next
best)
ΔAICc
(worst)
R2
α-diversity Elevation + elevation2 + habitat + MAT + MAP2 + elevation:MAT + elevation:elevation2
+ MAP:elevation2
570.96 0.1 1023.3 0.69
the proportion of
endemics
Elevation + elevation2 + habitat + MAT + MAP2 + elevation:MAT + elevation:elevation2
+ elevation:MAT2 + elevation:elevation2:MAT
-162.50 1.9 111.7 0.41
β-diversity Elevation + elevation2 + change + MAT + MAP + elevation:change + elevation:MAT
+ elevation2:change + elevation2:MAT + change:MAT + MAT:MAP + elevation:change:
MAT + elevation2:change:MAT
287.41 1.7 450.6 0.49
https://doi.org/10.1371/journal.pone.0182893.t003
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 7 / 17
vegetation, but more severe habitat conversion, e.g. from forest to intensive cropland,
decreases species richness.
Proportion of endemic species
The best model for the proportion of endemic species was the one that included elevation, ele-
vation2, habitat, MAT, MAP2 and the interaction between elevation and MAT, elevation and
elevation2, elevation2 and MAT and a triple interaction with elevation, elevation2 and MAT
(Fig 4, Table 3). The proportion of endemic species showed a non-linear pattern along the ele-
vational gradient with the highest proportion at 500 m (Fig 4, Table 4). There is no effect of
forest-use intensity on the proportion of endemic species, the lack of a pattern on the endemic
species can be attributed to the adaptation of the endemic herbaceous plant’s species of Mexico
to several patterns of disturbance or the extinction of previous endemic species. This model of
the proportion of endemic species had an R2 of 0.41.
Beta and gamma diversity
In contrast to richness, forest-use intensity had marked effects on floristic composition, which
was most visible when looking at habitat transitions (Table 5). Within-habitat β-diversity was
generally lower (0.42–0.52) than between-habitat β-diversity (0.58–0.66). Unsurprisingly, we
Fig 3. Alpha-diversity of herbaceous angiosperms along gradients of elevation and forest-use
intensity at the Cofre de Perote, central Veracruz, Mexico. We fit the lines from a negative generalized
linear model (GLM), the shaded area marks confidence intervals (CI = 1.96 times standard error). Difference
to zero is significant for the intercept (xy, OG at 50 m, p < 0.001), but it is not significant for the effect of habitat
(DE: p = 0.453, SE: p = 0.446). Observed species richness on 20 m × 20 m plots along the elevational
gradient for OG (green), DE (blue), and SE (red).
https://doi.org/10.1371/journal.pone.0182893.g003
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 8 / 17
found the highest dissimilarity in the transition from OG to SE, whereas the most homogenous
species pool was within OG. The high β-diversity between habitats indicated wider environ-
mental differences [58] than the heterogeneity of plots within the same habitat type.
The best model for β-diversity included the main effects of elevation, elevation2, change,
MAT, MAP and interactions between elevation and change, elevation and MAT, elevation2
and change, elevation2 and MAT, change and MAT, MAT and MAP, as well as triple interac-
tions among elevation, change, MAT and elevation2, change, MAT (Table 3). This model
explains little variation (R2 = 0.49). There was a marked effect of elevation on β-diversity (Fig
5, Table 4). Within-habitat β-diversity showed a clear humped-shaped pattern for OG and SE
with peaks between 2500 m and 3000 m. DE, however, had their highest β-diversity at 650 m
with a subsequent decline. Obviously, DE at lower elevations exhibits a different response to
environmental conditions compared to OG and SE. Degradation may lead to a higher hetero-
geneity of environmental conditions and, consequently, offer diverse niches triggering com-
munity differentiation [59]. Beta-diversity between-habitats was generally high but varied with
the type of habitat transition. In the transition from OG to SE, there was a turnover of c. 50%
of the species at both extremes of the elevational gradient. Between 1500 m and 2500 m, how-
ever, 75% of the species were different when comparing OG to SE. Habitat transitions related
to degradation (OG-DE, DE-SE) showed highest β-diversity at 650 m and 2500 m and declined
with increasing elevation. Endemic species contribute to this pattern only at 650 m where we
found their peak (Fig 4). Especially at 3100 m and 3500 m, the change in species composition
Table 4. Parameter estimates from generalized linear models. The parameters link herb α-diversity, the proportion of endemics and β-diversity to envi-
ronmental explanatory variables, along an elevational gradient at central Veracruz, Mexico. Estimates are on the standardized scale ± standard error. We
marked significant estimates with an asterisk (* < = 0.05, ** < = 0.01 and *** < 0.001). Empty cells indicate terms not included in the best model for a given
response variable.
Estimates
Term α-diversity the proportion of endemics β-diversity
Elevation -5.66 ± 0.04*** 7.22 ± 1.25 -2.89 ± 0.38***
Elevation2 7.09 ± 0.28** -5.74 ± 1.72 1.98 ± 0.24***
Habitat.DE 0.01 ± 7.78E-2 0.08 ± 0.02 0.05 ± 3.72E-3
Habitat.SE 16.76 ± 0.85 0.27 ± 0.02 -7.23 ± 0.28*
MAT -1.23 ± 0.02*** -0.84 ± 1.25 -2.47 ± 0.28***
MAP 29.06 ± 6.68
MAP2 -0.04 ± 1.77E-03*** 0.70 ± 0.22***
Elevation:MAT 9. 23E-4 ± 2.24E-5** 5.26 ± 3.60 0.26 ± 0.03***
Elevation:Habitat.DE 0.35 ± 0.01
Elevation:Habitat.SE 0.57 ± 0.07
Elevation:elevation2 -7.96E-4 ± 6.49E-5* -0.37 ± 0.06
Elevation2:Habitat.DE -53.4 ± 6.65*
Elevation2:Habitat.SE 5.42E-6 ± 8.04E-7
Elevation2:MAT -0.07 ± 6.49E-3 -8.41 ± 4.13*** 0.03 ± 6.67E-3***
Habitat.DE:MAT -71.2 ± 7.82
Habitat.SE:MAT 28.46 ± 4.17*
MAT:MAP 34.74 ± 1.94***
Elevation:elevation2:MAT 2.10 ± 0.89**
Elevation:Habitat.DE:MAT -0.05 ± 1.94E-3*
Elevation:Habitat.SE:MAT 10.24 ± 2.29*
Elevation2:Habitat.DE:MAT 1.23E-5 ± 1.94E-6
Elevation2:Habitat.SE:MAT -3.70E-3 ± 2.33E-4*
https://doi.org/10.1371/journal.pone.0182893.t004
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 9 / 17
with the transition from OG to DE was relatively low. This indicates that present environmen-
tal conditions favour a spectrum of adapted species [60], which thrive regardless of the habitat
type. Above 3100 m there are fewer species, which are adapted to extreme climate events, such
as days below 0˚C, lower temperature and precipitation (Table 1).
We revealed the relative contribution of relative α, relative within-habitat β, and relative
between-habitats β-diversity to γ using the multiplicative partitioning approach (Fig 6). The
contribution of relative α-diversity was only between 3 and 13 (Fig 6). Diverse forest habitats
support high species diversity and lead to high relative β-diversity between-habitats (βb)
Fig 4. The proportion of endemic species of herbaceous angiosperms along gradients of elevation
and forest-use intensity at the Cofre de Perote, central Veracruz, Mexico. We fit the lines from a GLM
with a beta error family, the shaded area marks confidence intervals (CI = 1.96 times standard error).
Difference to zero is not significant for the intercept (xy, OG at 50 m, p = 0.066) as well as for the effect of
habitat (DE: p = 0.771, SE: p = 0.226). Observed proportion of endemic species on 20 m × 20 m plots along
the elevational gradient for OG (green), DE (blue), and SE (red).
https://doi.org/10.1371/journal.pone.0182893.g004
Table 5. Average effects of the habitat change. Mean β-diversity at every habitat transition, letters in
superscript differences in groups after Tukey posthoc test (HSD = 0.138).
Habitat transition β-diversity (1-S)
Old-growth to secondary 0.66a
Degraded to secondary 0.61ab
Old-growth to degraded 0.58ab
Degraded to degraded 0.52bc
Secondary to secondary 0.48bc
Old-growth to old-growth 0.42c
https://doi.org/10.1371/journal.pone.0182893.t005
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 10 / 17
confirming the importance of these habitats. We found a hump-shaped pattern of relative α-
diversity (number of species per plot) with a peak at 2500 m and decreases towards the
extremes. There was also a hump-shaped pattern of the relative contribution of between-habi-
tat β-diversity (difference in species composition between plots within a habitat) and a decline
of within-habitat β-diversity with elevation. The adaptation of some herbs to cold climates
could increase of relative α-diversity at higher elevations [55], which is similar to the pattern
found by Cicuzza et al. [34] concerning the effect of forest structure (e.g. more open canopies)
[61]. Yang et al. [62] also found the pattern of relative β-diversity, which was explained by
changes in climatic variables as elevation-related vegetation zones reflect climatic zones. This
is observable in our results as the most remarkable changes in values of relative βb-diversity
occurred between different climatic zones (Fig 6), which demonstrates the effects of an eleva-
tion-related climate gradient on relative βb-diversity patterns [62]. The increasingly stronger
control of climate over other environmental factors (e.g. soil factors) and the decrease of forest
heterogeneity at higher elevations may explain the decline in relative βw-diversity which is con-
sistent with the results of Akhtar and Bergmeier [63], who worked with herbs in the mountains
of Northern Pakistan.
The most remarkable effect of forest-use intensity was on β-diversity and it was highest
between 2100 m and 2500 m (Fig 5). It has been reported that forest-use intensity decreases β-
diversity due to the propagation of exotic and opportunist species that can lead to a ‘biotic
homogenization’ [64]. Beta-diversity was, on average, in our sample of landscapes, lower
within habitats than between habitats (Table 5 and Figs 5 and 6). Additionally, the α-diversity
Fig 5. Compositional heterogeneity (as a measure for β-diversity) among different changes in forest habitats along
the elevational gradient at the Cofre de Perote, central Veracruz, Mexico. Values are 1-Sørensen values as means
across all plots. Shadows are standard errors computed by a GLM with log-normal error family.
https://doi.org/10.1371/journal.pone.0182893.g005
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 11 / 17
in our study did not change, but the dissimilarity between habitats was high (0.594 to 0.665)
(Table 5).
Here, we argue that all habitats and even anthropogenic habitats contribute to the herba-
ceous angiosperm richness in our study region with seven species endemic to Mexico and
seven native species exclusive from these habitats [36]. It is, of course, important to note that
our study does not include spatially weighted information about the abundance of different
forest habitats. Pressure on the primary or OG forest is often higher than their ability to regen-
erate. Therefore, there is the risk that OG or even DE will be converted into SE [65]. Although
different successional stages of SE may also harbor high species richness [66], a homogeniza-
tion of habitats will consequently lead to species homogenization by decreasing β-diversity.
The most vulnerable location is the Pinus-Quercus forest at 2500 m because high βb-diver-
sity implies a loss of many OG species during forest degradation. This means that conversion
of a certain area increases the chance that a unique flora is changed in composition and inva-
sive species appear. In addition, this elevation contains the largest number of endemic species
compared to the rest of locations [36], leading to increased vulnerability. Therefore, the Pinus-
Quercus forest at 2500 m should be considered as a priority for conservation, especially because
according to Mittermeier et al. [67] this vegetation type in Mexico has the lowest levels of
Fig 6. Multiplicative diversity partitioning. Multiplicative version isolating pure relative differentiation (e.g. β is
independent of α). Here, the number of distinct units of the lower level of partition can explain the relative β-diversity and
multiply to equal γ-diversity.
https://doi.org/10.1371/journal.pone.0182893.g006
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 12 / 17
protection [68]. For the herbaceous angiosperm group, however, it seems that a well-designed
management plan instead of pure conservation would be beneficial because high habitat het-
erogeneity is required to achieve high species richness.
For the asymptotic analysis, we plotted the estimated asymptotic diversity profiles when q is
between 0 and 3 (Fig 7). The empirical diversities imply that the eight locations differ in each
of the q orders. In contrast, the plots in Fig 7 reveal that for the asymptotic q = 0 the 2100 m
location is substantially more diverse than the other locations. However, for the Shannon
diversity (q = 1), the 2500 m location is substantially more diverse than the other locations. A
similar conclusion is also valid for the Simpson diversity (q = 2), confirming our other analyses
(Figs 3, 5 and 6).
Conclusions
In order to understand how forest-use intensity and elevation affect species diversity and com-
munity composition of herbs, we focused on the three components of diversity (α, β, and γ)
that allowed us to understand the different patterns at the landscape level. We did not find sig-
nificant differences in α-diversity among the three forest systems, a finding that is not in accor-
dance with previous studies [1,4,6,15]. However, forest-use intensity affected the floristic
composition, which varied markedly between habitats. The most important component driv-
ing γ-diversity was the β-diversity between habitats. Thus, different forest-use intensities,
which coexist, increased the species richness in the landscape.
Fig 7. Asymptotic analysis: The asymptotic diversity profile as a function of order q based on the adjusted
data. The estimated diversity profiles for q between 0 and 3 based on the sample frequency counts of herbaceous
angiosperms from eight locations.
https://doi.org/10.1371/journal.pone.0182893.g007
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 13 / 17
Some elevations, and especially the location at 2500 m, were evidently vulnerable, whereas
species richness still depends on a certain degree of forest-use intensity. Our findings clearly
showed that OG at mid-elevations contributed more to regional diversity than DE. In the
case of herbaceous angiosperms, sustainable forest management, such as forest certification
instead of strict protection may be a good way to conserve herbaceous forest plants in the
region. Forest protection systems should consider the important influence of β-diversity to
regional species richness rather than put emphasis completely on the protection at local scale
(α) diversity.
Supporting information
S1 File. Presence/absence data of the herbaceous angiosperms found at the plots sampled
along the elevational gradient at the Cofre de Perote, central Veracruz, Mexico.
(CSV)
Acknowledgments
We thank M. Ruiz-Go´mez, F. Calixto-Benites, C. Alavez-Tadeo, and V. Guzma´n-Jacob for
their help during fieldwork, A. Taylor and M. Campbell for their review of English, and an
anonymous reviewer and D. Waller for their helpful comments. We appreciate the working
facilities provided at the Centro de Investigaciones Tropicales (CITRO), Universidad Veracru-
zana, Xalapa.
Author Contributions
Conceptualization: Jorge Antonio Go´mez-Dı´az, Thorsten Kro¨mer, Ce´sar Isidro Carvajal-
Herna´ndez.
Data curation: Thorsten Kro¨mer.
Formal analysis: Jorge Antonio Go´mez-Dı´az, Holger Kreft, Gerhard Gerold.
Funding acquisition: Thorsten Kro¨mer, Gerhard Gerold, Felix Heitkamp.
Investigation: Jorge Antonio Go´mez-Dı´az, Thorsten Kro¨mer, Gerhard Gerold, Ce´sar Isidro
Carvajal-Herna´ndez, Felix Heitkamp.
Methodology: Jorge Antonio Go´mez-Dı´az, Thorsten Kro¨mer, Gerhard Gerold, Ce´sar Isidro
Carvajal-Herna´ndez, Felix Heitkamp.
Project administration: Thorsten Kro¨mer, Holger Kreft, Gerhard Gerold, Felix Heitkamp.
Resources: Thorsten Kro¨mer, Holger Kreft, Gerhard Gerold, Felix Heitkamp.
Software: Jorge Antonio Go´mez-Dı´az.
Supervision: Thorsten Kro¨mer, Holger Kreft, Gerhard Gerold, Felix Heitkamp.
Validation: Jorge Antonio Go´mez-Dı´az, Thorsten Kro¨mer, Holger Kreft, Gerhard Gerold,
Felix Heitkamp.
Visualization: Jorge Antonio Go´mez-Dı´az, Holger Kreft, Felix Heitkamp.
Writing – original draft: Jorge Antonio Go´mez-Dı´az, Felix Heitkamp.
Writing – review & editing: Jorge Antonio Go´mez-Dı´az, Thorsten Kro¨mer, Holger Kreft,
Gerhard Gerold, Ce´sar Isidro Carvajal-Herna´ndez, Felix Heitkamp.
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 14 / 17
References
1. Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, et al. Global consequences of land
use. Science. 2005; 309: 570–574. https://doi.org/10.1126/science.1111772 PMID: 16040698
2. Godfray HCJ, Beddington JR, Crute IR, Haddad L, Lawrence D, Muir JF, et al. Food security: the chal-
lenge of feeding 9 billion people. Science. 2010; 327: 812–818. https://doi.org/10.1126/science.
1185383 PMID: 20110467
3. Lindenmayer DB, Franklin JF, Fischer J. General management principles and a checklist of strategies
to guide forest biodiversity conservation. Biol Conserv. 2006; 131: 433–445. https://doi.org/10.1016/j.
biocon.2006.02.019
4. Ko¨ster N, Friedrich K, Nieder J, Barthlott W. Conservation of epiphyte diversity in an Andean landscape
transformed by human land use. Conserv Biol. 2009; 23: 911–919. https://doi.org/10.1111/j.1523-1739.
2008.01164.x PMID: 19210304
5. Wright SJ. Tropical forests in a changing environment. Trends Ecol Evol. 2005; 20: 553–560. https://
doi.org/10.1016/j.tree.2005.07.009 PMID: 16701434
6. Gibson L, Lee TM, Koh LP, Brook BW, Gardner TA, Barlow J, et al. Primary forests are irreplaceable for
sustaining tropical biodiversity. Nature. 2011; 478: 378–381. https://doi.org/10.1038/nature10425
PMID: 21918513
7. Flynn DFB, Gogol-Prokurat M, Nogeire T, Molinari N, Richers BT, Lin BB, et al. Loss of functional diver-
sity under land use intensification across multiple taxa. Ecol Lett. 2009; 12: 22–33. https://doi.org/10.
1111/j.1461-0248.2008.01255.x PMID: 19087109
8. Murphy GEP, Romanuk TN. A meta-analysis of declines in local species richness from human distur-
bances. Ecol Evol. 2014; 4: 91–103. https://doi.org/10.1002/ece3.909 PMID: 24455164
9. Ellis E, Antill E, Kreft H. All is not loss: plant biodiversity in the Anthropocene. PLoS One. 2012; 7:
e30535. https://doi.org/10.1371/journal.pone.0030535 PMID: 22272360
10. Becker A, Ko¨rner C, Brun J, Guisan A, Tappeiner U. Ecological and land use studies along elevational
gradients. Mt Res Dev. International Mountain Society; 2007; 27: 58–65. https://doi.org/10.1659/0276-
4741(2007)27[58:EALUSA]2.0.CO;2
11. Willingho¨fer S, Cicuzza D, Kessler M. Elevational diversity of terrestrial rainforest herbs: when the
whole is less than the sum of its parts. Plant Ecol. 2012; 213: 407–418. https://doi.org/10.1007/s11258-
011-9986-z
12. Mayfield MM, Daily GC. Countryside biogeography of neotropical herbaceous and shrubby plants. Ecol
Appl. 2005; 15: 423–439. https://doi.org/10.1890/03-5369
13. Barlow J, Gardner T. Quantifying the biodiversity value of tropical primary, secondary, and plantation
forests. Proc Natl Acad Sci U S A. 2007; 104: 18555–18560. https://doi.org/10.1073/pnas.0703333104
PMID: 18003934
14. Marin-Spiotta E, Silver W, Ostertag R. Long-term patterns in tropical reforestation: plant community
composition and aboveground biomass accumulation. Ecol Appl. 2007; 17: 828–839. PMID: 17494400
15. Buscardo E, Smith G, Kelly D. The early effects of afforestation on biodiversity of grasslands in Ireland.
Biodivers Conserv. 2008; 17: 1057–1072.
16. Naeem S, Raffaelli D, Schmid B, Bengtsson J, Grime JP, Hector A, et al. Biodiversity and ecosystem
functioning: current knowledge and future challenges. Science. American Association for the Advance-
ment of Science; 2001; 294: 804–808. https://doi.org/10.1126/science.1064088 PMID: 11679658
17. Cadotte MW, Borer ET, Seabloom EW, Cavender-Bares J, Harpole WS, Cleland E, et al. Phylogenetic
patterns differ for native and exotic plant communities across a richness gradient in Northern California.
Divers Distrib. 2010; 16: 892–901. https://doi.org/10.1111/j.1472-4642.2010.00700.x
18. Schuldt A, Wubet T, Buscot F, Staab M, Assmann T, Bo¨hnke-Kammerlander M, et al. Multitrophic diver-
sity in a biodiverse forest is highly nonlinear across spatial scales. Nat Commun. Nature Publishing
Group; 2015; 6: 10169. https://doi.org/10.1038/ncomms10169 PMID: 26658136
19. Decocq G, Beina D, Jamoneau A, Gourlet-Fleury S, Closset-Kopp D. Don’t miss the forest for the trees!
Evidence for vertical differences in the response of plant diversity to disturbance in a tropical rain forest.
Perspect Plant Ecol Evol Syst. 2014; 16: 279–287. https://doi.org/10.1016/j.ppees.2014.09.001
20. Gentry AH. Changes in plant community diversity and floristic composition on environmental and geo-
graphical gradients. Ann Missouri Bot Gard. 1988; 75: 1. https://doi.org/10.2307/2399464
21. Kreft H, Jetz W. Global patterns and determinants of vascular plant diversity. Proc Natl Acad Sci U S A.
2007; 104: 5925–30. https://doi.org/10.1073/pnas.0608361104 PMID: 17379667
22. Clinebell RR, Phillips OL, Gentry AH, Stark N, Zuuring H. Prediction of neotropical tree and liana spe-
cies richness from soil and climatic data. Biodiversity and Conservation. 1995. pp. 56–90. https://doi.
org/10.1007/BF00115314
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 15 / 17
23. Gentry AH. Tropical forest biodiversity: distributional patterns and their conservational significance.
Oikos. 1992; 63: 19. https://doi.org/10.2307/3545512
24. McCain C, Grytnes J. Elevational gradients in species richness. Encyclopedia of Life Sciences (ELS).
Chichester: John Wiley & Sons, Ltd; 2010. p. 10. https://doi.org/10.1002/9780470015902.a0022548
25. Kro¨mer T, Acebey A, Kluge J, Kessler M. Effects of altitude and climate in determining elevational plant
species richness patterns: A case study from Los Tuxtlas, Mexico. Flora—Morphol Distrib Funct Ecol
Plants. 2013; 208: 197–210. https://doi.org/10.1016/j.flora.2013.03.003
26. Luna-Vega I, Espinosa D. Geographical patterns and determinants of species richness in Mexico
across selected families of vascular plants: implications for conservation. Syst Biodivers. 2013; 11:
237–256.
27. Muñoz-Villers LE, Lo´pez-Blanco J. Land use/cover changes using Landsat TM/ETM images in a tropi-
cal and biodiverse mountainous area of central-eastern Mexico. Int J Remote Sens. 2008; 29: 71–93.
https://doi.org/10.1080/01431160701280967
28. Williams-Linera G. Tree species richness complementarity, disturbance and fragmentation in a Mexican
tropical montane cloud forest. Biodivers Conserv. 2002; 11: 1825–1843.
29. Leopold A. Vegetation zones of Mexico. Ecology. 1950; 31: 507–518.
30. SMN. Servicio Meteorolo´gico Nacional. Normales climatolo´gicas. In: Servicio Meteorolo´gico Nacional
[Internet]. 2016 [cited 8 Mar 2016].: http://smn.cna.gob.mx/index.php?option=com_content&view=
article&id=164&tmpl=component
31. Poulsen AD. Species richness and density of ground herbs within a plot of lowland rainforest in north-
west Borneo. J Trop Ecol. 1996; 12: 177–190.
32. Carvajal-Herna´ndez CI, Kro¨mer T. Riqueza y distribucio´n de helechos y lico´fitos en el gradiente altitudi-
nal del Cofre de Perote, centro de Veracruz, Me´xico. Bot Sci. 2015; 93: 601. https://doi.org/10.17129/
botsci.165
33. Kessler M, Bach K. Using indicator families for vegetation classification in species-rich Neotropical for-
ests. Phytocoenologia. 1999; 29: 485–502.
34. Cicuzza D, Kro¨mer T, Poulsen AD, Abrahamczyk S, Delhotal T, Piedra HM, et al. A transcontinental
comparison of the diversity and composition of tropical forest understory herb assemblages. Biodivers
Conserv. 2013; 22: 755–772. https://doi.org/10.1007/s10531-013-0447-y
35. Newbold T, Hudson LN, Hill SLL, Contu S, Lysenko I, Senior RA, et al. Global effects of land use on
local terrestrial biodiversity. Nature. 2015; 520: 45–50. https://doi.org/10.1038/nature14324 PMID:
25832402
36. Go´mez-Dı´az JA, Kro¨mer T, Carvajal-Herna´ndez CI, Gerold G, Heitkamp F. Richness and distribution of
herbaceous angiosperms along gradients of elevation and forest disturbance in central Veracruz,
Mexico. Bot Sci. 2017; 95: 1–22. https://doi.org/10.17129/botsci.859
37. Dice LR. Measures of the Amount of Ecologic Association Between Species. Ecology. 1945; 26: 297–
302. https://doi.org/10.2307/1932409
38. Bishop TR, Robertson MP, van Rensburg BJ, Parr CL. Elevation-diversity patterns through space and
time: ant communities of the Maloti-Drakensberg Mountains of southern Africa. Gillman L, editor. J Bio-
geogr. 2014; 41: 2256–2268. https://doi.org/10.1111/jbi.12368
39. Tuomisto H. A consistent terminology for quantifying species diversity? Yes, it does exist. Oecologia.
2010; 164: 853–60. https://doi.org/10.1007/s00442-010-1812-0 PMID: 20978798
40. Anderson MJ, Ellingsen KE, McArdle BH. Multivariate dispersion as a measure of beta diversity. Ecol
Lett. 2006; 9: 683–93. https://doi.org/10.1111/j.1461-0248.2006.00926.x PMID: 16706913
41. Chao A, Chazdon RL, Colwell RK, Shen T-J. A new statistical approach for assessing similarity of spe-
cies composition with incidence and abundance data. Ecol Lett. 2005; 8: 148–159. https://doi.org/10.
1111/j.1461-0248.2004.00707.x
42. Colwell R. EstimateS: Statistical estimation of species richness and shared species from samples.
2013.
43. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical
Computing, Vienna, Austria [Internet]. 2017 [cited 20 Jan 2016].: https://www.r-project.org/
44. Jost L. Partitioning diversity into independent alpha and beta components. Ecology. 2007; 88: 2427–
2439. https://doi.org/10.1890/06-1736.1 PMID: 18027744
45. Burghardt KT, Tallamy DW. Not all non-natives are equally unequal: Reductions in herbivore??-diver-
sity depend on phylogenetic similarity to native plant community. Ecol Lett. 2015; 18: 1087–1098.
https://doi.org/10.1111/ele.12492 PMID: 26271480
46. Baselga A, Leprieur F. Comparing methods to separate components of beta diversity. Methods Ecol
Evol. 2015; 6: 1069–1079. https://doi.org/10.1111/2041-210X.12388
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 16 / 17
47. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara RB, et al. vegan: Community ecol-
ogy package [Internet]. 2016 [cited 3 May 2016] p. 12.: https://cran.r-project.org/package=vegan
48. Chao A, Jost L. Estimating diversity and entropy profiles via discovery rates of new species. Chisholm
R, editor. Methods Ecol Evol. 2015; 6: 873–882. https://doi.org/10.1111/2041-210X.12349
49. Chiu C-H, Chao A. Estimating and comparing microbial diversity in the presence of sequencing errors.
PeerJ. PeerJ Inc.; 2016; 4: e1634. https://doi.org/10.7717/peerj.1634 PMID: 26855872
50. Chao A, Ma KH, Hsieh TC, Chiu CH. SpadeR (Species-richness Prediction And Diversity Estimation in
R): an R package in CRAN. 2016.
51. Eiserhardt W, Svenning J, Kissling WD, Balslev H. Geographical ecology of the palms (Arecaceae):
determinants of diversity and distributions across spatial scales. Ann Bot. 2011; 108: 1391–1416.
https://doi.org/10.1093/aob/mcr146 PMID: 21712297
52. Salas-Morales SH, Meave JA. Elevational patterns in the vascular flora of a highly diverse region in
southern Mexico. Plant Ecol. 2012; 213: 1209–1220. https://doi.org/10.1007/s11258-012-0077-6
53. Lovett JC, Marshall AR, Carr J. Changes in tropical forest vegetation along an altitudinal gradient in the
Udzungwa Mountains National Park, Tanzania. Afr J Ecol. 2006; 44: 478–490. https://doi.org/10.1111/j.
1365-2028.2006.00660.x
54. Kluge J, Kessler M. Fern endemism and its correlates: Contribution from an elevational transect in
Costa Rica. Divers Distrib. 2006; 12: 535–545. https://doi.org/10.1111/j.1366-9516.2006.00231.x
55. Hawkins BA, Rodrı´guez MA´ , Weller SG. Global angiosperm family richness revisited: linking ecology
and evolution to climate. J Biogeogr. 2011; 38: 1253–1266. https://doi.org/10.1111/j.1365-2699.2011.
02490.x
56. Salazar L, Homeier J, Kessler M, Abrahamczyk S, Lehnert M, Kro¨mer T, et al. Diversity patterns of
ferns along elevational gradients in Andean tropical forests. Plant Ecol Divers. Taylor & Francis; 2015;
8: 13–24. https://doi.org/10.1080/17550874.2013.843036
57. Holeksa J, Saniga M, Szwagrzyk J, Dziedzic T, Ferenc S, Wodka M. Altitudinal variability of stand struc-
ture and regeneration in the subalpine spruce forests of the Pol’ana biosphere reserve, Central Slova-
kia. Eur J For Res. 2007; 126: 303–313. https://doi.org/10.1007/s10342-006-0149-z
58. Wang G, Zhou G, Yang L, Li Z. Distribution, species diversity and life-form spectra of plant communities
along an altitudinal gradient in the northern slopes of Qilianshan Mountains, Gansu, China. Plant Ecol.
Kluwer Academic Publishers; 2003; 165: 169–181. https://doi.org/10.1023/A:1022236115186
59. Warren SD, Holbrook SW, Dale DA, Whelan NL, Elyn M, Grimm W, et al. Biodiversity and the heteroge-
neous disturbance regime on military training lands. Restor Ecol. 2007; 15: 606–612. https://doi.org/10.
1111/j.1526-100X.2007.00272.x
60. Sa´nchez-Gonza´lez A, Lo´pez-Mata L. Clasificacio´n y ordenacio´n de la vegetacio´n del norte de la Sierra
Nevada, a lo largo de un gradiente altitudinal. An del Inst Biol Ser Bota´nica. 2003; 74: 47–71.
61. Sabatini FM, Burrascano S, Tuomisto H, Blasi C. Ground Layer Plant Species Turnover and Beta Diver-
sity in Southern-European Old-Growth Forests. Bond-Lamberty B, editor. PLoS One. 2014; 9: e95244.
https://doi.org/10.1371/journal.pone.0095244 PMID: 24748155
62. Yang Y, Shen Z, Han J, Zhongyong C. Plant diversity along the Eastern and Western slopes of Baima
Snow Mountain, China. Forests. 2016; 7: 89. https://doi.org/10.3390/f7040089
63. Akhtar N, Bergmeier E. Species richness, alpha and beta diversity of trees, shrubs and herbaceous
plants in the woodlands of swat, Pakistan. Pakistan J Bot. 2015; 47: 2017–2113.
64. Vellend M, Verheyen K, Flinn KM, Jacquemyn H, Kolb A, Van Calster H, et al. Homogenization of forest
plant communities and weakening of species-environment relationships via agricultural land use. J
Ecol. 2007; 95: 565–573. https://doi.org/10.1111/j.1365-2745.2007.01233.x
65. Brown S, Lugo AE. Tropical secondary forests. J Trop Ecol. 1990; 6: 1–32. https://doi.org/10.1017/
S0266467400003989
66. Valencia V, Naeem S, Garcı´a-Barrios L, West P, Sterling EJ. Conservation of tree species of late suc-
cession and conservation concern in coffee agroforestry systems. Agric Ecosyst Environ. 2016; 219:
32–41. https://doi.org/10.1016/j.agee.2015.12.004
67. Mittermeier R, Gil R, Hoffman M, Pilgrim J, Brooks TM, Mittermeier CG, et al. Hotspots revisited: Earth’s
biologically richest and most endangered terrestrial ecoregions. Boston: University of Chicago Press;
2005.
68. Go´mez-Mendoza L, Arriaga L. Modeling the effect of climate change on the distribution of oak and pine
species of Mexico. Conserv Biol. 2007; 21: 1545–1555. https://doi.org/10.1111/j.1523-1739.2007.
00814.x PMID: 18173478
Herb diversity along gradients of elevation and forest-use intensity
PLOS ONE | https://doi.org/10.1371/journal.pone.0182893 August 8, 2017 17 / 17