Rodriguésia 72: e00142020.

2021
http://rodriguesia.jbrj.gov.br
DOI: http://dx.doi.org/10.1590/2175-7860202172136

Original Paper
Environmental factors driving plant trait distributions
in coastal zones of Atlantic Forest

Lays Lins1,4, Juliana Da Silva-Pinheiro1,5, Ricardo A. Correia1,2, Laurício Endres3,6,

Ana Cláudia Mendes Malhado1,7, Gilson Moura Filho3,8, Flávia de Barros Prado Moura1,9
& Gilberto Costa Justino1,10,11

Abstract
Environmental filtering has been defined as the effect of environmental gradients on species in a plant
community and can be the dominant driver of community assembly. Here, we evaluate the relationship between
plant communities and the environment in the Restinga vegetation. For this, we measured 11 functional traits
of plant species present along transects covering a marked edaphic environmental gradient. This gradient
was characterized through Principal Component Analysis of soil characteristics. The relationships between
the edaphic gradient and functional traits were evaluated using linear models. Finally, we compared the
contributions of species turnover and intraspecific variation to among-site variation in functional traits. The
gradients associated with soil nutrients (PCA axis 1) and soil acidity and organic matter (PCA axis 2) were then
used to test the observed changes in community composition and were significant predictors of the distribution
of water potential, leaf dry matter content and K content, height and chlorophyll index. Decomposing the total
variation in the distribution of functional traits between species turnover and intraspecific variation revealed
that species turnover explains a greater proportion of the observed variation. We conclude that community
assembly is strongly limited by environmental filters and mediated by functional traits at the species level.
Key words: assembly rules, community composition, dune, functional trait, Restinga, turnover.
Resumo
A filtragem ambiental é o efeito dos gradientes ambientais nas espécies e pode ser o principal fator de montagem
de comunidades de plantas. Aqui, avaliamos a relação entre comunidades vegetais e o solo na vegetação
Restinga do nordeste do Brasil. Para isso, medimos 11 características funcionais das espécies vegetais presentes
ao longo dos transectos, cobrindo um gradiente ambiental marcado. Este gradiente foi caracterizado através
da Análise de Componentes Principais (PCA) da composição química do solo. As relações entre o gradiente
edáfico e as características funcionais foram avaliadas usando modelos lineares. Finalmente, comparamos
as contribuições da substituição de espécies e variação intraespecífica na comunidade vegetal. Os gradientes
associados aos nutrientes do solo (eixo 1 da PCA) e acidez do solo e matéria orgânica (eixo 2 da PCA) foram
então utilizados para testar as mudanças observadas na composição da comunidade e uma relação significativa
foi encontrada com o eixo 2. Os eixos foram preditores significativos da distribuição da altura, teor de clorofila,
potencial hídrico, massa seca e teor de K. A decomposição da variação total na distribuição de características
funcionais entre a rotatividade de espécies e a variação intraespecífica revelou que a rotatividade de espécies

See supplementary material at <https://doi.org/10.6084/m9.figshare.16879834.v1>

1
Universidade Federal de Alagoas, Instituto de Ciências Biológicas e da Saúde, Tabuleiro dos Martins, Maceió, AL, Brasil.
2
Helsinki Institute of Sustainability Science (HELSUS), Department of Geosciences and Geography, Helsinki Lab of Interdisciplinary Conservation Science,
University of Helsinki, Helsinki, Finland. ORCID: <https://orcid.org/0000-0001-7359-9091>.
3
Centro de Ciências Agrárias, Cidade Universitária, Maceió, AL, Brasil.
4
ORCID: <https://orcid.org/0000-0001-8074-7291>. 5 ORCID: <https://orcid.org/0000-0002-3656-8189>. 6 ORCID: <https://orcid.org/0000-0003-1215-4456>.
7
ORCID: <https://orcid.org/0000-0003-3621-779X>. 8 ORCID: <https://orcid.org/0000-0003-0951-959X>. 9 ORCID: <https://orcid.org/0000-0003-0014-9561>.
10
ORCID: <https://orcid.org/0000-0001-9754-8280>.
11
Author for correspondence: [email protected]
2 de 12 Lins L et al.

explica uma proporção maior da variação observada. Concluímos que a montagem da comunidade é fortemente
limitada por filtros ambientais e mediada por características funcionais no nível das espécies.
Palavras-chaves: regras de montagem, estrutura da comunidade, duna, atributos funcionais, Restinga,
variação interespecífica.

Introduction to environmental variation (Jung et al. 2010; De

A central goal of ecology is to understand Bello et al. 2011; Siefert et al. 2015; Zorger et al.
the processes that influence community assembly 2019). While some researchers have proposed that
and that control species distributions across competition for limited resources is the main driver
environmental gradients (McGill et al. 2006; of intraspecific trait variation of co-occurring
Fortunel et al. 2014). Two contrasting deterministic species ( Lepš et al. 2011; Auger & Shipley 2013),
processes have been implicated in structuring plant others have suggested that environment filtering
communities: environmental filters (Weiher & is the dominant driver of community assembly
Keddy 1995; Cornwell et al. 2006) and limiting (Barros et al. 2017; Zhang et al. 2018).
similarity (MacArthur & Levins 1967; Stubbs & In heterogeneous environments such as sand
Wilson 2004). In relation to the former, community dune ecosystems, vegetation mosaics are often
assemblage can be conceptualized as consequence associated with environmental filtering imposed
of successive environmental filters which exclude by strong edaphic gradients (Cardoso & Lomônaco
species lacking certain functional traits (adaptations 2003). The mosaic of plant communities of the
to local abiotic/biotic conditions) from the species coastal zones of Brazil, generally known as Restinga,
pool (Kraft et al. 2015). As a consequence, species is a good example of such a heterogeneous habitat
within a filtered community tend to have similar subject to intense biophysical selection pressures.
functional attributes that allow them to survive Specifically, Restinga communities are subject to
within the environment. In contrast, the concept a wide array of environmental conditions, such
of limiting similarity suggests that co-occurring as high temperatures, flooding, high salinity and
species need to have somewhat different ecological lack of nutrients (Assumpção & Nascimento 2000;
strategies (and traits) in order to avoid high levels Scarano 2002). Plant survival and reproduction
of interspecific competition (MacArthur & Levins is therefore closely associated with levels of soil
1967). exposure, edaphic factors, and vegetation critical
Functional traits can represent ecological mass required to stabilize nutrient relations of the
strategies and determine how plants respond to system (Reinert et al. 1997). Thus, the Restinga
environmental factors (Pérez-Harguindeguy et al. ecosystem provides an interesting opportunity
2013). The distribution of traits in the community to evaluate the role of environmental filtering in
should therefore broadly reflect processes that plant community composition, and to compare the
structure plant communities (Bochet & García- relative importance of intraspecific variation and
Fayos 2015). It follows that trait distributions species turnover in determining functional trait
should change along environmental gradients distributions.
because of environmental filtering, adaptive Several studies have already focused
phenotypic plasticity [e.g., individuals of many on functional trait variation and distribution
plant species are taller in fertile habitat (Lepš et al. contributes to structure the diverse communities
2011)], or a combination of these two effects ( Lepš of species rich tropical plant communities through
et al. 2011; Siefert et al. 2014; Zuo et al. 2017). differential niche-based processes (Kraft et al.
Numerous studies examining community 2008; Malhado et al. 2009; Neyret et al. 2016).
assembly have focused on trait variation Although some studies carried out on sandy soil and
due to changes in species composition along Restingas have demonstrated functional responses
environmental gradients, because it typically of plants to different environmental conditions
involves clear and easily measurable differences (see Mahdavi & Bergmeier 2016; Rosado &
in trait values among species ( Lepš et al. 2011; De Mattos 2010; Scarano 2002; Scarano et al.
Kichenin et al. 2013). However, recent studies 2001), information on how species turnover and
have shown that intraspecific variation may also intraspecific variation contribute to these responses
play an important role in community trait responses is still lacking. Preliminary observations suggest

Rodriguésia 72: e00142020. 2021

Plant trait responses to evironmental factors 3 de 12

that the distribution of plant traits in Restinga 1,726 mm. The natural vegetation is characterized
systems may be strongly influenced by edaphic by a mosaic of plant communities that occupy
factors, such as pH, aluminium content, organic sandy plains formed by marine deposits of the
matter and salinity (Giaretta et al. 2013; Santos- late Quaternary which lie between the sea and the
Filho et al. 2013). Given the strong environmental tabuleiros (large, flat areas formed from Tertiary
gradients present in the Restinga, we hypothesized deposits of crystalline rock) occupied by Atlantic
that environmental filters play a significant Forest (Scarano 2002). The first vegetation
role in community composition, mediated by encountered along the beach-to-inland gradient
the distribution of functional traits associated is a narrow strip of beach vegetation, in which
with survival-related pathways. Specifically, we creeping psammophytes grow on the shifting
predict that: 1) there will be significant changes in sands of the upper part of the beach. Behind these
community composition in response to observed embryo dunes is the foredune, a slightly higher and
environmental gradients; 2) changes in community older beach ridge dominated by grasses (Castanho
composition will reflect the distribution of plant et al. 2012).
functional traits along the same environmental Based on Reinert et al. (1997) and Souza
gradients, and; 3) observed functional changes will et al.(2008), we divided Restingas areas into
be mostly driven by community turnover rather four plots/zones (50 × 200 m each) characterized
than intraspecific variation. according edaphic factors and non-vegetation
types: zone a- areas nearest to the sea, where
Materials and Methods psammophilous and halophyte species dominate,
Study area and collection particularly rhizomatous, cespitose, and creepy
We conducted this study in six plots (50 × herbaceous; zone b- intermediate zones, 2
200 m) distributed in two Restingas areas (09º42’S, kilometres from the sea, composed of dense 0.5
35º47’W; 09º47’S, 35º52’W) in the Santa Rita to 1 m high shrub mosaics; zone c- zones located
Sustainable Development Reserve, Northeast furthest from the sea (3 kilometres), composed
Brazil. This reserve is within an area has a typical by shrubs and trees between 2 and 5 m high, and;
tropical climate, with a mean annual temperature (zone d) flooded areas (3,5 kilometres from the
around 24.7 °C and mean annual precipitation of sea) (Fig. 1).

Figure 1 – Representation of sampling in zones of Restingas.

Rodriguésia 72: e00142020. 2021
4 de 12 Lins L et al.
Plant morphophysiological traits were Measurement of environments
measured between March and June 2016, in order variables
to sample the plant community at peak standing We collected ten soil samples in each zone
biomass during the rainy season. In Restinga at a depth of 15 cm. We mixed the sample soil to
environments, nurse plants promote facilitation obtain a composite sample per zone. Organic matter
by enhancing fitness, survival and/or growth of content (OM, g/Kg), ion exchange capacity (IEC,
associated species (Callaway et al. 2002). This cmolc/ dm3), pH in water, Calcium (Ca2+, cmolc/
often results in the process of nucleation, i.e. the dm3), magnesium (Mg2+, cmolc/ dm3, phosphorus
formation of vegetation clumps or islands, and (P, mg/dm3), sodium (Na+, mg/dm3) and potassium
the vegetation becomes organized in small islands (K +, mg/dm 3) were extracted according to a
separated by sandy stretches with clumps (Pimentel methodology recommended by EMBRAPA (2013).
et al. 2007). Because of this, for each type of
zone, two transects (200 m) were established
Data analysis
perpendicular to the seashore. We collected all
We used a Principal Component Analysis
individuals in each transect. (PCA) with soil composition variables to
characterize the edaphic gradients present in the
Selection and measurements study site. The first two PCA axis explained a large
of functional traits proportion of the variation and were thus used to
Trait-based community analysis requires the represent the environmental gradients in subsequent
selection of traits that are critical to the community analyses. Community composition changes along
processes of interest (Cornelissen et al. 2003; the environmental gradient were detected using a
Baraloto et al. 2010). Our selection of traits thus multivariate model (ManyGLM) in the R package
represented the leaves and the overall life form mvabund (Wang et al. 2012; Warton et al. 2012).
of each species, covering a wide range of traits This methodology was specifically developed
frequently deemed essential for plant survival and for complex community composition data and
establishment in extreme environments (Tab. 1) was used for parameter estimates for generalised
(Pérez-Harguindeguy et al. 2013). We selected linear models simultaneously fitted to each of
adult individuals and minimum and preferred many variables, critically, takes into account the
number of replicates for different traits according interactions between species commonly that are
Pérez-Harguindeguy et al. (2013). absent in standard GLM approaches (Milazzo
Specifically, we measured leaf area (cm²) et al. 2016). According Warton et al. (2012) this
-l0 leaves per individual- with LAI 3000 (Li-Cor). approach detects differences in communities of
Leaf dry matter content (LDMC) (g) was measured less abundant species that may be more poorly
following drying to constant mass at 60 ºC (around represented by distance-based approaches.
72 h), and specific leaf area -SLA- was calculated Subsequently, we also analysed the relationships
for each lamina as the ratio of leaf area to leaf dry between edaphic gradients and functional traits
matter content (cm²/g). We considered the rachis using linear models. Trait averages were calculated
as part of the leaf for the SLA calculation (Pérez- from ‘specific average’ using trait values for
Harguindeguy et al. 2013). Chlorophyll index individual species that are specific to each habitat
was estimated from five measurements per lamina where the species are found (Lepš et al. 2011)
using a portable, non-destructive chlorophyll meter (Table S1, available on supplementary material
Minolta SPAD 502DL (Spectrum Technologies, <https://doi.org/10.6084/m9.figshare.16879834.
Plainfield, IL, USA). Maximum quantum efficiency v1>). Linear models were performed separately for
of PSII (Fv/Fm) was determined after 20 min of leaf each sampled trait and considered as explanatory
dark adaptation period using the saturation pulse of variables the two main axis resulting from the PCA
actinic light (8000 lmol/ m2) through a modulated described above.
fluorometer (PAM 2500). Leaf water potential Finally, we compared the relative contribution
(MPa) was measured around 12:00 PM using a of species turnover and intraspecific variation in
Scholander pressure chamber. Leaf phosphorus (P), explaining among-site variation in functional traits
sodium (Na+) and potassium (K+) were measured in by partitioning the variance explained by each
mg/g according to the methodology recommended component. For this, we used the method proposed
by EMBRAPA (Carmo et al. 2000). by Lepš et al. (2011) based on the decomposition

Rodriguésia 72: e00142020. 2021

 Table 1 – Measured functional traits, their respective biological significances and expected relation with different environmental factors

Trait Functional role References

Height Competitive vigour, resource availability. Taller plants are likely to settle in areas with nutrient-rich soils. (Lammerts et al. 1999; Cornelissen et al. 2003)

Leaf area Changes in the leaf area represent a functional strategy associated with the availability of water and/or soil (Sharma 1996; Ackerly et al. 2002)
nutrients. For example, small leaf size maintain leaf temperature and greater photosynthetic and water use

Rodriguésia 72: e00142020. 2021

efficiency under the combination of high solar radiation, low availability of water and nutrients and Salinity.
Plant trait responses to evironmental factors

Dry mass Is a predictor of nutrient availability in the soil. The increase in the availability of P and K in soils promotes a (Marschner 1995; Lawrence 2001)
gradual biomass allocation

SLA Reports on leaf longevity, photosynthetic rate and growth rate, (Wright et al. 2004; Maire et al. 2015)
Positively related with leaf nitrogen (N) concentration. Low SLA values have been associated with low soil N
and P

Chlorophyll index Chlorophylls absorb light energy and transfer it to the photosynthetic apparatus. In this way, its quantification (Sims & Gamon 2002; Spasojevic & Suding 2012)
can provide valuable information about the physiological performance of the leaves. Communities with greater
abiotic stress tend to have less chlorophyll content.

Fv/Fm Indicative of efficiency in the use of radiation by photochemistry and, consequently, the assimilation of carbon. (Tester & Bacic 2005)
Saline stress causes a complex effect on metabolism, resulting in ionic toxicity that interfere with the physical
conditions of the soil or the availability of other elements, indirectly affecting the development of plants.

Leaf water potential Is a simple indicator of the leaf’s water status; the more negative the value, the more dehydrated the leaf. The (Bréda et al. 2006)
water potential of the leaf tends to decrease as the evaporative demand increases. The hydraulic deficit results
in decreased growth and development.

P Leaf P concentration and leaf N:P were related to soil total P alone. The sequence of responses suggests that (Herbert & Fownes 1995; Ordoñez et al. 2009)
increased available P promoted an increase in photosynthetic area which led to increased wood production

Na+ Salinity reduced leaf area and number of tillers, and increased Na+ and Cl− concentrations in leaves. (Sharma 1996)

K+ The uptake of K by the leaves can increase the turgor and contribute to a greater water potential (Tanguilig et al. 1987).

N Leaf N concentration was related to both soil C:N and soil total P (Ordoñez et al. 2009)
5 de 12
6 de 12 Lins L et al.
of the sum of squares of the trait variation along and Ca variables, and explained 33.8% of the total
the gradient. To assess the respective contribution observed variance (Fig. 2). Both axes were then
of species turnover and intraspecific variability, we used to test the observed changes in community
used function “trait.flex.anova” developed by the composition between zones, and a significant
authors for the R software (R Development Core relationship was found with axis 2 (Tab. 2).
Team <https://www.r-project.org>). Plant functional traits were measured in
200 individuals of 18 wood plant species (Tab.
Results 3) and the environmental gradients represented
We found a marked environmental gradient by the two PCA axes were significant predictors
characterized by chemical changes in soil. The first of the distribution of five out of the eleven study
two PCA components explained 93% of the total traits (Fig. 3; Tab. 4). There was a relationship
variance of environmental attributes according with both predictors for water potential, LDMC
to soil composition variables (Fig. 2). PCA axis and K content, while two other traits (height and
1 was mainly associated with soil nutrients (P, chlorophyll index) only showed a significant
Na, K, Mg and IEC) and explained 59.2% of the relationship with axis 2. Effect sizes (standardized
variation observed between sampled sites, while estimates) were generally higher for axis 2 than for
the PCA axis 2 related mainly to pH, H+Al, OM axis 1 (Tab. 4). Higher values of PCA axis 2 were

Table 2 – ManyGLM testing the relationship between predictors (axis1 and axis 2 from PCA) and community
composition. Significant results are in bold.
Predictor Deviance P value
Axis 1 1.413 0.599
Axis 2 3.805 0.028

Figure 2 – Principal component analysis for edaphic attributes (Na, K, Mg, IEC, OM, Ca, Hal, pH) from two Restinga
study sites.
Rodriguésia 72: e00142020. 2021
Plant trait responses to evironmental factors 7 de 12

positively associated with height, but negatively variation and turnover was negative for chlorophyll
associated with chlorophyll index, water potential index, water potential and phosphorus.
and LDMC. In contrast, higher values of PCA axis
1 were associated with increasing water potential, Discussion
LDMC and K content (Fig. 3). Our results showed that the previously
Decomposition of total variation in the established vegetation zones are characterized
distribution of each functional trait demonstrated by a significant difference in the chemical soil
that species turnover accounts for a much larger characteristics, framing an environmental gradient.
proportion of total trait variation than intraspecific Furthermore, the observed changes in community
variation (Tab. 5). The percentage of variation composition and environmental gradients
explained by the turnover component varied correlated strongly with plant functional traits,
between ~87 and 100 whereas that of intraspecific generating additional insights into mechanisms
variability varied between 0 and 22%. For foliar behind the observed environmental filtering. Plant
area, chlorophyll index, nitrogen, sodium and community composition and associated functional
potassium concentration, all observed variation traits are known to be affected by a combination
was due to species turnover (Tab. 5). There was of environmental factors whose relative influence
a positive covariation between species turnover may be difficult to tease apart (Fortunel et al. 2014).
and intraspecific variation for height, Fv/Fm, leaf However, our results showed that some edaphic
area, LDMC and SLA, indicating that the effects factors have a significant effect on community
of species turnover and intraspecific variation composition in Restinga habitats.
reinforced each other (i.e., sites dominated by Soil pH is a factor that is known to affect the
species with high values of those traits also tended establishment of plant communities in Restingas
to have individuals with high trait values for (Santos-Filho et al. 2013), because acidic pH
their species). Co-variation between intraspecific reduces the rates of decomposition in the soil

Table 3 – Plants found in the six plots of Restinga

Family Specie
Anacardiaceae Schinus terebinthifolius Raddi.
Apocynaceae Hancornia speciosa Gomez.
Burseraceae Protium heptaphyllum (Aubl.) Marchand.
Capparidaceae Capparis cynophallophora L.
Chrisobalanaceae Chrysobalanus icaco L.
Dilleniaceae Tetracera breyniana Schltdl.
Erytrhroxylaceae Erythroxylum citrifolium A. St. –Hill.
Erythroxylum sp.
Euphorbiaceae Croton polyandrus Spreng.
Fabaceae Canavalia rosea (Sw.) DC.
Senna sp.
Crotalaria sp.
Goodeniaceae Scaevola plumeri Vahl.
Nyctaginaceae Guapira sp.
Ochnaceae Ouratea nitida Aubl.
Rubiaceae Chioccoca alba (L.) Hitchc.
Tocoyena selloana Schum.
Guettarda angelica Mart.

Rodriguésia 72: e00142020. 2021

8 de 12 Lins L et al.
and consequently decreases soil organic matter (van der Weele et al. 2000). Leaf water potential
(OM). Areas with higher OM availability are more is an indicator of the leaf’s water status and shows
favourable to the development of taller woody the balance between water availability in the
plants (Lammerts et al. 1999), which explains the soil. It tends to decrease as evaporative demand
relationship between plant height and PCA axis 2. (evapotranspiration potential) increases (Bréda et
Plant height is often related with water potential al. 2006), potentially explaining our observation of

Table 4 – Linear models testing the relationship between predictors (axis1 and axis 2 from PCA) and functional traits.
Significant results are in bold.
Axis 1 Axis 2
Trait
Estimate ± SE P value Estimate ± SE P value
Height -0.137 ± 0.046 0.06 0.200 ± 0.061 0.04
Chlorophyll index 0.333 ± 0.357 0.41 -1.712 ± 0.472 0.03
Water potential 0.070± 0.0193 0.03 -0.202 ± 0.025 0.004
Fv/Fm -0.015 ± 0.009 0.20 0.014 ± 0.012 0.32
Foliar Area 55.56 ± 29.37 0.15 -107.65 ± 38.88 0.06
Dry mass 0.53 ± 0.128 0.02 -0.753 ± 0.169 0.02
SLA 11.605 ± 6.912 0.19 -5.44 ± 9.152 0.59
N 1.160 ± 0.656 0.17 -2.192 ± 0.869 0.08
P 0.332 ± 0.329 0.38 0.421 ± 0.436 0.40
Na -0.066 ± 0.495 0.90 -0.872 ± 0.655 0.27
K 1.739 ± 0.478 0.03 -1.935 ± 0.632 0.05

Figure 3 – Relationship between PCA derived environmental gradients (axis 1 and 2) and height, chlorophyll, water
potential, LDMC and K contend.
Rodriguésia 72: e00142020. 2021
Plant trait responses to evironmental factors 9 de 12

higher water potential in leaves with lower levels richness of local species and a species turnover
of OM and pH in soil. among plots. The dominance of interspecific trait
A growing availability of P and K in soils variation for all eleven traits suggests that that
promotes a gradual biomass allocation (Marschner changes in the relative abundances of species are
1995; Lawrence 2001), as observed in the positive caused by habitat changes (Lepš et al. 2011; Zuo
relationship between the availability of such et al. 2017). It has been suggested that intraspecific
nutrients and LDMC. Nutrient availability is also effects are more important in short environmental
an important indicator of strategies of habitat gradients, where changes in species composition
preferences and of productivity of plants related to are small ( Wright et al. 2004; Albert et al. 2011).
environments under stress (Niklas & Christianson However, in resource-poor environments the effect
2011). Under conditions of water stress, such as of environmental filters tends to be more intense
found in Restinga (Scarano 2002; Neves et al. and the plant communities are correspondingly
2009), the survival of plants may be partly due to more influenced by species selection and show
their ability to maintain turgor as a result of the higher levels of functional convergence (Kraft et
slow decline in leaf water potential brought about al. 2008; Lohbeck et al. 2014).
by low transpiration rate and continued uptake Environmental filter should weaken the
of nutrients, especially K (Tanguilig et al. 1987). intraspecific variation, because rather than
Increased K uptake suggests that, under water being excluded from a habitat species can
stress conditions, K may be absorbed preferably adjust their characteristics to the environment
and may result in an osmotic adjustment which (Kichenin et al. 2013). Nevertheless, in our
contributes to higher leaf-water potential (Tanguilig study it is likely that intraspecific trait plasticity
et al. 1987). We observed a similar pattern in the is insufficient to buffer against environmental
positive relationship between K availability and filtering. The greatest contribution of intraspecific
water potential. variation was observed in water potential (22%).
These mechanisms suggest that the The intraspecific variation observed in this
environmental gradients observed in Restinga characteristic show negative covariation indicating
areas ultimately structure the distribution of local that areas dominated by species with high water
plant communities and their associated traits. potential values tend to contain individuals
Different environmental conditions result in an with low trait values for their species (Siefert et
increase of number of ecological niches within al. 2014). It follows that species turnover and
each site reflecting in a great variation in the intraspecific variation compensate for each other,

Table 5 – Percentage of total variation in community-weighted mean trait values explained due to species turnover,
intraspecific variation, and their covariation.
Trait Turnover Intraspecific Covariation
Height 88% 5% 7%
Chlorophyll index 100% 10% -10%
Water potential 87% 22% -9%
Fv/Fm 91% 2% 7%
Foliar Area 99% 0% 1%
Dry mass 91% 4% 5%
SLA 92% 2% 6%
N 100% 0% 0%
P 97% 14% -11%
Na 99% 0% 0%
K 99% 1% 0%

Rodriguésia 72: e00142020. 2021

10 de 12 Lins L et al.
leading to an absence of response of community- variability be considered in trait-based plant
average trait values (Lepš et al. 2011). ecology? Perspectives in Plant Ecology, Evolution
The effect of environmental filtering and Systematics 13: 217-225.
observed in our study implies similar ecological Assumpção J & Nascimento MT (2000) Estrutura e
needs of co-occurring species. Others studies composição florística de quatro formações vegetais
de restinga no complexo lagunar Grussaí/Iquipari,
have observed that species in dunes from the
São João da Barra, RJ, Brasil. Acta Botanica
same habitat type have similar traits as result Brasilica 14: 301-315.
of exposure to the same environmental stress Auger S & Shipley B (2013) Inter-specific and intra-
(Scarano et al. 2001; Mahdavi & Bergmeier 2016). specific trait variation along short environmental
These results do not support the action of limiting gradients in an old-growth temperate forest. Journal
similarity in dunes as proposed by Stubbs & of Vegetation Science 24: 419-428.
Wilson (2004). According to these authors species Baraloto C, Timothy Paine CE, Patiño S, Bonal D,
can more readily coexist if they differ in their Hérault B & Chave J (2010) Functional trait
resource use patterns. Rosado & De Mattos (2010) variation and sampling strategies in species-rich
also observed that traits of the leaf economic plant communities. Functional Ecology 24: 208-
spectrum showed small interspecific variation. 216.
However, these authors also noted that traits with Barros MF, Pinho BX, Leão T & Tabarelli M (2017)
Soil attributes structure plant assemblages across
physiological significance for the Restinga, such as
an Atlantic forest mosaic. Journal of Plant Ecology
water potential, showed considerable interspecific 11: 613-622.
variation and recommend considerable caution De Bello F, Lavorel S, Albert CH, Thuiller W, Grigulis K,
when selecting functional traits to describe whole- Dolezal J, Janeček Š & Lepš J (2011) Quantifying
plant responses to environmental drivers. In the relevance of intraspecific trait variability for
summary, the ability of a given species to colonize functional diversity. Methods in Ecology and
the dune habitat depends on integrative traits Evolution 2: 163-174.
resulting from multiple trait arrays, not necessarily Bochet E & García-Fayos P (2015) Identifying
captured by the leaf economic spectrum (Rosado plant traits: A key aspect for species selection
& De Mattos 2017). in restoration of eroded roadsides in semiarid
This study provides important insights environments. Ecological Engineering 83: 444-451.
about how environmental factors influence the Bréda N, Huc R, Garnier A, Dreyer E, Granier A &
Dreyer E (2006) Temperate forest trees and stands
distribution of plant traits in the Restinga. The
under severe drought: a review of ecophysiological
strong environmental gradients observed are
responses, adaptation processes and long-term
clear and intuitive. Still, it would be interesting to consequences. Annals of Forest Science 63: 625-
explore similar questions in other tropical systems 644.
where, like the Restinga, vegetation is established Callaway RM, Brooker RW, Choler P, Kikvidze Z, Lortie
in sandy soils. Such habitats are likewise somewhat CJ, Michalet R, Paolini L, Pugnaire FI, Newingham
poorly studied and may provide important insights B, Aschehoug ET, Armas C, Kikodze D & Cook BJ
into plant community dynamics under strong (2002) Positive interactions among alpine plants
environmental gradients. increase with stress. Nature 417: 844-848.
Cardoso G & Lomônaco C (2003) Variações fenotípicas
Acknowledgements e potencial plástico de Eugenia calycina Cambess.
(Myrtaceae) em uma área de transição cerrado-
We also thank Francesco de Bello, for vereda. Rev Bras Botânica. Brazilian Journal of
suggestions for data analysis; and Richard Ladle, Botany 26: 131-140.
for the English language review. RAC is currently Carmo CAFS, Araújo WS, Bernardi ACC & Saldanha
funded by a post-doctoral scholarship from FCT MFC (2000) Métodos de análise de tecidos vegetais
- Fundação para a Ciência e Tecnologia (SFRH/ utilizados na Embrapa Solos. 41. Embrapa Solos-
BPD/118635/2016). ACMM is funded by CNPq- Circular Técnica (INFOTECA-E), Rio de Janeiro.
Conselho Nacional de Desenvolvimento Científico 41p.
e Tecnológico (#309980/2018-6). Castanho CT, Oliveira AA & Ina P (2012) The
importance of plant life form on spatial associations
along a subtropical coastal dune gradient. Journal
References of Vegetation Science 23: 952-961.
Albert CH, Grassein F, Schurr FM, Vieilledent G & Cornelissen JHC, Lavorel S, Garnier E, Díaz S,
Violle C (2011) When and how should intraspecific Buchmann N, Gurvich DE, Reich PB, Ter Steege

Rodriguésia 72: e00142020. 2021

Plant trait responses to evironmental factors 11 de 12

H, Morgan HD, Van Der Heijden MG a, Pausas Gonzalez G, Baker TR, Anderson LO, Arroyo L,
JG & Poorter H (2003) A handbook of protocols Almeida S, Higuchi N, Killeen TJ, Monteagudo
for standardised and easy measurement of plant A, Neill DA, Pitman NCA, Prieto A, Salomao RP,
functional traits worldwide. Aust J Bot 51:335–380. Vasquez R, Laurance WF & Ramirez H (2009)
Cornwell WK, Schwilk DW & Ackerly DD (2006) A Spatial distribution and functional significance
trait-based test for habitat filtering: convex hull of leaf lamina shape in Amazonian forest trees.
volume. Ecology 87: 1465-1471. Biogeosciences 6: 1577-1590.
Fortunel C, Paine CET, Fine PVA, Kraft NJB, Baraloto Marschner H (1995) Mineral nutrition of higher plants.
C & Firme T (2014) Environmental factors predict Academic Press, London. 889p.
community functional composition in Amazonian McGill BJ, Enquist BJ, Weiher E & Westoby M (2006)
forests. Journal of Ecology 102: 145-155. Rebuilding community ecology from functional
Giaretta A, Menezes LFT De & Pereira OJ (2013) traits. Trends in Ecology & Evolution 21: 178-85.
Structure and floristic pattern of a coastal dunes Milazzo M, Quattrocchi F, Azzurro E, Palmeri A,
in southeastern Brazil. Acta Botanica Brasilica 27: Chemello R, Di Franco A, Guidetti P, Sala E,
87-107. Sciandra M, Badalamenti F & García-Charton JA
Jung V, Violle C, Mondy C, Hoffmann L & Muller (2016) Warming-related shifts in the distribution
S (2010) Intraspecific variability and trait-based of two competing coastal wrasses. Marine
community assembly. Journal of Ecology 98: Environmental Research 120: 55-67.
1134-1140. Neves NR, Oliva MA, Centeno C, Costa AC, Ribas RF &
Kichenin E, Wardle DA, Peltzer DA, Morse CW & Pereira EG (2009) Science of the Total Environment
Freschet GT (2013) Contrasting effects of plant Photosynthesis and oxidative stress in the restinga
inter- and intraspecific variation on community- plant species Eugenia uni fl ora L . exposed to
level trait measures along an environmental simulated acid rain and iron ore dust deposition :
gradient. Functional Ecology 27: 1254-1261. Potential use in environmental risk assessment.
Kraft NJB, Adler PB, Godoy O, James EC, Fuller Science of the Total Environment 407: 3740-3745.
S & Levine JM (2015) Community assembly, Neyret M, Bentley LP, Oliveras I, Marimon BS,
coexistence and the environmental filtering
Marimon-Junior BH, Almeida de Oliveira E,
metaphor. Functional Ecology 29: 592-599.
Barbosa Passos F, Castro Ccoscco R, dos Santos
Kraft NJB, Valencia R & Ackerly DD (2008) Functional
J, Matias Reis S, Morandi PS, Rayme Paucar G,
traits and niche-based tree community assembly in
Robles Cáceres A, Valdez Tejeira Y, Yllanes Choque
an Amazonian forest. Science 322: 580-582.
Y, Salinas N, Shenkin A, Asner GP, D??az S, Enquist
Lammerts EJ, Pegtel DM, Grootjans AP & Van Der Veen
BJ & Malhi Y (2016) Examining variation in the
A (1999) Nutrient limitation and vegetation changes
leaf mass per area of dominant species across two
in a coastal dune slack. Journal of Vegetation
contrasting tropical gradients in light of community
Science 10: 111-122.
assembly. Ecology and Evolution 6: 5674-5689.
Lawrence D (2001) Nitrogen and phosphorus enhance
growth and luxury consumption of four secondary Niklas KJ & Christianson ML (2011) Differences in
forest tree species in Borneo. Journal of Tropical the scaling of area and mass of Ginkgo biloba
Ecology 17: 859-869. (Ginkgoaceae) leaves and their relevance to the
Lepš J, de Bello F, Šmilauer P & Doležal J (2011) study of specific leaf area. American Journal of
Community trait response to environment: Botany 98: 1381-1386.
Disentangling species turnover vs intraspecific trait Pérez-Harguindeguy N, Díaz S, Garnier E, Lavorel
variability effects. Ecography (Cop) 34: 856-863. S, Poorter H, Jaureguiberry P, Bret-Harte MS,
Lohbeck M, Poorter L, Martínez-Ramos M, Rodriguez- Cornwell WK, Craine JM, Gurvich DE, Urcelay
Velázquez J, van Breugel M & Bongers F (2014) C, Veneklaas EJ, Reich PB, Poorter L, Wright IJ,
Changing drivers of species dominance during Ray P, Enrico L, Pausas JG, De Vos AC, Buchmann
tropical forest succession. Functional Ecology 28: N, Funes G, Quétier F, Hodgson JG, Thompson K,
1052-1058. Morgan HD, Ter Steege H, Van Der Heijden MGA,
MacArthur R & Levins R (1967) The limiting similarity, Sack L, Blonder B, Poschlod P, Vaieretti M V., Conti
convergence, and divergence of coexisting species. G, Staver AC, Aquino S & Cornelissen JHC (2013)
The American Naturalist 101: 377-385. New handbook for standardised measurement of
Mahdavi P & Bergmeier E (2016) Plant functional plant functional traits worldwide. Australian Journal
traits and diversity in sand dune ecosystems across of Botany 61: 167-234.
different biogeographic regions. Acta Oecologica Pescador DS, De Bello F, Valladares F & Escudero A
74: 37-45. (2015) Plant trait variation along an altitudinal
Malhado ACM, Whittaker RJ, Malhi Y, Ladle RJ, ter gradient in mediterranean high mountain grasslands:
Steege H, Butt N, Aragao LEOC, Quesada CA, Controlling the species turnover effect. PLoS One
Murakami-Araujo A, Phillips OL, Peacock J, Lopez- 10: 1-16.

Rodriguésia 72: e00142020. 2021

12 de 12 Lins L et al.
Pimentel MCP, Barros MJ, Cirne P, Mattos EADE, Souza CRDG, Hiruma ST, Sallun AEM, Ribeiro RR
Oliveira RC, Pereira MCA, Scarano FR, Zaluar & Sobrinho JMA (2008) “Restinga” - Conceitos
HLT & Araujo DSD (2007) Spatial variation in the e empregos do termo no Brasil e implicações na
structure and floristic composition of “restinga” legislação ambiental. Instituto Geológico, São
vegetation in southeastern Brazil 1. Brazilian Paulo.
Journal of Botany 30: 543-551. Stubbs WJ & Wilson JB (2004) Evidence for limiting
Reinert F, Roberts a, Wilson JM, Deribas L, Cardinot similarity in a sand dune community. Journal of
G & Griffith H (1997) Gradation in nutrient Ecology 92: 557-567.
composition and photosynthetic pathways across Tanguilig V., Yambao EB, O’Toole JC & De Datta SK
the restinga vegetation of Brazil. Acta Botanica (1987) Water stress effects on leaf elongation, leaf
Brasilica 110: 135-142. water potential, transpiration, and nutrient uptake of
Rosado B enrique P & De Mattos EA (2010) Interspecific rice, maize, and soybean. Plant Soil 103: 155-168.
variation of functional traits in a CAM-tree Wang Y, Naumann U, Wright ST & Warton DI (2012)
dominated sandy coastal plain. J Veg Sci 21: 43-54. Mvabund- an R package for model-based analysis
Rosado BHP & de Mattos EA (2017) On the relative of multivariate abundance data. Methods in Ecology
importance of CSR ecological strategies and and Evolution 3: 471-474.
integrative traits to explain species dominance at Warton DI, Wright ST & Wang Y (2012) Distance-
local scales. Functional Ecology 31: 1969-1974. based multivariate analyses confound location
Santos-Filho FS, Almeida Jr E & Zickel CS (2013) Do and dispersion effects. Methods in Ecology and
edaphic aspects alter vegetation structures in the Evolution 3: 89-101.
Brazilian restinga? Acta Botanica Brasilica 27: van der Weele CM, Spollen WG, Sharp RE & Baskin
613-623. TI (2000) Growth of Arabidopsis thaliana seedlings
Scarano FR (2002) Structure, function and floristic under water deficit studied by control of water
relationships of plant communities in stressful potential in nutrient-agar media. Journal of
habitats marginal to the Brazilian Atlantic rainforest. Experimental Botany 51: 1555-1562.
Annals of Botany 90: 517-524. Weiher E & Keddy PA (1995) Assembly rules, null
Scarano FR, Duarte HM, Ribeiro KT, Rodrigues models, and trait dispersion: new questions from
PJFP, Barcellos EMB, Franco AC, Brulfert J, old patterns. Oikos 74: 159-164.
DelÉens E & Lüttge U (2001) Four sites with Wright IJ, Reich PB, Westoby M, Ackerly DD, Baruch Z,
contrasting environmental stress in southeastern Bongers F, Cavender-Bares J, Chapin T, Cornelissen
Brazil: Relations of species, life form diversity, JHC, Diemer M, Flexas J, Garnier E, Groom PK,
and geographic distribution to ecophysiological Gulias J, Hikosaka K, Lamont BB, Lee T, Lee W,
parameters. Botanical Journal of the Linnean Lusk C, Midgley JJ, Navas M-L, Niinemets U,
Society 136: 345-364. Oleksyn J, Osada N, Poorter H, Poot P, Prior L,
Siefert A, Fridley JD & Ritchie ME (2014) Community Pyankov VI, Roumet C, Thomas SC, Tjoelker MG,
functional responses to soil and climate at multiple Veneklaas EJ & Villar R (2004) The worldwide leaf
spatial scales: When does intraspecific variation economics spectrum. Nature 428: 821-7.
matter? PLoS One 9: e111189. Zhang H, Chen HYH, Lian J, John R, Ronghua L, Liu H,
Siefert A, Violle C, Chalmandrier L, Albert CH, Taudiere Ye W, Berninger F & Ye Q (2018) Using functional
A, Fajardo A, Aarssen LW, Baraloto C, Carlucci trait diversity patterns to disentangle the scale-
MB, Cianciaruso M V., de L. Dantas V, de Bello F, dependent ecological processes in a subtropical
Duarte LDS, Fonseca CR, Freschet GT, Gaucherand forest. Functional Ecology 32: 1379-1389.
S, Gross N, Hikosaka K, Jackson B, Jung V, Zorger BB, Tabarelli M, de Queiroz RT, Rosado BHP &
Kamiyama C, Katabuchi M, Kembel SW, Kichenin Pinho BX (2019) Functional organization of woody
E, Kraft NJB, Lagerström A, Bagousse-Pinguet plant assemblages along precipitation and human
Y Le, Li Y, Mason N, Messier J, Nakashizuka T, disturbance gradients in a seasonally dry tropical
Overton JM, Peltzer DA, Pérez-Ramos IM, Pillar forest. Biotropica 51: 838-850.
VD, Prentice HC, Richardson S, Sasaki T, Schamp Zuo X, Yue X, Lv P, Yu Q, Chen M, Zhang J, Luo Y,
BS, Schöb C, Shipley B, Sundqvist M, Sykes MT, Wang S & Zhang J (2017) Contrasting effects
Vandewalle M & Wardle DA (2015) A global meta- of plant inter- and intraspecific variation on
analysis of the relative extent of intraspecific trait community trait responses to restoration of a sandy
variation in plant communities. Ecology Letters grassland ecosystem. Ecology and Evolution 7:
18: 1406-1419. 1125-1134.

Area Editor: Dr. Mário Luís Garbin

Received in February 12, 2020. Accepted in December 08, 2020.
This is an open-access article distributed under the terms of the Creative Commons Attribution License.

Rodriguésia 72: e00142020. 2021