Article
Changes in Plant Functional Groups during
Secondary Succession in a Tropical Montane
Rain Forest
Kexin Fan
1,2
, Jing Tao
1,3
, Lipeng Zang
1,2
, Jie Yao
1,2
, Jihong Huang
1,2
, Xinghui Lu
1,2
,
Yi Ding
1,2
, Yue Xu
1,2
and Runguo Zang
1,2,
*
1
Key Laboratory of Forest Ecology and Environment, State Forestry and Grassland Administration, Research
Institute of Forest Ecology, Environment and Protection, Chinese Academy of Forestry, Beijing 100091, China;
[email protected] (K.F.); [email protected] (J.T.); [email protected] (L.Z.); [email protected] (J.Y.);
[email protected] (J.H.); [email protected] (X.L.); [email protected] (Y.D.); [email protected] (Y.X.)
2
Co-Innovation Center for Sustainable Forestry in Southern China, Nanjing Forestry University,
Nanjing 210037, China
3
Yunnan Institute of Forestry Inventory and Planning, Kunming 650051, China
* Correspondence: [email protected]; Tel.: +86-010-6288-9546
Received: 18 October 2019; Accepted: 10 December 2019; Published: 12 December 2019

 
Abstract:
Aggregating diverse plant species into a few functional groups based on functional
traits provides new insights for promoting landscape planning and conserving biodiversity in
species-diverse regions. Ecophysiological traits are the basis of the functioning of an ecosystem.
However, studies related to the identification of functional groups based on plant ecophysiological
traits in tropical forests are still scarce because of the inherent difficulties in measuring them. In this
study, we measured five ecophysiological traits: net photosynthetic capacity (A
max
), maximum
stomatal conductance (g
max
), water use efficiency (WUE), transpiration rate (Trmmol), and specific
leaf areas (SLA) for 87 plant species dominant in a chronosequence of secondary succession, using four
time periods (5 year-primary, 15 year-early, and 40 year-middle successional stages after clear
cutting and old growth) in the tropical montane rainforest on Hainan Island, China. These species
were grouped using hierarchical cluster analysis and non-metric multidimensional scaling. Finally,
the changes in the composition of functional groups and species richness along the chronosequence
were analyzed. Results showed that the plant species in the tropical montane rainforest could be
classified into eight distinct functional groups. The richness of functional groups was low during
the initial early stage and increased as the early and middle stages progressed, and then declined in
the late successional stage. The dominant functional groups in the primary stages had the highest
A
max
, g
max
, Trmmol, and SLA, as well as the lowest WUE, while those in the early and middle
successional stages had functional traits at a moderate level, and at the late stage they had the lowest
A
max
, g
max
, Trmmol, and SLA, and highest WUE. Our study showed that the diverse plant species in
the tropical montane rainforest could be grouped into a few functional groups according to major
ecophysiological traits, and the composition and relative abundance of different groups changed with
the successional dynamics of the forest ecosystem.
Keywords:
plant functional group; forest dynamics; secondary succession; tropical montane rainforest
1. Introduction
Plant functional groups (PFGs) are groups of species that share similar morphological and
physiological attributes, use similar resources, and play similar roles in a particular ecosystem [
1
].
PFGs can be divided into functional effect and functional response groups based on each group’s
Forests 2019, 10, 1134; doi:10.3390/f10121134 www.mdpi.com/journal/forests
Forests 2019, 10, 1134 2 of 14
function and/or adaptive responses to environmental variables in an ecosystem. The former refers to
groups with a similar effect on one or several ecosystem functions such as primary production and
nutrient cycling [
2
,
3
], and the latter refers to groups with a similar response to particular environmental
factors such as resource availability, disturbance, and drought stress [
4
–
6
]. Aggregating species into
functional groups is a common method useful for reducing the complexity of diverse ecosystems
(e.g., tropical rain forest communities) [
7
,
8
]. The identification of PFGs has been given priority in
international research agendas [
9
] for two reasons. First, in modeling vegetation under changing
climatic conditions, a widely recognized need exists to move away from single-leaf to whole-plant
approaches [
10
]. Second, in doing so, the enormous complexity of individual species and populations
needs to be summarized into a relatively small number of general and recurrent patterns [
11
,
12
].
Research studies on the characteristics that define the main functional groups of tropical trees and their
relationship to forest dynamics and regeneration have proliferated since the mid-1970s [13–15].
PFGs can be defined based on species functional traits, such as life form, maximum potential
height, successional status, and seed dispersal pattern [
15
–
17
], and on species’ associations with
particular environmental factors such as light, disturbance [
18
,
19
] or on the ecological strategy of
species resources use (e.g., competitors (C), stress-tolerators (S), and ruderals (R), C-S-R strategy) [
20
,
21
].
PFGs are most commonly defined based on functional traits. The distribution of functional traits in
a community and the magnitude of their differences among species can shed light on the relative
influence of environmental filtering and competition [
22
]. Functional traits can be defined as any
attributes that have a potentially significant influence on establishment, survival, and fitness of a
species [
21
,
23
], and plant ‘functional traits’ are considered to reflect adaptations to variations in the
physical environment as well as ecophysiological and/or evolutionary trade-offs among different
functions within a plant [
24
]. Westoby [
25
] proposed the leaf-height-seed scheme, to define PFGs based
on specific leaf area (SLA), canopy height, and seed mass. The output of the PFG method agrees very
well with field studies, indicating that a particular functional trait used for PFG identification in this
method can fully reflect the survival strategy of plants [20,25].
Based on the definition of a functional group, the most appropriate way to determine the PFG of a
species is to group species based on the role of various species in ecosystem processes (e.g., the cycling
of carbon, water, and nitrogen). The plant–atmosphere interactions that are of prime interest for
regional and global simulations are the carbon, water, and nitrogen cycles. These cycles are strongly
linked to the ecophysiological process of different groups of plant species. Although the body of
literature available concerning PFGs is substantial, studies focused on tropical forest vegetation are
still scarce [
26
] and few studies were found addressing the identification of functional groups based
on ecophysiological attributes, especially in tropical forest ecosystems. Two of the most important
reasons why such studies are rarely conducted are that measuring the ecophysiological traits for
numerous individual species is difficult, and it is hard to control the comparability between individual
plant species. Furthermore, different PFGs are expected to play different roles in ecosystem processes.
Therefore, the identification and the estimation of their abundance are relevant to the assessment of
ecosystem function [27–29].
Plant communities recover from disturbances through ecological succession, a process that implies
sequential changes occur in the community attributes over time [
30
]. Although many studies have
been conducted to aid in understanding the processes of secondary forest succession in the tropics, few
studies have applied concepts related to PFGs to plant community succession in tropical forests. Letcher
et al. found that successional habitat specialization is a conserved trait for tropical forests, which
associate with many different pioneer lineages and a concomitant diversity of functional traits [
31
].
To assess plant community succession, an assessment of the changes in PFGs based on functional traits
during succession is necessary. Through the use of technologies designed to detect changes in plant
functional traits, we can also develop tools for inferring the functions and successional status of plant
communities [
32
]. This is especially important for tropical systems, where functional recovery has
been poorly explored.
Forests 2019, 10, 1134 3 of 14
In the present study, we attempted to define functional groups of a tropical montane rain forest
(TMRF) on Hainan Island, China, based on five ecophysiological traits: net photosynthetic capacity
(A
max
,
µ
mol m
−2
s
−1
), maximum stomatal conductance (g
max
, mol m
−2
s
−1
), water use efficiency
(WUE,
µ
mol mol
−1
), transpiration rate (Trmmol, mmol m
−2
s
−1
), and specific leaf area (SLA, m
2
kg
−1
).
These traits are directly related to plant–atmosphere interactions as well as plant resource acquisition
and use. A
max
refers to the net photosynthetic rate of a mature leaf under saturated irradiances
that directly indicate the interception and assimilation of resources. g
max
, WUE, and Trmmol are
traits that are directly related to the plant–atmosphere interactions, and plants have to modulate
their photosynthetic and transpiration rates in differing environmental conditions by adjusting their
stomatal conductance. SLA is the ratio between leaf area and leaf dry mass, and it is related to resource
interception and use. This ratio gives a measure of a plant’s investment into photosynthetic processes
as well as into a plant’s participation in the carbon and water cycles
PFGs were identified using standard multivariate analysis techniques based on these five
ecophysiological traits. Then we explored the variation of functional group compositions during
ecological succession. The composition of PFGs and the dynamics along differing successional stages
of TMRF were analyzed based on identified functional groups. The objectives of the study were: 1)
To aggregate the diverse tree species in the TMRF of Hainan island into a few functional groups based
on the field measured functional traits so that future studies related to the ecosystem functioning and
their simulation at different scales could be simplified or made more convenient. 2) To understand the
change of functional group composition during the process of succession to provide some theoretical
bases for the sustainable management and effective restoration of tropical montane forests on Hainan
Island, southern China.
2. Materials and Methods
2.1. Study Sites and Sampling
The research site was located in the TMRF near the Jianfengling Long-term Research Station of
Tropical Forest Ecosystems (18
◦
20
0
–18
◦
57
0
N, 108
◦
41
0
–109
◦
12
0
E) (JRSTF, hereafter) at approximately
800–960 m elevation in Southwest Hainan, China. The region’s tropical monsoon climate has a distinct
dry and wet season, a mean annual temperature at the study site of 19.7
◦
C, a mean annual rainfall of
2651.6 mm, and a mean annual potential evaporation of 1303.7 mm [
33
]. The study site has a lateritic
yellow soil [34].
Field measurements were carried out within areas of the TMRF located close to the JFSTF by
establishing sampling plots in secondary stands with differing times after clear cutting and in old
growth stands. For convenience, four successional stages were arbitrarily defined based on time since
harvest: primary (stage I), early (II), middle (III), and late successional or old-growth forest stages (IV)
with time since harvest = 5 years, 15 years, 40 years, and old growth (no recent harvest), respectively.
A total of 70 plots were previously delimited: three were 10
×
10 m
2
with abandonment ages of 5 years,
30 were 10
×
10 m
2
with abandonment ages of 15 years, 12 were 10
×
10 m
2
with abandonment ages
of 40 years, and 25 were 20
×
20 m
2
old growth sites (Table 1). At the beginning of the study, all
free-standing plants (trees, shrubs, and herbs) in each plot were recorded, identified, mapped and
measured to the lowest possible taxonomic level between November, 2005 and March, 2006. Based on
the results, we selected 87 of the most abundant and representative vascular plant species in the TMRF
of JRSTF and measured ecophysiological traits.
Forests 2019, 10, 1134 4 of 14
Table 1. Sampling plots for the stands of different successional stages.
Successional Stage Time Years Since Harvest No. of Plots
Plot Size (m
2
)
I Primary 5 3 10 × 10
II Early 15 30 10 × 10
III Middle 40 12 10 × 10
IV Late Old growth 25 20 × 20
2.2. Measurement of Ecophysiological Traits
Measurements were made on 2–6 individuals for each species and 3–5 leaves per individual in
low- to mid-crown positions of trees (
≤
5 m in height). All measurements were taken on fully spread
mature leaves and we spent 3–5 min for each leaf in an attempt to ensure that photosynthetic induction
had already occurred. Leaf gas-exchange rates were measured in the field in April and May 2006
using an LI-6400 portable photosynthesis measurement system (LI-6400, Li-Cor, Lincoln, NE, USA);
A
max
, g
max
, and Trmmol were measured under ambient CO
2
concentrations and relative humidity [
35
].
The saturating irradiances (1000
µ
mol m
−2
s
−1
) were obtained using an artificial LED light source
(LI-6400). The saturating irradiances were identified by the light response curve of light-demanding
species, such as Sapium discolor. The measurements were carried out over 9:00–12:30 and 14:00–15:30
on clear days to minimize the influence of depressed stomatal conductance, which may lead to a
reduction in CO
2
assimilation.
Leaves were harvested after measurements were obtained and single-sided leaf areas were
measured for each fresh leaf with a leaf area meter (LI-3100, LI-COR and then leaves were oven dried to
a constant weight at 65
◦
C. These data were used for the calculation of SLA. Averaged field parameters
of each species or individual were used for statistical analysis. The ecophysiological traits related to
photosynthesis of major tree species of differing recovery stages were measured in April–May, 2006.
2.3. Data Analysis
Ecophysiological traits of 87 species were measured in stands of different age, as harvest and
averages of parameters for individual species were used for statistical analysis. PFGs were defined
based on five functional traits by applying hierarchical cluster analysis. Data were log transformed to
meet normality prior to the hierarchical cluster analysis. The relative Euclidean distance and Ward
Linkage methods were used in the analysis, and cluster analysis using other distances and linkage
methods was also carried out; the results were generally in agreement with each other.
MRPP (multi-response permutation procedures) are non-parametric procedures for testing the
hypothesis of no difference between two or more groups of entities. The MRPP algorithm first calculates
all pairwise distances in the entire dataset, then calculates
δ
. It then permutes the sampling units and
their associated pairwise distances, and recalculates them based on the permuted data. It repeats the
permutation step 999 times. The significance test is the fraction of permuted deltas that are less than
the observed delta, with a small sample correction. The function also calculates the change-corrected
within-group agreement A = 1 –
δ
E(
δ
), where E(
δ
) is the expected
δ
assessed as the average of
dissimilarities. A value greater than 0 indicates that the difference between groups is greater than the
difference within the groups, and less than 0 indicates that the difference within the groups is greater
than the difference between groups.
We used the function “agnes” in the {cluster} R package [
36
], available on CRAN (https://svn.r-
project.org/R-packages/trunk/cluster), to accomplish the Hierarchical Cluster analysis. Multi-Response
Permutation Procedures (MRPP) were accomplished using the function “MRPP” in the {vegan} R
package [37], available on CRAN (https://cran.r-project.org, https://github.com/vegandevs/vegan).
Forests 2019, 10, 1134 5 of 14
3. Results
3.1. Plant Functional Groups of the TMRF
The 87 main species in the TMRF in Jianfengling, Hainan Island were aggregated into eight PFGs
based on five functional traits (A
max
, g
max
, WUE, Trmmol, and SLA) using hierarchical cluster analysis
(Figure 1). The results of MRPP showed a significant difference between the groups (p = 0.001) and the
difference between groups is greater than the difference within the groups (A = 0.5197), which indicated
that the functional groups classified by the ecophysiological traits were distinct groups.
Forests 2019, 10, x FOR PEER REVIEW 5 of 14
3. Results
3.1. Plant Functional Groups of the TMRF
The 87 main species in the TMRF in Jianfengling, Hainan Island were aggregated into eight PFGs
based on five functional traits (A
max, gmax, WUE, Trmmol, and SLA) using hierarchical cluster analysis
(Figure 1). The results of MRPP showed a significant difference between the groups (p = 0.001) and
the difference between groups is greater than the difference within the groups (A = 0.5197), which
indicated that the functional groups classified by the ecophysiological traits were distinct groups.
Figure 1. Dendrogram depicting functional groups derived from data of 87 species. Variables used to
create cluster include net photosynthetic capacity (A
max), maximum stomatal conductance (gmax),
water use efficiency (WUE), transpiration rate (Trmmol), and specific leaf areas (SLA). Plant
functional groups (PFGs).
*Note:
Sp1
Blastus cochinchinensis, Lour. Sp30
Syzygium hancei Merr. et
Perry
Sp59 Antidesma montanum Bl.
Sp2
Lithocarpus longipedicellatus
(Hick. et A. Camus) A.
Camus
Sp31
Phoebe hungmaoensis S.
Lee
Sp60
Evodia glabrifolia (Champ. ex
Benth.) Huang
Sp3
Elaeocarpus petiolatus (Jack)
Wall. ex Kurz
Sp32 Illicium ternstroemioides Sp61 Helicia formosana
Sp4
Acronychia pedunculata (L.)
Miq.
Sp33
Litsea elongata (Wall. ex
Nees) Benth. et Hook. f.
Sp62
Eriobotrya deflexa (Hemsl.) Nakai f.
koshunensis (Kanehira et Sasaki)
Li
Sp5 Ormosia balansae Drake Sp34
Lasianthus koi Merr. et
Chun
Sp63
Sapium discolor (Champ. ex Benth.)
Muell. Arg.
Sp6
Chassalia curviflora Thwaites
var. longifolia
Sp35
Endospermum chinense
Benth.
Sp64
Ampelocalamus actinotrichus (Merr.
et Chun) S. L. Chen. T. H. Wen et
G. Y. Sheng
Sp7
Neolitsea oblongifolia Merr. et
Chun
Sp36
Symplocos paniculata
(Thunb.) Miq.
Sp65 Calamus simplicifolius C. F. Wei
Sp8
Mallotus hookerianus (Seem.)
Muell. Arg.
Sp37 Dasymaschalon rostratum Sp66
Rhaphiolepis indica (L.) Lindl. ex
Ker
*Note: Sp1
Blastus cochinchinensis, Lour.
Sp30
Syzygium hancei Merr. et
Perry
Sp59
Antidesma montanum Bl.
Sp2
Lithocarpus longipedicellatus
(Hick. et A. Camus) A.
Camus
Sp31
Phoebe hungmaoensis S. Lee
Sp60
Evodia glabrifolia (Champ. ex
Benth.) Huang
Sp3
Elaeocarpus petiolatus (Jack)
Wall. ex Kurz
Sp32
Illicium ternstroemioides
Sp61
Helicia formosana
Sp4
Acronychia pedunculata (L.)
Miq.
Sp33
Litsea elongata (Wall. ex
Nees) Benth. et Hook. f.
Sp62
Eriobotrya deflexa (Hemsl.)
Nakai f. koshunensis
(Kanehira et Sasaki) Li
Sp5 Ormosia balansae Drake
Sp34
Lasianthus koi Merr. et Chun
Sp63
Sapium discolor (Champ. ex
Benth.) Muell. Arg.
Sp6
Chassalia curviflora Thwaites
var. longifolia
Sp35
Endospermum chinense Benth.
Sp64
Ampelocalamus actinotrichus
(Merr. et Chun) S. L. Chen.
T. H. Wen et G. Y. Sheng
Sp7
Neolitsea oblongifolia Merr. et
Chun
Sp36
Symplocos paniculata (Thunb.)
Miq.
Sp65
Calamus simplicifolius C. F.
Wei
Sp8
Mallotus hookerianus (Seem.)
Muell. Arg.
Sp37
Dasymaschalon rostratum
Sp66
Rhaphiolepis indica (L.) Lindl.
ex Ker
Sp9
Lasianthus chinensis (Champ.)
Benth.
Sp38
Lasianthus hirsutus (Roxb.)
Merr.
Sp67
Olax wightiana Wall. ex
Wight et Arn.
Sp10
Gironniera subaequalis
Planch.
Sp39
Sterculia lanceolata Cav.
Sp68
Pygeum topengii Merr.
Sp11
Mastixia pentandra Blume
subsp. cambodiana (Pierre)
Matthew
Sp40
Symplocos lancifolia Sieb. et
Zucc.
Sp69
Miscanthus floridulus (Lab.)
Warb. ex Schum. et Laut.
Figure 1. Cont.
Forests 2019, 10, 1134 6 of 14
Sp12 Winchia calophylla A. DC.
Sp41
Psychotria rubra (Lour.) Poir.
Sp70
Trema angustifolia (Planch.)
Bl.
Sp13
Lithocarpus brachystachyus
Chun
Sp42
Castanopsis fissa (Champ. ex
Benth.) Rehd. et Wils
Sp71
Neolitsea phanerophlebia Merr.
Sp14
Parapyrenaria multisepala
(Merr. et Chun) Chang
Sp43
Cyclobalanopsis blakei (Skan)
Schott.
Sp72
Dillenia pentagyna Roxb.
Sp15 Ardisia nervosa Walker
Sp44
Ormosia semicastrata Hance f.
litchifolia How
Sp73
Castanopsis chinensis Hance
Sp16
Polyosma cambodiana
Gagnep.
Sp45
Reevesia pubescens Mast.
Sp74
Artocarpus styracifolius Pierre
Sp17 Eupatorium odoratum L.
Sp46
Pithecellobium lucidum Benth.
Sp75
Glochidion coccineum
(Buch.-Ham.) Muell. Arg.
SP18 Elaeocarpus dubius A. DC.
Sp47
Heliciopsis lobata (Merr.)
Sleum.
Sp76
Pinanga discolor Burret
Sp19
Diplospora dubia (Lindl.)
Masam.
Sp48
Eurya groffii Merr.
Sp77
Melastoma candidum D. Don
Sp20 Ervatamia hainanensis Tsiang
Sp49
Machilus salicina Hance
Sp78
Olea dioica Roxb.
Sp21
Lasianthus curtisii King et
Gamble
Sp50
Ardisia quinquegona Bl.
Sp79
Cinnamomum burmanni
(Nees et T.Nees) Blume
Sp22
Lindera kwangtungensis
(Liou) Allen
Sp51
Melastoma sanguineum Sims
Sp80
Symplocos pseudobarberina
Gontsch.
Sp23
Lithocarpus fenestratus
(Roxb.) Rehd.
Sp52
Ilex pubilimba Merr. et Chun
Sp81 Alseodaphne hainanensis Merr.
Sp24
Lindera robusta (Allen) H. P.
Tsui
Sp53
Beilschmiedia laevis Allen
Sp82
Canthium dicoccum (Gaertn.)
Teysmann et Binnedijk
Sp25 Helicia hainanensis
Sp54
Syzygium jambos (L.) Alston
Sp83
Castanopsis tonkinensis Seem.
Sp26
Nephelium topengii (Merr.) H.
S. Lo
Sp55
Cryptocarya chingii Cheng
Sp84
Linociera ramiflora (Roxb.)
Wall. ex G. Don
Sp27
Adinandra hainanensis Hayata
Sp56
Pentaphylax euryoides Gardn.
et Champ.
Sp85
Cryptocarya chinensis (Hance)
Hemsl.
Sp28
Drypetes indica (Muell. Arg.)
Pax et Hoffm
Sp57
Evodia lepta
Sp86
Ardisia crenata Sims
Sp29
Lithocarpus fenzelianus
A.Camus
Sp58
Prismatomeris tetrandra
(Roxb.) K. Schum.
Sp87
Thysanolaena maxima (Roxb.)
Kuntze)
Figure 1.
Dendrogram depicting functional groups derived from data of 87 species. Variables used
to create cluster include net photosynthetic capacity (A
max
), maximum stomatal conductance (g
max
),
water use efficiency (WUE), transpiration rate (Trmmol), and specific leaf areas (SLA). Plant functional
groups (PFGs).
The composition and functional traits of different PFGs varied greatly (Table 2). PFG1 included
one species: Eupatorium odoratum L., an invasive ruderal in deforested areas in tropical forest areas.
Their main functional traits were comparatively high A
max
, g
max
, Trmmol, SLA, and low WUE. PFG2
also included one species: Evodia glabrifolia (Champ. ex Benth.) Huang, a fast-growing species,
had comparatively high A
max
, Trmmol, and SLA but moderate g
max
and low WUE. The group PFG3
was mainly composed of shrubs and tall grasses of the Poaceae, along with pioneer species with higher
than average A
max
, moderate g
max
, Trmmol, WUE, and SLA. The group PFG4 was mainly composed of
evergreen trees and shrubs, such as Elaeocarpus petiolatus (Jack) Wall. ex Kurz and Adinandra hainanensis,
characterized by comparatively lower g
max
, moderate A
max
, WUE, Trmmol, and lower SLA. PFG5 and
PFG6 included most of the tree species in the genera and families typical of the TMRF. Representative
species such as Cryptocarya chinensis (Hance) Hemsl., Diplospora dubia (Lindl.) Masam, and Lithocarpus
brachystachyus Chun, were mostly common species in the middle successional stages. Functional traits
of these two groups were low A
max
, g
max
, Trmmol, and moderate SLA, while the WUE of PFG5 was
generally higher in contrast with PFG6. The groups PFG7 and PFG8 were mainly composed of late
successional species, such as Dillenia pentagyna Roxb. and Nephelium topengii (Merr.) H. S. Lo. They had
Forests 2019, 10, 1134 7 of 14
similar functional traits such as low A
max
, g
max
, and Trmmol values and higher WUE, while the SLA
of PFG8 were relatively higher than that of PFG7.
Table 2.
The distributions of the five traits in each plant functional group (Number of species in each
PFGs indicated in brackets).
PFGs Growth Form Characteristic of Functional Traits Representative Species
PFG1(1) Perennial herbs
High A
max
(>10), High g
max
(>10),
Low WUE (<10), High Trmmol (>5),
High SLA (>20).
Eupatorium odoratum L.
PFG2(1) Trees
High A
max
(>10), Middle g
max
(1–10), low WUE (<10), High
Trmmol (>5), High SLA (>20).
Evodia glabrifolia (Champ. ex Benth.)
Huang
PFG3(5) Shrubs and herbs
High A
max
(>10), Middle g
max
(1–10), Middle WUE (10–20), High
Trmmol (>5), Middle SLA (10–20).
Sapium discolor (Champ. ex Benth.) Muell.
Arg.
Thysanolaena maxima
PFG4(4)
Evergreen Trees and
shrubs
Middle A
max
(5–10), Low g
max
(<1),
Middle WUE (10–20), Middle
Trmmol (2–5), Middle SLA (10–20).
Elaeocarpus petiolatus (Jack) Wall. ex Kurz
Adinandra hainanensis
PFG5(28)
Shrubs, Trees and Liana
Low A
max
(<5), Low g
max
(<1),
High WUE (>20), Low Trmmol (<2),
Middle SLA (10–20).
Cryptocarya chinensis (Hance) Hemsl.
Diplospora dubia (Lindl.) Masam.
PFG6(5)
Evergreen Trees and
Shrubs
Low A
max
(<5), Low g
max
(<1), Low
WUE (<10), Low Trmmol (<2),
Middle SLA (10–20).
Lithocarpus brachystachyus Chun
Cyclobalanopsis blakei (Skan) Schott.
PFG7(34)
Shrubs or small trees
Low A
max
(<5), Low g
max
(<1),
High WUE (>20), Low Trmmol (<2),
Middle SLA (10–20).
Dillenia pentagyna Roxb. Nephelium
topengii (Merr.) H. S. Lo
PFG8(9) Shrubs or saplings
Low A
max
(<5), Low g
max
(<1),
High WUE (>20), Low Trmmol (<2),
High SLA (>20).
Blastus cochinchinensis Lour. Ervatamia
hainanensis Tsiang
3.2. Analysis of Species Richness and Occurrence Frequency of PFGs
The occurrence frequency in the plots at different times after harvest for the different functional
groups of TMRF was analyzed (Figure 2. table S). The result showed that almost all high occurrence
species were present in PFG5 and PFG7.
Forests 2019, 10, x FOR PEER REVIEW 7 of 14
Table 2. The distributions of the five traits in each plant functional group (Number of species in each
PFGs indicated in brackets).
PFGs Growth form Characteristic of functional traits Representative species
PFG1(1)
Perennial
herbs
High A
max (＞10), High gmax (＞10),
Low WUE (＜10), High Trmmol (＞
5), High SLA (＞20).
Eupatorium odoratum L.
PFG2(1) Trees
High A
max (＞10), Middle gmax (1–10),
low WUE (＜10), High Trmmol (＞
5), High SLA (＞20).
Evodia glabrifolia (Champ. ex
Benth.) Huang
PFG3(5)
Shrubs and
herbs
High A
max (＞10), Middle gmax (1–10),
Middle WUE (10–20), High Trmmol
(＞5), Middle SLA (10–20).
Sapium discolor (Champ. ex
Benth.) Muell. Arg.
Thysanolaena maxima
PFG4(4)
Evergreen
Trees and
shrubs
Middle A
max (5–10), Low gmax (＜1),
Middle WUE (10–20), Middle
Trmmol (2–5), Middle SLA (10–20).
Elaeocarpus petiolatus (Jack)
Wall. ex Kurz
Adinandra hainanensis
PFG5(28)
Shrubs，Trees
and Liana
Low A
max (＜5), Low gmax (＜1), High
WUE (＞20), Low Trmmol (＜2),
Middle SLA (10–20).
Cryptocarya chinensis (Hance)
Hemsl.
Diplospora dubia (Lindl.)
Masam.
PFG6(5)
Evergreen
Trees and
Shrubs
Low A
max (＜5), Low gmax (＜1), Low
WUE (＜10), Low Trmmol (＜2),
Middle SLA (10–20).
Lithocarpus brachystachyus
Chun Cyclobalanopsis blakei
(Skan) Schott.
PFG7(34)
Shrubs or
small trees
Low A
max (＜5), Low gmax (＜1), High
WUE (＞20), Low Trmmol (＜2),
Middle SLA (10–20).
Dillenia pentagyna Roxb.
Nephelium topengii (Merr.) H.
S. Lo
PFG8(9)
Shrubs or
saplings
Low A
max (＜5), Low gmax (＜1), High
WUE (＞20), Low Trmmol (＜2),
High SLA (＞20).
Blastus cochinchinensis Lour.
Ervatamia hainanensis Tsiang
3.2. Analysis of Species Richness and Occurrence Frequency of PFGs
The occurrence frequency in the plots at different times after harvest for the different functional
groups of TMRF was analyzed (Figure 2. table S). The result showed that almost all high occurrence
species were present in PFG5 and PFG7.
Figure 2. The occurrence frequency of 87 species belonging to different plant functional groups. The
87 species corresponding to each symbol in Figure 2 are listed in Table S.
3.3. PFGs Richness and Functional Composition in Different Successional Stages
The richness of PFGs varied significantly within differing successional stages (χ2 = 213.774, df =
21, p < 0.001). The statistical analysis showed that the richness of PFGs was lowest in the primary
Figure 2.
The occurrence frequency of 87 species belonging to different plant functional groups. The 87
species corresponding to each symbol in Figure 2 are listed in Table S1.
3.3. PFGs Richness and Functional Composition in Different Successional Stages
The richness of PFGs varied significantly within differing successional stages (
χ
2 = 213.774, df = 21,
p < 0.001). The statistical analysis showed that the richness of PFGs was lowest in the primary stage.
The dominant PFGs of this stage were PFG3 and PFG4 (high relative abundance), while PFG5, PFG6,
PFG7, and PFG8 were not found. PFG richness increased rapidly after the restoration of the forest
Forests 2019, 10, 1134 8 of 14
environment and PFG richness of the early and middle successional stages were highest; the dominant
PFGs of these stages were PFG5 and PFG7. The richness of PFGs decreased in the late successional
stage as a result of the disappearance of pioneer species; the dominant PFGs of this stage were PFG5
and PFG7 (Figure 3).
Forests 2019, 10, x FOR PEER REVIEW 8 of 14
stage. The dominant PFGs of this stage were PFG3 and PFG4 (high relative abundance), while PFG5,
PFG6, PFG7, and PFG8 were not found. PFG richness increased rapidly after the restoration of the
forest environment and PFG richness of the early and middle successional stages were highest; the
dominant PFGs of these stages were PFG5 and PFG7. The richness of PFGs decreased in the late
successional stage as a result of the disappearance of pioneer species; the dominant PFGs of this stage
were PFG5 and PFG7 (Figure 3).
Figure 3. PFG richness and functional composition in different successional stages. Table 1 lists the
successional stages.
3.4. Relative Abundance of PFGs along the Successional Stages
Generally, the relative abundance of PFG1, PFG2, PFG3, and PFG4 decreased as time progressed
through the successional stages, and that of PFG5 peaked at late stage while the relative abundance
of PFG6, PFG7, and PFG8 increased during the successional stages (Figure 4). The PFG1 and PFG2
occurred only in the primary stage, while PFG3 and PFG4 were dominant in the primary stage. The
species found in these four stages were most found in the early and middle successional stages. The
relative abundance of PFG5, PFG6, PFG7, and PFG8 increased gradually as time progressed through,
PFG5, PFG6, and PFG7 dominated PFGs of the late successional stages, while PFG8 dominated in the
middle stage.
Figure 3.
PFG richness and functional composition in different successional stages. Table 1 lists the
successional stages.
3.4. Relative Abundance of PFGs along the Successional Stages
Generally, the relative abundance of PFG1, PFG2, PFG3, and PFG4 decreased as time progressed
through the successional stages, and that of PFG5 peaked at late stage while the relative abundance
of PFG6, PFG7, and PFG8 increased during the successional stages (Figure 4). The PFG1 and PFG2
occurred only in the primary stage, while PFG3 and PFG4 were dominant in the primary stage.
The species found in these four stages were most found in the early and middle successional stages.
The relative abundance of PFG5, PFG6, PFG7, and PFG8 increased gradually as time progressed
through, PFG5, PFG6, and PFG7 dominated PFGs of the late successional stages, while PFG8 dominated
in the middle stage.
Forests 2019, 10, 1134 9 of 14
Figure 4.
Box plot of relative abundance of PFGs in difference successional stages of tropical montane
rain forest (TMRF). Table 1 lists the successional stages.
4. Discussion
The concept of PFGs has been regarded as a framework for predicting ecosystem response to
environmental changes on a global scale, without detailed information about each species [
8
,
38
].
Therefore, research related to PFGs is strongly tied to the modeling of ecosystem processes [
39
].
The challenge in functional grouping is how to find the most appropriate indices and methods to use
for the aggregation of PFGs. PFGs are commonly defined based on a series of related quantitative
traits [
39
–
41
]. Theoretically, one can properly define PFGs based on ecophysiological traits related
to plant–atmosphere interactions. A study of shrub vegetation in Florida showed that PFGs based
on physiological traits were spatially and temporally robust [
42
]. A study in the eastern Amazonian
region demonstrated that ecophysiological traits were significantly different between functional groups.
Additionally, PFGs identified by combining hard traits (ecophysiological traits) with soft traits (e.g.,
leaf chemical, anatomical characteristics) were highly consistent with groups based on life forms [
26
].
One can reasonably, practically, and feasibly define functional groups using ecophysiological traits [
43
].
However, measuring large sets of ecophysiological parameters in the field for grouping is time intensive
and expensive, making it necessary to find one or a set of easy-to-measure soft traits that can be used
as substitutes of ecophysiological traits, and the grouping result based on these indicators are expected
to agree with that of the ecophysiological traits. A study of shrub habitat in Florida [
44
] 50showed
that groups defined by life forms were similar to ecophysiologically-based groups. We attempted to
identify PFGs based on a set of ecophysiological traits related to photosynthetic parameters obtained
from field measurements. These traits are indices of plant–atmosphere interaction (gas and water
exchange), and are therefore directly related to plant use of light, water, and other resources. Thus,
this grouping is expected to enable successful prediction of responses to environmental variations in
climate, atmospheric chemistry, or site disturbances.
Community structure can strongly influence ecosystem function [
44
], and the functions of species
in ecosystems are highly related to the size and density of the individual species [
45
]. According to
L
ó
pez-Mart
í
nez’s study [
46
], variation in floristic composition was greatest for shrubs and lowest
for trees. We found that most of the species with higher occurrence frequency were smaller than
the relatively rare species in the same group. For example, in group 7, Heliciopsis lobata (Merr.)
Sleum, Cryptocarya chingii Cheng, and Antidesma montanum
Bl.
occurred only in the middle and
Forests 2019, 10, 1134 10 of 14
late successional stages were larger than relatively found shrubs like Lasianthus chinensis (Champ.)
Benth., Symplocos paniculata (Thunb.) Miq., and Pentaphylax euryoides. Meanwhile, we can interpret
varied community characteristics along the succession gradient by studying key functional traits of
dominant species. Similarly, community characteristics can be interpreted by studying their functional
compositions and the main characteristics of functional groups [
8
]. We attempted to study succession
by comparing sites of various ages in the montane rain forest in Jianfengling using an approach called
a “space for time substitution.” Our results indicate that the richness of PFGs is lowest in the primary
successional stage, increases rapidly after the restoration of a forested environment, and peaks during
the early and middle stages then decreases in the late stage. The results indicate that the influence of
an established species in its habitat is the main force driving succession [
47
] and the establishment
and growth of new species strongly depends on particular physical environments created by earlier
species [
48
]. The establishment of new species will cause changes to the existing habitats. Luc
í
a et al.
found that although almost all species can be established along the complete environmental gradient,
species that dominated early in succession had acquisitive functional traits, while those that dominated
at later successional ages and hills showed more conservative traits [
24
].The establishment of PFGs
during the primary stage facilitates the establishment and growth of new PFGs, which conversely
inhibit the survival of former; the decrease in functional richness during the late successional stage is
probably caused mainly by the heavily shaded original forest environment, which is unfavorable for
the growth and regeneration of light-demanding pioneer species. This constitutes a foundation for
further exploration of the application of remote sensing technologies to the study of tropical succession.
Community changes during succession include not only the increase in functional group richness
but also changes in the composition of PFGs. Statistical results show that the relative abundance of
PFG1, PFG2, PFG3, and PFG4 decrease as ecological succession progresses and that PFG5 peaks at the
late stage while the relative abundances of PFG6, PFG7, and PFG8 increase as succession continues.
Analysis shows that the dominant PFGs of the primary stage are those with higher photosynthetic
capacity, fast-growing rates, and low water use efficiency. These PFGs appear to be associated with
relative high values of A
max
, Trmmol, and SLA, as well as with lower values of WUE. This particular
combination of traits generally occurs as a response to a particular combination of environmental
conditions: low water availability, high light availability, and high temperature [
49
]. Dominant PFGs
of the early and middle stages were those with moderate photosynthetic capacity and WUE and
comparatively high growth rates; the main PFGs of the late successional stage were those with low
A
max
, high WUE, and low growth rates.
We found that the SLA, which had been above average throughout the succession, rose above
20 in group 8 again, possibly because group 8 was mainly composed of shade tolerant species in
the middle and late succession. The facilitation and inhibition model of succession proposes that
early successional species are more vulnerable to variety of physical and biological factors that cause
mortality. Thus, the relative abundance of PFGs with pioneer species (gap species) decreases during
succession, while that of PFGs, mainly composed of shade-tolerant species, increases during ecological
succession. Surprisingly, in PFG3, a shrub-like herbaceous C4 plant thysanolaena maxima, which looks
like bamboo, is grouped with three other shrub species. This may be part of the reason, as it may
have some important leaf trait differences with shrubs. This difference points to a potential problem,
in that using leaf traits to define PFGs is somewhat limited. It may be more meaningful and useful
to add dimensions to the leaf traits we studied such as heights and seeds; these could perhaps be
used in addition to leaf traits to define them [
42
]. Zhang and Zang [
13
] classified tropical forest
vegetation of Hainan Island, China into six functional groups based on successional status and potential
maximum height. There is still much room for research and in-depth exploration on leaf traits in
the future. Variation of dominant PFGs in differing stages indicates that what really determines the
occurrence and relative abundance of PFGs in a particular community is the diversity of microhabitats
and the abundance of shaded and gap habitats in the meta community over long periods of time.
The replacement of species will cause variations in the functional groups present and the variations
Forests 2019, 10, 1134 11 of 14
in the richness and relative abundances of PFGs and in the progress of ecological succession of a
community are cause and effect related.
5. Conclusions
Based on five field measured ecophysiological traits, major plant species in the TMRF of
Jianfengling, Hainan Island, were grouped into eight PFGs by applying quantitative techniques.
MRPP tests indicated that the identified PFGs are distinct groups with high heterogeneity between
groups but high homogeneity within each group. The richness of PFGs and functional groups
composition of TMRF in the Jianfengling areas varied significantly during ecological succession.
Species composition and PFGs richness were relatively low in the primary stage as the closed canopy
forested environment was lacking, and because the plant community of this stage was dominated by
fast-growing pioneer groups with high A
max
and low WUE. The richness of PFGs increased rapidly
after establishment of a forested environment and reached a peak in the middle successional stages.
Additionally, the related dominant PFGs of both these stages were faster-growing species with moderate
A
max
and lower WUE. The richness of PFGs decreased in the late successional stage because the dense
canopy limited the regeneration of light-demanding species, and the dominant PFGs of this stage were
slow-growing species with low A
max
, but high WUE. A functional approach should be incorporated
as a regular descriptor of forest succession because it provides a richer understanding of vegetation
dynamics than is offered by either the floristic or structural approach alone.
Tropical montane rainforests are one of the most highly diverse types of forests in the world,
and they are considered extremely vulnerable to disturbance and climate changes. Our study, which
aggregated the diverse plant species into a few functional groups based on ecophysiological functional
traits, provided new insights elucidating the structure and function of tropical forest ecosystems [
13
].
It is also an effective way of making further landscape planning, improving forest conservation
planning and conserving biodiversity in species-rich tropical forests [50].
Supplementary Materials:
The following are available online at http://www.mdpi.com/1999-4907/10/12/1134/s1,
Figure S1: Box plot for the distributions of the five traits in each plant functional group, Table S1: The 87 species
correspongding to each symbol.
Author Contributions:
R.Z. and J.H. designed this study; K.F., J.T., and R.Z. wrote the first draft of manuscript.
J.T., and L.Z. did the field work; K.F., Y.D., and J.Y. performed the data analysis; X.L., J.Y. and Y.X. improved
the English language and grammatical editing. All the coauthors contributed to the discussion, revision and
improvement of the manuscript.
Funding:
This research was funded by the Fundamental Research Funds for the Central Non-profit Research
Institution of CAF (CAFYBB2019ZA002, CAFYBB2016QA006), the National Natural Science Foundation of China
(31270474) and the National Forestry Research Project for Public Welfare (201304308).
Acknowledgments:
We would like to thank Mingxian Lin for his help during field measurement. We thank
Yide Li and Han Xu for their help during the design of this study. We are grateful to the Jianfengling Experimental
Station of the RITF and the Jianfengling Long-term Research Station of Tropical Forest Ecosystem (JRSTF) for
logistical support.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Hawkins, C.P.; MacMahon, J.A. Guilds: The multiple meanings of a concept. Annu. Rev. Entomol.
1989
, 34,
423–451. [CrossRef]
2.
Gitay, H.; Noble, I.R.; Connell, J.H. Deriving functional types for rain-forest trees. J. Veg. Sci.
1999
, 10,
641–650. [CrossRef]
3.
Pos, E.; Guevara Andino, J.E.; Sabatier, D.; Molino, J.F.; Pitman, N.; Mogoll
ó
n, H.; Neill, D.; Cer
ó
n, C.;
Rivas-Torres, G.; Di, F.A.J.E.; et al. Estimating and interpreting migration of Amazonian forests using
spatially implicit and semi-explicit neutral models. Ecol. Evol. 2017, 7, 4254–4265. [CrossRef] [PubMed]
Forests 2019, 10, 1134 12 of 14
4.
Garc
í
a-Palacios, P.; Maestre, F.T.; Gallardo, A. Soil nutrient heterogeneity modulates ecosystem responses
to changes in the identity and richness of plant functional groups. J. Ecol.
2011
, 99, 551–562. [CrossRef]
[PubMed]
5.
Lavorel, S.; D
í
az, S.; Cornelissen, J.H.C.; Garnier, E.; Harrison, S.P.; Mcintyre, S.; Pausas, J.G.;
P
é
rezharguindeguy, N.; Roumet, C.; Urcelay, C. Plant Functional Types: Are We Getting Any Closer
to the Holy Grail? Terr. Ecosyst. Chang. World 2007, 17, 149–164.
6.
Muler, A.L.; Canham, C.A.; Etten, E.J.B.V.; Stock, W.D.; Froend, R.H.J.F.E. Using a functional ecology approach
to assist plant selection for restoration of Mediterranean woodlands. For. Ecol. Manag.
2018
, 424, 1–10.
[CrossRef]
7.
Hubbell, S.P. Neutral theory in community ecology and the hypothesis of functional equivalence. Funct. Ecol.
2005, 19, 166–172. [CrossRef]
8.
McLaren, J.R.; Turkington, R. Ecosystem properties determined by plant functional group identity. J. Ecol.
2010, 98, 459–469. [CrossRef]
9.
Kooyman, R.; Rossetto, M. Definition of plant functional groups for informing implementation scenarios in
resource-limited multi-species recovery planning. Biodivers. Conserv. 2008, 17, 2917–2937. [CrossRef]
10.
Körner, C. CO
2
Fertilization: The Great Uncertainty in Future Vegetation Development. In Vegetation
Dynamics Global Change; Solomon, A., Shugart, H., Eds.; Springer: New York, NY, USA, 1993; pp. 53–70.
[CrossRef]
11.
D
í
az, S.; Noy-Meir, I.; Cabido, M. Can grazing response of herbaceous plants be predicted from simple
vegetative traits? J. Appl. Ecol. 2010, 38, 497–508. [CrossRef]
12.
Barradas, M.C.D.; Zunzunegui, M.; Tirado, R.; Ain-Lhout, F.; Novo, F.G. Plant functional types and ecosystem
function in Mediterranean shrubland. J. Veg. Sci. 2010, 10, 709–716. [CrossRef]
13.
Zhang, Z.; Zang, R. Relationship between species richness of plant functional groups and landscape patterns
in a tropical forest of Hainan Island, China. J. Trop. For. Sci. 2011, 23, 289–298.
14.
Li, N.; He, N.; Yu, G.; Wang, Q.; Jian, S. Leaf non-structural carbohydrates regulated by plant functional
groups and climate: Evidences from a tropical to cold-temperate forest transect. Ecol. Indic.
2016
, 62, 22–31.
[CrossRef]
15.
Chazdon, R.L.; Finegan, B.; Capers, R.S.; Salgado-Negret, B.; Casanoves, F.; Boukili, V.; Norden, N.
Composition and dynamics of functional groups of trees during tropical forest succession in northeastern
Costa Rica. Biotropica 2010, 42, 31–40. [CrossRef]
16.
Eisenhauer, N.; Scheu, S. Invasibility of experimental grassland communities: The role of earthworms,
plant functional group identity and seed size. Oikos 2008, 117, 1026–1036. [CrossRef]
17.
Spasojevic, M.J.; Harline, K.; Stein, C.; Mangan, S.A.; Myers, J.A. Landscape context mediates the relationship
between plant functional traits and decomposition. Plant Soil 2019, 438, 377–391. [CrossRef]
18.
Katabuchi, M.; Kurokawa, H.; Davies, S.J.; Tan, S.; Nakashizuka, T. Soil resource availability shapes
community trait structure in a species-rich dipterocarp forest. J. Ecol. 2012, 100, 643–651. [CrossRef]
19.
Hong, H.N.; Uria-Diez, J.; Wiegand, K. Spatial distribution and association patterns in a tropical evergreen
broad-leaved forest of north-central Vietnam: Official organ of the International Association for Vegetation
Science. J. Veg. Sci. 2016, 27, 318–327.
20.
Brym, Z.T.; Lake, J.K.; Allen, D.; Ostling, A. Plant functional traits suggest novel ecological strategy for an
invasive shrub in an understorey woody plant community. J. Appl. Ecol. 2011, 48, 1098–1106. [CrossRef]
21.
Reich, P.B.; Buschena, C.; Tjoelker, M.G.; Wrage, K.; Knops, J.; Tilman, D.; Machado, J.L. Variation in growth
rate and ecophysiology among 34 grassland and savanna species under contrasting N supply: A test of
functional group differences. New Phytol. 2003, 157, 617–631. [CrossRef]
22.
Luc
í
a, S.V.; Juan, D.; Jos
é
, A.; Casandra, R.-G.; Horacio, P.; Paula, J. Functional Diversity of Small and Large
Trees along Secondary Succession in a Tropical Dry Forest. Forests 2016, 7, 163.
23.
Storkey, J.; Brooks, D.; Haughton, A.; Hawes, C.; Smith, B.M.; Holland, J.M.; Lavorel, S. Using functional traits
to quantify the value of plant communities to invertebrate ecosystem service providers in arable landscapes.
J. Ecol. 2013, 101, 38–46. [CrossRef]
24.
Sanaphre-Villanueva, L.; Dupuy, J.M.; Andrade, J.L.; Reyes-Garc
í
a, C.; Jackson, P.C.; Paz, H. Patterns of plant
functional variation and specialization along secondary succession and topography in a tropical dry forest.
Environ. Res. Lett. 2017, 12, 055004. [CrossRef]
Forests 2019, 10, 1134 13 of 14
25.
Cornelissen, J.H.C.; Cerabolini, B.; Castro-D
í
ez, P.; Villar-Salvador, P.; Montserrat-Mart
í
, G.; Puyravaud, J.P.;
Maestro, M.; Werger, M.J.A.; Aerts, R. Functional traits of woody plants: Correspondence of species rankings
between field adults and laboratory-grown seedlings? J. Veg. Sci. 2003, 14, 311–322. [CrossRef]
26.
Westoby, M. A leaf-height-seed (LHS) plant ecology strategy scheme. Plant Soil
1998
, 199, 213–227. [CrossRef]
27.
Domingues, T.F.; Martinelli, L.A.; Ehleringer, J.R. Ecophysiological traits of plant functional groups in forest
and pasture ecosystems from eastern Amazonia, Brazil. Plant Ecol. 2007, 193, 101–112. [CrossRef]
28.
D
í
az, S.; Cabido, M. Plant functional types and ecosystem function in relation to global change. J. Veg. Sci.
1997, 8, 463–474. [CrossRef]
29.
Meers, T.L.; Bell, T.L.; Enright, N.J.; Kasel, S. Role of plant functional traits in determining vegetation
composition of abandoned grazing land in north-eastern Victoria, Australia. J. Veg. Sci.
2008
, 19, 515–524.
[CrossRef]
30.
Mechelen, C.V.; Dutoit, T.; Kattge, J.; Hermy, M.J.E.E. Plant trait analysis delivers an extensive list of potential
green roof species for Mediterranean France. Ecol. Eng. 2014, 67, 48–59. [CrossRef]
31.
Lima, T.A.; Vieira, G. High plant species richness in monospecific tree plantations in the Central Amazon.
For. Ecol. Manag. 2013, 295, 77–86. [CrossRef]
32.
Letcher, S.G.; Lasky, J.R.; Chazdon, R.L.; Norden, N.; Wright, S.J.; Meave, J.A.; P
é
rez-Garc
í
a, E.A.; Muñoz, R.;
Romero-P
é
rez, E.; Andrade, A.; et al. Environmental gradients and the evolution of successional habitat
specialization: A test case with 14 Neotropical forest sites. J. Ecol. 2015, 103, 1276–1290. [CrossRef]
33.
Zeng, Q.B.; Li, Y.D.; Chen, B.F.; Wu., Z.M.; Zhou, G.Y. Research and Management of Tropical Ecosystem;
China Forestry Press: Beijing, China, 1997. (In Chinese)
34.
Jiang, Y.X.; Lu, J.P. Tropical Forest Ecosystem in Jianfengling, Hainan Island, China; Science Press: Beijing, China,
1991. (In Chinese)
35.
Alvarez-Anorve, M.Y.; Quesada, M.; Sanchez-Azofeifa, G.A.; Avila-Cabadilla, L.D.; Gamon, J.A. Functional
regeneration and spectral reflectance of trees during succession in a highly diverse tropical dry forest
ecosystem. Am. J. Bot. 2012, 99, 816–826. [CrossRef] [PubMed]
36.
Maechler, M.; Rousseeuw, P.; Struyf, A.; Hubert, M.; Hornik, K. cluster: Cluster Analysis Basics and
Extensions. R Package Version 2012, 1, 56.
37.
Jari Oksanen, F.; Blanchet, G.; Friendly, M.; Kindt, R.; Legendre, P.; McGlinn, D.; Minchin, P.R.; O’Hara, R.B.;
Simpson, G.L.; Solymos, P.; et al. vegan: Community Ecology Package. R Package Version 2.5-5. Available
online: https://CRAN.R-project.org/package=vegan (accessed on 1 September 2019).
38.
Mclaren, J.R.; Novoplansky, A.; Turkington, R.J.P.E. Few effects of plant functional group identity on
ecosystem properties in an annual desert community. Plant. Ecol. 2016, 217, 1–15. [CrossRef]
39.
Holl, K.D.; Zahawi, R.A.; Cole, R.J.; Ostertag, R.; Cordell, S. Planting Seedlings in Tree Islands Versus
Plantations as a Large-Scale Tropical Forest Restoration Strategy. Restor. Ecol.
2011
, 19, 470–479. [CrossRef]
40.
Reich, P.B.; Walters, M.B.; Ellsworth, D.S. From tropics to tundra: Global convergence in plant functioning.
Proc. Natl. Acad. Sci.USA 1997, 94, 13730–13734. [CrossRef]
41. Craine, J.; Tilman, D.; Wedin, D.; Reich, P.; Tjoelker, M.; Knops, J. Functional traits, productivity and effects
on nitrogen cycling of 33 grassland species. Funct. Ecol. 2002, 16, 563–574. [CrossRef]
42.
Westoby, M.; Falster, D.; Moles, A.; Vesk, P.; Wright, I. Plant ecological strategies: Some leading dimensions
of variation between species. Annu. Rev. Ecol. Syst. 2002, 33, 125–159. [CrossRef]
43.
Foster, T.E.; Brooks, J.R. Functional groups based on leaf physiology: Are they spatially and temporally
robust? Oecologia 2005, 144, 337–352. [CrossRef]
44.
H
é
rault, B.; Bachelot, B.; Poorter, L.; Rossi, V.; Bongers, F.; Chave, J.; Paine, C.E.T.; Wagner, F.; Baraloto, C.
Functional traits shape ontogenetic growth trajectories of rain forest tree species. J. Ecol.
2011
, 99, 1431–1440.
[CrossRef]
45. Risser, P.G. Biodiversity and ecosystem function. Conserv. Biol. 1995, 9, 742–746. [CrossRef]
46.
L
ó
pez-Mart
í
nez, J.; Sanaphre-Villanueva, L.; Dupuy, J.; Hernandez-Stefanoni, J.L.; Meave, J.; Gallardo, A.
β
-Diversity of Functional Groups of Woody Plants in a Tropical Dry Forest in Yucatan. PLoS ONE
2013
,
8, e73660. [CrossRef] [PubMed]
47.
Chapin, F. Physiological controls over plant establishment in primary succession. In Primary Succession on
Land; Miles, J., Walton, D.W.H., Eds.; Blackwell Scientific Publications: Oxford, UK, 1993; pp. 161–178.
48.
Pickett, S.; Collins, S.; Armesto, J. A hierarchical consideration of causes and mechanisms of succession.
Vegetatio 1987, 69, 109–114. [CrossRef]
Forests 2019, 10, 1134 14 of 14
49.
Lebrija-Trejos, E.; P
é
rez-Garc
í
a, E.A.; Meave, J.A.; Poorter, L.; Bongers, F. Environmental changes during
secondary succession in a tropical dry forest in Mexico. J. Trop. Ecol. 2011, 27, 477–489. [CrossRef]
50.
Zhang, Z.-D.; Zang, R.-G.; Convertino, M. Predicting the distribution of potential natural vegetation based
on species functional groups in fragmented and species-rich forests. Plant Ecol. Evol.
2013
, 146, 261–271.
[CrossRef]
©
2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).