Journal of Threatened Taxa |
www.threatenedtaxa.org | 26 April 2024 | 16(4): 25057–25068
ISSN 0974-7907
(Online) | ISSN 0974-7893 (Print)
https://doi.org/10.11609/jott.7702.16.4.25057-25068
#7702 | Received 18
October 2021 | Final received 10 February 2024 | Finally accepted 10 March 2024
Elliptic Fourier analysis of leaf
shape of Callicarpa pedunculata and Callicarpa
rubella (Lamiaceae)
Jennifer S. Danila 1 & Grecebio
Jonathan D. Alejandro 2
1 The Graduate School, University
of Santo Tomas. España Blvd., 1015 Manila,
Philippines.
1,2 College of Science and Research
Centre for the Natural and Applied Sciences, University of Santo Tomas, España Blvd., 1015 Manila, Philippines.
1 jennifer.danila.gs@ust.edu.ph
(corresponding author), 2 gdalejandro@ust.edu.ph
Abstract: Leaves play an important role in
species discrimination. An elliptic Fourier analysis (EFA) based
morphometric technique was used to assess divergence between the poorly
differentiated species, Callicarpa pedunculata
and C. rubella. Using leaf specimen images from herbarium collections,
principal components (PCs) were extracted from the Fourier coefficients and
used to describe leaf outline and leaf shape descriptors: circularity, aspect
ratio, and solidity. The results indicate that symmetric (54%) and asymmetric
(35%) components of the leaves of C. pedunculata
and C. rubella are sources of shape variation, as shown in the width and
leaf tips among the samples. MANOVA revealed significant interspecific
differences (P = 0.03) between C. pedunculata
and C. rubella. The jack-knife cross-validation showed 71% of correctly
classified species both in C. pedunculata and C.
rubella. Furthermore, the results of this study were able to reveal
significant leaf shape descriptors like aspect ratio, circularity, and solidity
as important diagnostic characters in discriminating C. pedunculata
and C. rubella. Thus, in conclusion, leaf serrations, leaf size, and
leaf lobes are important characteristics in discriminating between C. pedunculata and C. rubella.
Keywords: Aspect ratio, Callicarpa,
circularity, correlation, evolution, geometric morphometrics, leaf, principal
component analysis, solidity, symmetry.
Editor: Mandar Paingankar, Government Science College, Gadchiroli,
India. Date of publication: 26 April
2024 (online & print)
Citation: Danila, J.S., & G.J.D. Alejandro (2024). Elliptic
Fourier analysis of leaf shape of Callicarpa pedunculata
and C. rubella (Lamiaceae). Journal of Threatened Taxa 16(4):
25057–25068. https://doi.org/10.11609/jott.7702.16.4.25057-25068
Copyright: © Danila & Alejandro 2024. Creative Commons Attribution 4.0
International License. JoTT allows unrestricted use, reproduction, and
distribution of this article in any medium by providing adequate credit to the
author(s) and the source of publication.
Funding: University of Santo Tomas, Graduate School.
Competing interests: The authors declare no competing interests.
Author
details: Jennifer S. Danila is a graduate of Doctor of Philosophy major in Biology at the University of Santo Tomas. She has expertise in plant science including systematics, taxonomy, and biodiversity conservation. Grecebio Jonathan D. Alejandro is a
professor at the College of Science and is currently the Director of the Office for Graduate Research, Graduate School, University of Santo Tomas (UST). He established the Thomasian Angiosperm Phylogeny and Barcoding Group (TAPBG) in the UST – Research Center for the Natural and Applied Sciences.
Author
contributions: JSD—contributed to field collection, data analysis, discussion of results, and conclusion of the manuscript. GJDA—led the discussion, editing, and paper review for the manuscript.
Acknowledgements: The authors would like to thank the curators of AMD,
FLMNH, K, MSU, NY, US, and USTH, and the Global Biodiversity Information
Facility (GBIF) for allowing us to access their digital collections for study.
We thank the curator of the University of Santo Tomas (USTH) for giving us
accessions to our specimens. Likewise, we would like to thank the Department of
Environment and Natural Resources (DENR) for granting us the gratuitous permit
to collect specimens in Isabela Province, Philippines.
INTRODUCTION
Callicarpa is a genus of Lamiaceae characterized by branched hair; inflorescences
axillary; flowers polysymmetric, 4(--−5) merous; anthers porose; stigma
peltate or capitate; and fruit a drupe (Linnaeus 1753; Munir 1982; Leeratiwong et al. 2009; Bramley
2013). Several species of Callicarpa have been classified and formally
recognized from different parts of the world, including the Philippines and
Borneo. C. pedunculata R. Br. and C.
rubella Lindl. show extensive distribution in the
southeastern Asian region, but both are geographically and taxonomically
controversial. C. pedunculata is not found in
Sumatra, Java and Borneo, while C. rubella is rather more extensive,
occupying a wider range in the Asian continent. In contrast, C. pedunculata is widely distributed in the Philippines,
while C. rubella is not present (BGCI 2024; Arvidsson
2020). Taxonomically, the relationship between the two taxa was not clear due
to ambiguous morphological characters.
C. pedunculata
and C. rubella were usually differentiated by their leaf size and
presence of glandular hairs (Bramley 2013): C. pedunculata has wider leaves and lacks glandular hairs,
while C. rubella has narrower leaves and hairs are present. Although its
morphology has been previously described by Bramley
(2013, 2019), C. pedunculata is easily
confused with C. rubella due to misleading morphological characters.
Likewise, several taxonomists have linked other species with C. pedunculata and C. rubella, e.g., the
long-established C. caudata Maxim and doubtful
C. cuspidata Roxb.
were linked to C. rubella based on indumentum and leaf serrations (Roxburgh 1820; Lam & Bakhuizen
1921) and leaf bases (Bramley 2013), while C. cuspidata has been reported as a synonym of C. pedunculata (Munir 1982) which adds to the confusion
between the two taxa. Likewise, no direct studies have identified the
relationship between C. pedunculata and C.
rubella to further separate or combine the two species. Thus, the taxonomic
status of C. pedunculata and C. rubella
was becoming uncertain due to the overlapping of morphological characters.
The taxonomic transcription among
C. pedunculata and C. rubella and its
closely related species were originally described by Roxburgh
(1820) and revised by Munir (1982), but, according to Bramley
(2013), they did not indicate any specimen or type to describe the species.
Consequently, Bramley (2013), considered the
description of Roxburgh (1820) and Munir (1982)
unsuitable for correct identification due to lack of data and poor vouchering.
In a previous study of Callicarpa in Thailand and the Philippines (Leeratiwong et al. 2009; Bramley
2013), C. rubella was recognized as distinct from other Callicarpa
species through its cordate or obliquely cordate leaf base, while C. pedunculata was defined by its attenuation to cuneate
leaf bases. Currently, our knowledge of these two species is known only from
collections made early in the twentieth century, and recent studies were mostly
based on herbarium specimens. The lack of updated distribution listings and
exhaustive data contributes to species taxonomic challenges. This also raises
several questions on the current conservation status and taxonomic relationship
of C. pedunculata and C. rubella. While
C. rubella is thought to be absent in the Philippines, its current
natural distribution is also difficult to determine with precision because of
the potential impact of human use in different countries. In the southeastern
Asian region, C. pedunculata and C. rubella
were reported to have medicinal properties (Brown 1920; Tu et al. 2013)
collected from twigs, roots, and leaves, while their fruits are used for human
consumption. Thus, the natural distribution of most species may have been
changed by its dispersal based on human actions affecting local or even global
distributions (Di Marco & Santini 2015; Newbold et al. 2015). The change in
the environment and distribution of species were highly influential in plant
structures, especially on leaves which serve as indicators of environmental
change (Gupta et al. 2019; Zhang & Li 2019).
In this paper, the authors
discuss leaf morphometrics using a more comprehensive quantification of leaf
shape, where measurements of individual parameters were obtained as a basis of
species discrimination. This technique, elliptical Fourier descriptors (EFD)
utilizes the sum of ellipses over contours to quantify outlines and silhouettes
in an image (McLellan & Endler 1998; Hearn 2009; Godefroy et al. 2012), based on the instructions taken from
Klein and Svoboda (2017) on geometric morphometric analysis. Aside from the
typical leaf extraction, leaf shape descriptors: Circularity, measured as 4π
(area/perimeter²) related to serrations and lobing;
Aspect ratio (AR), the ratio of the major to the minor axis and influenced by
length and width; and, Solidity, measured as area or convex hull and sensitive
to leaf deep lobes (Cope et al. 2012) were incorporated into the downstream
analysis. As leaf shapes vary among or within species, it is also important to
quantify leaf shapes to understand broader aspects of plant adaptation to the
environment (Chitwood, et al. 2014). Leaf morphological traits such as length,
width, and veins are controlled by the environment, whether to stabilize or to
adjust to certain environmental conditions (Alonso-Forn
et al. 2020). This study describes for the first-time accessions of C. pedunculata and C. rubella through leaf
morphometrics, contributing to a better understanding of the species variation
through leaf shapes. Furthermore, this study aimed to discriminate C. pedunculata and C. rubella leaf shape
descriptors: circularity, aspect ratio (AR), and solidity between the two taxa,
and predict the correlation among the three leaf descriptors.
MATERIALS AND METHODS
Study Site
A total of 46 individual
herbarium samples of C. pedunculata and C.
rubella were used in the study (Image 1). Twenty samples of C. pedunculata were collected in the secondary forests and
forest edges of Palanan, Isabela in the Philippines while 26 samples of C.
rubella were carried out from selected digital herbarium of AMD, FLMNH, K,
MSU, NY, US, and USTH (Image 2) through online accessions in the Global
Biodiversity Information Facility (GBIF) database via the web interface (Table
1). The online images and details were downloaded using the ‘Darwin Core
Archive’ format which contains the URLs and information of the samples in GBIF
(Table 9). On the other hand, samples from the fieldwork have undergone
herbarium protocol from the securing of the permit for the collection of
specimens, preparation of materials, pressing of the specimen, mounting in
herbarium sheet, identification, and labeling to the deliberation of voucher
specimen to the University of Santo Tomas Herbarium (USTH) in the Philippines.
Procedures
In this study, herbarium samples
were the main source of datasets to build shape descriptors from the leaf
outline. The collected digital images were subjected to leaf isolation using
Adobe Photoshop version 22.0.0 (Adobe System San Jose, USA). After all leaves
have been isolated from the scans, the software SHAPE (Iwata & Ukai 2002) which uses binary leaf outline image files in
BMP format converts images to black and white. SHAPE converts the image
outlines to chain code and then normalized EFDs. A maximum number of harmonics
were set to 20 to recapitulate leaf shape and the normalization method was set
to the longest radius for the initial orientation of the images. From the
obtained EFD coefficients, the analysis focused on coefficients a and d, as
well as coefficients b and c. These correspond to the symmetric and asymmetric
components of leaf shapes, respectively, following the approach outlined by Lexer et al. (2009). Subsequently, principal component
analysis (PCA) was conducted on the EFD coefficients to identify variations in
leaf shape across the entire set of leaf samples. Prinprint
program was used to view the Eigen leaves or leaf contours of each principal
component. Then an analysis of leaf shape descriptors was obtained using ImageJ
version 1.52a, Java 1.8.0_112 (64-bit) (Ambramoff et
al. 2004) software. After all images of C. pedunculata
and C. rubella were measured based on AR, circularity, and solidity,
the resulting data were imported to PAST version 4.06b software (Hammer et al.
2001) for further analysis.
RESULTS AND
DISCUSSION
Principal
Component Analysis (PCA)
Independent
shape variables were identified by PCA of EFD. Table 2 shows the relative
contributions of the first 10 PCs of the whole dataset are accounted for 93% of
the total variance while significant variations in the first four PCs (PC1,
PC2, PC3, and PC4) equal to 74% cumulative variance based on broken stick method
(MacArthur 1957). Most of the samples of C. pedunculata
and C. rubella were densely overlapping than scattered in the scatter
plot. (Figure 1). The ordination plot of the two taxa in a two-dimensional
space was highly defined by PC1 and PC2. It suggests that the plots of C. pedunculata and C. rubella are similar along PC1
and PC2, with positive values but few data points were positioned in the
negative values in both PCs which results in overlap in the interspecific
comparison suggesting similarities between the two taxa. Likewise, the
similarity in leaf shape has been reflected in the discriminant analysis (DA),
where there is no significant difference between the means (Figure 3, Hotelling’s T² = 36.83, F = 2.2419, P = 0.08146) of C.
pedunculata and C. rubella. Additionally,
the jack-knife cross-validation showed 71% of correctly classified species both
in C. pedunculata and C. rubella (Table
4). Despite similarities in the ordination of plots between the two
taxa, the comparisons showed relevant variations in their leaf mean shapes in
multivariate analysis of variance (MANOVA) as the significant difference
between the C. pedunculata and C. rubella
exists based on leaf shapes (Wilk’s λ = 0.6196, F = 2.272, d.f.
= 10 and 37, P = 0.03431) (Figure 3).
The effects
of shape variables in the Eigen leaves or leaf contours were determined based
on the scores of the first four PCs to identify symmetric (54%) and asymmetric
variations (35%). In Figure 4, symmetric variation highlights PC1 (85%) which
explains leaf shape changes in width and leaf tips among samples of C. pedunculata and C. rubella. These variations
were represented by discernible width expansion and transformation of leaf tips
from acuminate to acute. Since PC1 accounts for the highest variations, it
revealed that leaf tips and width expansion contribute to the overlapping of
the two taxa. PC2 score (8.9%) describes cuneate, oblique to cordate leaf bases
among samples, whereas PC3 (1.75%) and PC4 (1.53%) describe fine leaf changes
along its margin that exhibit variations in the basal portion of the leaf. On
the other hand, asymmetrical outline reconstruction shows basal and apical leaf
variations on PC1 (51.4%) while remaining PCs (PC2 20.5%; PC3 9.32%; PC4 6.54%)
revealed imperceptible variations across species. Thus, multivariate analyses
were more restricted to the symmetric dataset due to the inadequate
contribution of the asymmetric component.
In the recent
study of two closely related genera, Callicarpa and Geunsia,
the effect of environment and genetic factors were mentioned as the probable
cause of the taxonomic overlap between the two taxa (Danila & Alejandro
2021). In geometric morphometrics, this overlap indicates morphological
similarities among species and may occur due to the presence of hybrid among
samples (Adebowale et al. 2012). In recent years, there has been an increase in
the number of hybrids in the genus Callicarpa, e.g., C. japonica Thunb. with C. kochiana
Makino or C. mollis Siebold & Zucc., and C. dichotoma (Lour.) K.Koch with C. kwangtungensis Chun. (Yamanaka 1988, Tsukaya et al. 2003). The emergence of hybrids has brought
several consequences in the population including introgression of plant traits
or even the formation of new species which affect the interaction between plants
and the environment (Orians 2000).
Despite the
overlap, one clear finding in this study showed that symmetric variations on
the leaf bases play a key role in determining leaf shape variations between C.
pedunculata and C. rubella. In contrast to
the symmetric variations, asymmetric PC1 also showed an interspecific variation
focusing on the appearance of lobes in the basal portion of the leaf (Figure
2). In the leaf shape morphometric study conducted by Danila & Alejandro
(2021) of the genus Geunsia and Callicarpa,
the two taxa showed the possible occurrence of fluctuating asymmetry (FA). This
results when the same species were unable to go through an identical
development of the body organ on both sides resulting in uneven growth (van Valen 1962). Likewise, the occurrence of FA in leaves is a
poor sign of environmental and genetic stress which happens when two closely
related species mate and produce offspring (Sander & Matthies
2017). Hence, evidence of overlap in leaf shape variations and FA suggests that
environmental and genetic factors affect variations in the leaf shape of C. pedunculata and C. rubella.
Analysis of
Leaf Shape Descriptors
In this study, the first two
principal components (PC1 and PC2) showed the most variation among the three leaf
shape descriptors having 74% and 23%, respectively (Table 6). However, it shows
that shape trends in most samples were mostly observed in PC1 (74%). The bar
plot (Figure 5) and coefficient of correlation (Table 7) among PCs showed a
significant relationship among the three leaf-shape descriptors. PC1 is more
related to circularity (0.65380) and solidity (0.50692) but inversely related
to AR (-0.56176) while PC2 is more associated with AR (0.62891) and solidity
(0.77511) but inversely related to circularity (-0.06060). On the other hand,
PC2 marked a high coefficient of correlation in solidity and AR, but the
proportion of variability in PC2 is relatively low (23.17%). Therefore, the
first principal component (PC1) was considered a statistically significant PC
based on the broken stick method (MacArthur 1957) (Figure 4–6).
The overall
results showed that AR is the most variable leaf shape descriptor with a
Phenotypic Coefficient of Variation (PCV; ((standard deviation/mean) × 100),
estimates indicated the existence of a significant amount of variability among
species, with 19.33% followed by circularity with 16.60% (Table 5).
Additionally, both AR and circularity have a high distribution range of
1.90–4.02 and 0.29–0.57, respectively, meaning a high degree of variation was
observed among samples. On the other hand, solidity is the least variable with
the narrowest distribution (0.87–0.97) and the lowest PCV of 2.89%. Almost all
samples of C. pedunculata and C. rubella
exhibited a high AR (>1.90) which manifested an increase in leaf width
relative to the length, or vice versa. However, it shows that C. rubella
has higher PCV values (24.12%) compared to C. pedunculata
(12.54%) which indicates that the former has higher diversity in length-width
ratio. While an increase in AR manifests an increase in the size of the leaf
width relative to length, or vice-versa (Gupta et al. 2019). Some leaves of C.
rubella were narrower but with high AR, that is, a larger major axis either
on its length or width, affects the overall AR of the taxa. On the other hand,
variations in circularity were observed in all accessions, where 30 samples
indicated a low circularity (<0.50) while 16 samples had moderate
circularity (0.50–0.57), meaning the lower the circularity values, the more
prominent serrations are. Based on the observations, more specimens in C. pedunculata (45%) have more prominent serrations than
in C. rubella (30%). Thus, these observations revealed that serrations
and leaf size were useful in discriminating the two taxa. Moreover, the results
showed a significant relationship between leaf serrations to leaf size, that
is, as the leaf size increases, serrations decrease, or vice versa. Lastly,
solidity showed a narrow distribution (0.87–0.97) and low PCV values (2.89%) indicating
that most samples of C. pedunculata and C.
rubella do not have lobed leaves. However, few accessions of C. rubella
have been observed to show slightly rounded projections from the base of the
leaf blade. Likewise, these samples of C. rubella were observed to have
a lower solidity value representing cordate to oblique-cordate leaf bases. In
the study (Bramley 2013) of Callicarpa species
in the Philippines, it has been noted that most Callicarpa species have
either acute, acuminate, rounded, cuneate, oblique, or obtuse leaf bases which
are all features of species with a high solidity (>0.87). Thus, we can
conclude that solidity is also globally important as a diagnostic character to
distinguish species between C. pedunculata and
C. rubella.
Correlation among leaf shape
descriptors
Figure 5 presents a biplot that
simultaneously draws information from 46 individual samples of Callicarpa
based on three leaf shape descriptors: AR, Circularity, and Solidity. The three
leaf shape descriptors were positioned on the first, second, and fourth
quadrants while data points of samples were distributed in all four quadrants
based on their PCA scores. However, the distribution among individuals of C.
pedunculata and C. rubella has found a
minimal group differentiation due to a large degree of overlap. Although
overlap has been observed among samples, the three leaf-shape descriptors
produced a comparable level of relationship. In Table 8, the vectors of the
variables circularity and solidity were closer to each other which suggests a
positive correlation (+0.6784) between them. On the other hand, the greater
distance close to 180 degrees found between circularity and AR suggests a
negative correlation (-0.8067) while vectors of solidity and AR show almost an
angle of 90 degrees which indicates that the variables were weakly correlated
(-0.3106).
As mentioned
above, AR and circularity were found to be the two most important variables in
the discrimination of C. pedunculata and C.
rubella. These leaf shape descriptors were highly influenced by length,
width, and leaf margin. Since AR and circularity were found to be negatively
correlated, variables like the length and width of the leaf were inversely
proportional to the presence of serrations, that is, when the magnitude of the
leaf decreases, the degree of serrations increases or vice versa. These
observations exist among samples of C. pedunculata
and C. rubella, where each taxon exhibits a corresponding trait relative
to leaf serration and size. On the other
hand, circularity and solidity indicate a moderate positive correlation that
shows an impact of serrations in the projections of the leaf blade. Although a
positive correlation was found between circularity and solidity the interval
between the PCV values (circularity 16.6%; solidity 2.89%) is high, the two
variables are related but exhibit different percentages in terms of their
effects on the leaf shapes. This observation was evident among samples of C.
rubella in the occurrence of fine leaf lobes and discernible leaf
serration. While the weak correlation was observed between solidity and AR
where the former, unlike circularity, is little or not affected by serrations
and leaf lobes (Figure 5).
Several
studies (Thomas & Bazzaz 1996; Piazza 2005; Royer
& Wilf 2006; Chitwood et al. 2013) have identified several factors in the
evolution of leaf shapes and sizes, including the adaptation of plants to
various types of environments. Likewise, different environmental factors showed
a significant effect on morphological characters of closely related species
(Jones 1995; Wolfe & Liston 1998; Royer et al. 2008). However, the
adaptation mechanism in response to environmental variation in most species is
still incomprehensible (Jump & Panuelas 2005).
Since C. pedunculata and C. rubella
have been identified to grow in a different environment, the two taxa showed
distinct characteristics to discriminate the two species of Callicarpa.
However, it also revealed that C. pedunculata
and C. rubella showed similar leaf traits which can be considered as a
plesiomorphic character of the two taxa. C. pedunculata
has been described to show more serrations than C. rubella, while C.
rubella exhibits a larger leaf size than C. pedunculata
based on AR values. According to Peppe et al. (2011),
leaf characters including sizes and shapes strongly correlate to environmental
factors and prove that there is a biological basis for this relationship. The
variations in serrations and leaf size between C. pedunculata
and C. rubella are likely adaptations suited to specific environments.
These distinctive features contribute to the species’ ability to thrive in
different ecological niches. To gain a more comprehensive understanding of the
distribution and evolutionary relationships within the Callicarpa genus,
it is strongly recommended to undertake a thorough phylogenetic study. This
broader investigation will offer valuable insights into the geographic
distribution of Callicarpa species and enhance our understanding of their
adaptive evolution.
Conclusion
A
statistically significant difference in leaf shape between C. pedunculata and C. rubella was observed,
although there is considerable interspecific assessment, possibly due to
environmental and genetic factors. Nevertheless, this study identifies aspect
ratio and circularity as the two most informative variables in discrimination
between the two species, emphasizing the importance of length, width, and leaf
serrations as key diagnostic characteristics. The finding suggests leaf
serrations and leaf size were important to C. pedunculata
and C. rubella, respectively, and considered as an adaptive feature of
the two taxa. Likewise, fine-scale variations in the basal region, e.g.,
presence of leaf lobes, also show significance in the discrimination of the two
taxa. Thus, this research provides new experimental support for future
taxonomic, genetics, or even ecological studies of Callicarpa species in
the relevance of leaf size, leaf serrations, and leaf lobes.
Table 1. Populations and samples of C.
rubella and C. pedunculata were used in
this study.
Species |
Localities |
Accession number |
Herbarium |
C. rubella Lindl. |
Myanmar |
2648823 |
The New York Botanical Garden
(NY) |
|
China |
2787428 |
United States National
Herbarium, Smithsonian Institution (US) |
|
China |
FLAS 269814 |
Florida Museum of Natural
History (FLMNH) |
|
China |
FLAS 269815 |
Florida Museum of Natural
History (FLMNH) |
|
Thailand |
L 0534717 |
Naturalis Biodiversity Center
(AMD) |
|
Thailand |
L 0534080 |
Naturalis Biodiversity Center
(AMD) |
|
Malaysia |
L 2754590 |
Naturalis Biodiversity Center
(AMD) |
|
Malaysia |
L 2754591 |
Naturalis Biodiversity Center
(AMD) |
|
China |
L4212486 |
Naturalis Biodiversity Center
(AMD) |
|
Malaysia |
L0534846 |
Naturalis Biodiversity Center
(AMD) |
|
Vietnam |
P00991455 |
The New York Botanical Garden
(NY) |
|
Taiwan |
K000674727 |
Royal Botanic Gardens Kew (K) |
|
Indonesia |
K000194757 |
Royal Botanic Gardens Kew (K) |
|
Indonesia |
K000194756 |
Royal Botanic Gardens Kew (K) |
|
Vietnam |
MW0756909 |
Moscow State University (MSU) |
|
Vietnam |
MW0757612 |
Moscow State University (MSU) |
|
China |
103972 |
The New York Botanical Garden
(NY) |
|
China |
193971 |
The New York Botanical Garden
(NY) |
|
China |
103960 |
The New York Botanical Garden
(NY) |
|
China |
103959 |
The New York Botanical Garden
(NY) |
|
China |
103961 |
The New York Botanical Garden
(NY) |
|
China |
525329 |
The New York Botanical Garden
(NY) |
|
Vietnam |
2808318 |
The New York Botanical Garden
(NY) |
|
Vietnam |
2808046 |
The New York Botanical Garden
(NY) |
|
Myanmar |
3231815 |
The New York Botanical Garden
(NY) |
C. pedunculata
R.Br. |
Philippines |
JDS001 |
University of Santo Tomas
Herbarium (USTH) |
Table 2. Eigenvalues and contribution of the
first 10 principal components before data partitioning.
Component |
Eigenvalue |
Proportion (%) |
Cumulative (%) |
1 |
0.000682534 |
36.65 |
36.65* |
2 |
0.000385342 |
20.69 |
57.34* |
3 |
0.000166940 |
8.96 |
66.30* |
4 |
0.000144030 |
7.73 |
74.04* |
5 |
0.000106694 |
5.73 |
79.77 |
6 |
0.000085726 |
4.60 |
84.37 |
7 |
0.000063535 |
3.41 |
87.78 |
8 |
0.000048451 |
2.60 |
90.38 |
9 |
0.000029872 |
1.60 |
91.99 |
10 |
0.000025266 |
1.36 |
93.34 |
*Only the first four are significant based on
the broken stick method.
Table 3. The relative contribution of
symmetric and asymmetric components to leaf shape in two Callicarpa
species.
Eigenvalues |
|
|
|
|
|
|
PC1 |
PC2 |
PC3 |
PC4 |
Percentage contribution to
overall shape * |
Symmetric |
6.76 × 10-4 |
1.36 × 10-4 |
1.13 × 10-4 |
8.11 × 10-5 |
54.01% |
Asymmetric |
3.80 × 10-4 |
1.51 × 10-4 |
6.88 × 10-5 |
4.83 × 10-5 |
34.80% |
* Total percentage contribution from PC1 to
PC4 only.
Table 4. Cross-validation matrices from
canonical variates analysis (CVA) of leaf shape in C. pedunculata
and C. rubella.
|
C. rubella |
C. pedunculata |
Total |
% correct |
A. confusion matrix without the
jackknife |
||||
C. rubella |
20 |
4 |
24 |
83 |
C. pedunculata |
5 |
19 |
24 |
79 |
Total |
25 |
23 |
48 |
|
B. confusion matrix with the
jackknife |
||||
C. rubella |
17 |
7 |
24 |
71 |
C. pedunculata |
7 |
17 |
24 |
71 |
Total |
24 |
24 |
48 |
|
Classification
using PC scores computed from the original matrix. B. Jackknife classification.
Computed in
PAST ver. 4.06b. (Hammer et al. 2001).
Table 5. Leaf shape trait values across 46
selected species of (A) C. pedunculata, (B) C.
rubella, and (C) overall accessions.
Trait |
Range |
Mean |
SD |
PCV (%) |
Circularity |
|
|
|
|
(A) |
0.36−0.57 |
0.47 |
0.06 |
13.51 |
(B) |
0.29−0.57 |
0.44 |
0.09 |
19.68 |
(C) |
0.29−0.57 |
0.45 |
0.08 |
16.6 |
Aspect ratio |
|
|
|
|
(A) |
2.05−3.73 |
2.76 |
0.35 |
12.54 |
(B) |
1.90−4.12 |
2.92 |
0.71 |
24.12 |
(C) |
1.90−4.02 |
2.84 |
0.55 |
19.33 |
Solidity |
|
|
|
|
(A) |
0.87−0.97 |
0.94 |
0.02 |
2.46 |
(B) |
0.87−0.97 |
0.93 |
0.03 |
3.28 |
(C) |
0.87−0.97 |
0.94 |
0.03 |
2.89 |
* Only the first PC is significant based on
the broken stick method.
Table 7. Coefficients of correlation among
PC1 to PC3 and the leaf shape descriptors
|
PC 1 |
PC 2 |
PC 3 |
Circ |
0.65380 |
-0.06060 |
0.75423 |
AR |
-0.56176 |
0.62891 |
0.53749 |
Solidity |
0.50692 |
0.77511 |
-0.37714 |
Table 8. Pearson correlation coefficients
between three leaf-shape descriptors.
|
Circularity |
Aspect ratio |
Solidity |
Circularity |
|
3.91 × 10-11 |
6.64 × 10-7 |
Aspect ratio |
-0.8067 |
|
0.10703 |
Solidity |
0.67839 |
-0.31056 |
|
Table 9. Specimen examined. Authors and URLs
of the digital images obtained from the online herbaria used in this study.
Author |
HTTP url |
Bijmoer. R., M. Scherrenberg & J. Creuwels
(2021). Naturalis Biodiversity Center (NL) - Botany. Naturalis Biodiversity
Center. Occurrence dataset https://doi.org/10.15468/ib5ypt accessed via
GBIF.org on 2021-08-29. |
https://www.gbif.org/occurrence/2516551448 |
https://www.gbif.org/occurrence/2516532469 |
|
https://www.gbif.org/occurrence/2516548469 |
|
https://www.gbif.org/occurrence/2517253874 |
|
https://www.gbif.org/occurrence/2516548469 |
|
https://www.gbif.org/occurrence/2516381494 |
|
MNHN & S. Chagnoux (2021). The vascular plants collection (P) at
the Herbarium of the Muséum national d'Histoire Naturelle (MNHN -
Paris). Version 69.223. MNHN - Museum national d'Histoire
naturelle. Occurrence dataset
https://doi.org/10.15468/nc6rxy accessed via GBIF.org on 2021-08-29 |
https://www.gbif.org/occurrence/2270292394 |
Orrell, T &
Informatics Office (2021). NMNH Extant Specimen Records. Version 1.45.
National Museum of Natural History, Smithsonian Institution. Occurrence
dataset https://doi.org/10.15468/hnhrg3 accessed via GBIF.org on 2021-08-29. |
https://www.gbif.org/occurrence/1852124824 |
Perkins, K.D. (2021).
University of Florida Herbarium (FLAS). Version 11.1454. Florida Museum of
Natural History. Occurrence dataset https://doi.org/10.15468/v5wjn7 accessed
via GBIF.org on 2021-08-29 |
https://www.gbif.org/occurrence/2433456102 |
https://www.gbif.org/occurrence/2433458107 |
|
Ramirez, J., K. Watson, B.
Thiers & L. McMillin (2021). The New York Botanical Garden Herbarium
(NY). Version 1.38. The New York Botanical Garden. Occurrence dataset
https://doi.org/10.15468/6e8nje accessed via GBIF.org on 2021-08-29. |
https://www.gbif.org/occurrence/1929638283 |
https://www.gbif.org/occurrence/1930601756 |
|
https://www.gbif.org/occurrence/1930296336 |
|
https://www.gbif.org/occurrence/1930106241 |
|
https://www.gbif.org/occurrence/1929663090 |
|
https://www.gbif.org/occurrence/1929049006 |
|
https://www.gbif.org/occurrence/1928131180 |
|
https://www.gbif.org/occurrence/1929940867 |
|
https://www.gbif.org/occurrence/1931232274 |
|
Royal Botanic Gardens, Kew
(2021). Royal Botanic Gardens, Kew - Herbarium Specimens. Occurrence dataset
https://doi.org/10.15468/ly60bx accessed via GBIF.org on 2021-08-29. |
https://www.gbif.org/occurrence/912528324 |
https://www.gbif.org/occurrence/912176780 |
|
https://www.gbif.org/occurrence/912176785 |
|
Seregin, A. (2021). Moscow
University Herbarium (MW). Version 1.195. Lomonosov Moscow State University.
Occurrence dataset https://doi.org/10.15468/cpnhcc accessed via GBIF.org on
2021-08-29. |
https://www.gbif.org/occurrence/3004116377 |
https://www.gbif.org/occurrence/3004100339 |
FOR
FIGURES & IMAGES - - CLICK HERE FOR FULL PDF
References
Abrammoff, M.D., P.J. Magalhaes
& S.J. Ram (2004). Image Processing with Image Processing with Image. Journal of Biophotonics 11(7): 36–42.
Adebowale,
A., A. Nicholas, J. Lamb & Y. Naidoo (2012). Elliptic Fourier analysis
of leaf shape in southern African Strychnos section Densiflorae (Loganiaceae).
Botanical Journal of the Linnean Society 170(4):
542–553. https://doi.org/10.1111/j.1095-8339.2012.01308
Alonso-Forn, D., D. Sancho-Knapik, J.P. Ferrio, J.J. Peguero-Pina, A.
Bueno, Y. Onoda & E. Gil-Pelegrín (2020). Revisiting the
Functional Basis of Sclerophylly Within the Leaf
Economics Spectrum of Oaks: Different Roads to Rome. Current Forestry Reports
6: 260–281. https://doi.org/10.1007/s40725-020-00122-7
Arvidsson, C. (2020). Callicarpa pedunculata. The IUCN Red List of Threatened
Species 2020: e.T158679327A158679743. https://doi.org/10.2305/IUCN.UK.20203.RLTS.T158679327A158679743.en. Downloaded
on 14 August 2021.
BGCI (2024). PlantSearch.
Botanic Gardens Conservation International. Richmond, U.K. Available at
https://plantsearch.bgci.org. Accessed on 07/02/2024.
Bramley, G.L.C. (2013). The genus Callicarpa (Lamiaceae) in the Philippines. Kew Bulletin 68(3):
369–418.
Bramley, G.L.C. (2019). Flora Malesiana
23: 1-444. Noordhoff-Kolff N.V., Djakarta.
Brown, W.H.
(1920). Minor
Products of Philippine Forests. Department of Agriculture and Natural
Resources. Bureau of Forestry. Vol. 1 Bulletin No. 22. Bureau of Printing,
Manila, 473 pp.
Chitwood,
D.H., A. Ranjan, C. Martinez, L. Headland, T. Thiem,
R. Kumar, M.F. Covington, T. Hatcher, D.T. Naylor, S. Zimmerma,
N. Downs, N. Raymundo, E.S. Buckler, J.N. Maloof, M. Aradhya,
B. Prins, L. Li, S. Myles & N. Sinha (2014). A modern ampelography:
A genetic basis for leaf shape and venation patterning in grape. Plant Physiology
164: 259–272.
Cope, J.S.,
D. Corney, J.Y. Clark, P. Remagnino
& P. Wilkin (2012). Plant species identification using digital morphometrics: A
review. Expert Systems with Applications 39(8): 7562–7573. https://doi.org/10.1016/j.eswa.2012.01.073
Danila, J.S.
& G.J.D Alejandro (2021). Leaf geometric morphometric analyses of Callicarpa and Geunsia (Lamiaceae) in the
Malesian region. Biodiversitas
22: 4379–4390.
Di Marc, M.
& L. Santini (2015). Human pressures predict species’ geographic range size
better than biological traits. Global Change Biology 21(6): 2169–2178. https://doi.org/10.1111/gcb.12834
Godefroy, J.E., F. Bornert,
C.I. Gros & A. Constantinesco (2012). Elliptical Fourier
descriptors for contours in three dimensions: A new tool for morphometrical
analysis in biology. Comptes Rendus
Biologies 335(3): 205–213. https://doi.org/10.1016/j.crvi.2011.12.004
Gupta, S.,
D.M. Rosenthal, J.R. Stinchcombe & R.S. Baucom (2019). The remarkable morphological diversity of
leaf shape in sweetpotato (Ipomoea batatas): the
influence of genetics, environment, and G×E. New Phytologist
225(5): 2183–2195. https://doi.org/10.1111/nph.16286
Iwata, H.
& Y. Ukai (2002). SHAPE: a computer program
package for quantitative evaluation of biological shapes based on elliptic
Fourier descriptors. Journal of Heredity 93: 384–385.
Hammer, Ø.,
D.A.T. Harper & P.D. Ryan (2001). PAST: Paleontological statistics
software package for education and data analysis. Palaeontologia
Electronica 2001(4): 1–9.
Hearn, D.J.
(2009). Shape
analysis for the automated identification of plants from images of leaves. Taxon
58(3): 934–954.
Jones, C.S.
(1995). Does shade
prolong juvenile development: a morphological analysis of leaf shape changes in
Cucurbita argyrosperma ssp. sororia
(Cucurbitaceae). American Journal of Botany
82: 346–359.
Jump, A.S.
& J. Penuelas (2005). Running to stand still:
adaptation and the response of plants to rapid climate change. Ecology
Letters 8(9): 1010–1020. https://doi.org/10.1111/j.1461-0248.2005.00796
Klein, L.L.
& H.T. Svoboda (2017). Comprehensive Methods for Leaf
Geometric Morphometric Analyses. Bio-protocol 7(9): e2269. https://doi.org/10.21769/BioProtoc.2269
Leeratiwong, C., P. Chantaranothai
& A.J. Paton (2009). A synopsis of the genus Callicarpa L. (Lamiaceae)
in Thailand. Thai Forest Bulletin (Botany) 37: 36–58.
Lexer, C., J. Joseph, M. Van Loo, G. Prenner, B. Heinze, M.W. Chase & D. Kirkup
(2009). The use of digital
image based morphometrics to study the phenotypic mosaic in taxa with porous
genomes. Taxon 58: 5–20.
Linnaeus, C.
(1753). Species
plantarum: exhibentes plantas
rite cognitas, ad genera relatas,
cum differentiis specificis,
nominibus trivialibus, synonymis selectis, locis natalibus, secundum systema sexuale digestas. Impensis Laurentii Salvii, Holmiae 1: 560.
MacArthur,
R.H. (1957). On the relative abundance of bird species. Proceedings of the National
Academy of Sciences 43(3): 293–295. https://doi.org/10.1073/pnas.43.3.293
McLellan, T. & J.A. Endler
(1998). The relative
success of some methods for measuring and describing the shape of complex
objects. Systematic Biology 47(2): 264–281.
Mirouze, E., A. Staquet
& R. Vezy (2012). Morphometric and Morphological
analysis of leaves in two species of Araceae: Montrichardia linifera
and Montrichardia arborescens
among different French Guiana populations. Journal of Ecofog
3: 1–15.
Munir, J.
(1982). A taxonomic
revision of the genus Callicarpa in Australia. Journal of the Adelaide
Botanic Gardens. 6: 5–39.
Newbold, T.,
L.N. Hudson, S.L.L. Hill, S. Contu, I. Lysenko, R.A.
Senior & A. Purvis (2015). Global effects of land use on local
terrestrial biodiversity. Nature 520(7545): 45–50. https://doi.org/10.1038/nature14324
Orians, C.M. (2000). The effects of hybridization in
plants on secondary chemistry: implications for the ecology and evolution of
plant-herbivore interactions. American Journal of Botany 87(12): 1749–1756.
Peppe, D.J., D.L. Royer, B. Cariglino, S.Y. Oliver, S. Newman, E. Leight
& I.J. Wright, (2011). Sensitivity of leaf size and
shape to climate: global patterns and paleoclimatic applications. New Phytologist 190(3): 724–739. https://doi.org/10.1111/j.1469-8137.2010.03615
Piazza, P.,
S. Jasinski & M. Tsiantis
(2005). Evolution of
leaf developmental mechanisms. New Phytologist. 167
693–710.
Roxburgh, W. (1820). Flora Indica. Vol. I.
Thacker, Spink & Co., Calcutta, 489 pp.
Royer D.L.
& P. Wilf (2006). Why do toothed leaves correlate with cold climates? Gas exchange at
leaf margins provides new insights into a classic paleotemperature proxy. International
Journal of Plant Sciences 167: 11–18.
Royer, D.L.,
J.C. McElwain, J.M. Adams & P. Wilf (2008). Sensitivity of leaf size and
shape to climate within Acer rubrum and Quercus kelloggii.
New Phytologist.179(3): 808–817.
Sandner, T.M. & D. Matthies (2017). Fluctuating asymmetry of
leaves is a poor indicator of environmental stress and genetic stress by
inbreeding in Silene vulgaris. Ecological Indicators 79:
247–253. https://doi.org/10.1016/j.ecolind.2017.04.030
Thomas, S.C.
& F.A. Bazzaz (1996). Elevated CO2 and leaf shape: Are
dandelions getting toothier? American Journal of Botany. 83: 106–111.
Tsukaya, H., T. Fukuda & J. Yokoyama
(2003). Hybridization
and introgression between Callicarpa japonica and C. mollis
(Verbenaceae) in central Japan, as inferred from
nuclear and chloroplast DNA sequences. Molecular Ecology 12(11): 3003–3011.
https://doi.org/10.1046/j.1365-294x.2003.01961
Tu, S.H.,
C.T. Ho, M.F. Liu, C.S. Huang, H.W. Chang, C.H. Chang, C.H. Wu & Y.S. Ho
(2013). Luteolin sensitises drug-resistant human breast cancer cells to
tamoxifen via de inhibition of cyclin E2 expression. Food Chemistry
141(2): 1553–1561.
Van Valen, L. (1962). A study of fluctuating asymmetry.
Evolution 16: 125–142. https://doi.org/10.1111/j.1558-5646.1962.tb03206
Vieira, M., S.J.
Mayo & I.M. de Andrade (2014). Geometric morphometrics of leaves of Anacardium
microcarpum Ducke and A. occidentale L. (Anacardiaceae)
from the coastal region of Piauí, Brazil. Revista Brasileira
de Botanica 37(3): 315–327. https://doi.org/10.1007/s40415-014-0072-3
Wolfe, A. &
A. Liston (1998). Contributions of PCR-Based
Methods to Plant Systematics and Evolutionary Biology, pp. 43–86. In: Soltis,
D., P. Soltis & J. Doyle (eds.). Molecular Systematics of Plants II.
Springer US, 574 pp.
Yamanaka, T. (1988). On Callicarpa tosaensis. Makino Journal of Japanese
Botany 63: 15–17.
Zhang, J. & P. Li (2019). Response of plant
functional traits to climate change. IOP Conference Series: Earth and
Environmental Science 300(032078): 1–4. https://doi.org/10.1088/1755-1315/300/3/032078