ArticlePDF Available

Morphological Variation in Quinchamalium (Schoepfiaceae) is Associated with Climatic Patterns along Its Andean Distribution

  • INIBIOMA, CONICET, Universidad Nacional del Comahue, Bariloche, Argentina

Abstract and Figures

Quinchamalium (Schoepfiaceae) is a root hemiparasite with a broad geographic range throughout the Andes. Regional studies have used various vegetative and floral traits to describe and identify species, but there has been no detailed analysis of the continuum of morphological variation across the entire geographic range of this genus. Currently 21 species names are being used in the genus but their taxonomic distinctiveness is unclear. The aim of this study was to use multivariate analyses to identify patterns of morphological variation, assess the existence of morpho-species, and correlate variation with climatic and geographic factors. Two putative species were initially circumscribed based on corolla length, and this hypothesis was tested using principal component and discriminant analyses of 17 vegetative and floral characters obtained from 117 herbarium specimens. No statistically significant support was obtained through multivariate analyses for the existence of the two morpho-species, thus, only one species is recognized, a widespread and variable Q. chilense. Patterns of co-variation between several morphological traits and climate were identified. Taller plants with larger flowers were associated with sites with higher precipitation, and narrower leaves with higher temperatures. The presence of thrum flowers (floral morphs with relatively short styles) was correlated with higher latitudes and lower temperatures. Nevertheless, we have not determined whether these variations are genetically fixed ecotypes or are a consequence of phenotypic plasticity.
Content may be subject to copyright.
Systematic Botany (2015), 40(4)
© Copyright 2015 by the American Society of Plant Taxonomists
DOI 10.1600/036364415X690085
Date of publication December 15, 2015
Morphological Variation in Quinchamalium (Schoepfiaceae) is Associated with Climatic
Patterns along its Andean Distribution
Rita M. Lopez Laphitz,
Cecilia Ezcurra,
and Romina Vidal-Russell
Instituto de Investigaciones en Biodiversidad y Medioambiente, CONICET-UNComahue, Quintral 1250,
S. C. de Bariloche, 8400 Río Negro, Argentina.
Author for correspondence (
Communicating Editor: Andrea Weeks
AbstractQuinchamalium (Schoepfiaceae) is a root hemiparasite with a broad geographic range throughout the Andes. Regional studies
have used various vegetative and floral traits to describe and identify species, but there has been no detailed analysis of the continuum of
morphological variation across the entire geographic range of this genus. Currently 21 species names are being used in the genus but their
taxonomic distinctiveness is unclear. The aim of this study was to use multivariate analyses to identify patterns of morphological variation,
assess the existence of morpho-species, and correlate variation with climatic and geographic factors. Two putative species were initially
circumscribed based on corolla length, and this hypothesis was tested using principal component and discriminant analyses of 17 vegetative
and floral characters obtained from 117 herbarium specimens. No statistically significant support was obtained through multivariate analy-
ses for the existence of the two morpho-species, thus, only one species is recognized, a widespread and variable Q. chilense. Patterns of
co-variation between several morphological traits and climate were identified. Taller plants with larger flowers were associated with sites
with higher precipitation, and narrower leaves with higher temperatures. The presence of thrum flowers (floral morphs with relatively short
styles) was correlated with higher latitudes and lower temperatures. Nevertheless, we have not determined whether these variations are
genetically fixed ecotypes or are a consequence of phenotypic plasticity.
KeywordsBioclimatic variables, ecotypes, geography, heterostyly, morphological continuum, morpho-species, phenotypic plasticity.
Identifying and delimiting species has always been a com-
plex endeavor in systematics. The question of what defines a
species is contentious, as evidenced by the existence of a large
number of different species concepts (de Queiroz 2007). In
systematics, being able to efficiently and accurately delimit
species is basic because this taxonomic level is the funda-
mental study unit of different fields such as biogeography,
ecology, and conservation. However, one of the major
problems linked to species delimitation has been to distin-
guish between species concepts and species criteria (Wiens
and Servedio 2000). One of the practical yet conservative
strategies taken by many researchers has been to apply
evidence from diverse sources to support the recognition of
species. These include fixed or non-overlapping differences
in morphological, behavioral, or ecological characters, mole-
cular divergence, or geographic isolation. In this study, we
recognize non-overlapping patterns of morphological varia-
tion as the primary criterion for inferring species boundaries.
This is based on the idea that morphological discontinuities
suggest that some evolutionary force may be preventing two
distinct lineages from homogenizing (Stuessy 1990; Wiens
and Servedio 2000). Morpho-species (Cronquist 1978) or
pheno-species (Sokal and Sneath 1963) are those concepts that
follow this basic criterion that has been applied in most of
the morphological systematic studies (e.g. Lehnebach 2011;
Lopez Laphitz et al. 2011; Nagahama et al. 2014). As posed
by Wheeler and Meier (2000), even though a species
definition may or may not make reference to characters
(morphological or not), all concepts use character data to infer
species boundaries, and most of those data are morphological.
In this sense, we use the morpho-species criterion in an
empirical example focusing on Quinchamalium Molina, a
morphologically variable, geographically extended, and tax-
onomically understudied genus from southern South America.
Quinchamalium is a small genus of yellow-flowered hemi-
parasitic perennial herbs (Fig. 1) distributed throughout
the Andes in open habitats across a wide elevational range
(03,800 m a.s.l.). The genus was established by Molina in
1782 based on a specimen from Chile, and since then, 33 species
names have been published within it. At present, Quinchamalium
is thought to include approximately 21 species (Brako and
Zarucchi 1993; Zuloaga et al. 2009; Jørgensen et al. 2014;
Appendix 1), but the limits between these species are in
many cases obscure (Dawson 1944). The genus is distrib-
uted from northern Peru to southern Chile, Argentina and
Bolivia, covering a wide geographic area (Appendix 1). Despite
this, the 21 accepted South American species mentioned
before have not yet been treated in one single taxonomic
study that includes them all. Consequently, no diagnostic
key is available where the total number of recognized species
of Quinchamalium are comparable. In the past, taxonomic
treatments (Presl 1849; Philippi 1857; Miers 1880) included
quantitative morphological traits related to flower and leaf
morphology. More recently, Dawson (1944) and Navas (1976)
developed regional taxonomic keys including some other
quantitative characters. However, none of these taxonomic
studies performed numerical analyses based on these quan-
titative traits when identifying key characters. In addition,
there are recent examples in Quinchamalium where taxo-
nomic distinctions were focused only on specimens from
limited areas (Navas 1976: Cuenca de Santiago; Dawson
1984: Patagonia; Ulibarri 1994: San Juan, Argentina). Such
studies may unnecessarily split a taxon into multiple taxa.
On the other hand, a preliminary morphological examina-
tion of Quinchamalium specimens revealed considerable con-
tinuous variation in the characters used to differentiate these
purported species.
Geographic variation in plant morphology can be a func-
tion of genetic changes (e.g. Quiroga et al. 2002; Premoli
et al. 2007) or of phenotypic plasticity (e.g. Sultan 2000) in
response to local environmental conditions. For this reason,
the study and interpretation of patterns of geographic varia-
tion in morphological characters can potentially solve taxo-
nomic problems as well as help to infer environmental
effects. Moreover, the study of variation within widespread
species at large spatial scales is the first step in determining
the relative importance of different factors that promote
phenotypic differentiation (Ezcurra et al. 1997; Mascó et al.
2004; Chalcoff et al. 2008). Quinchamalium offers an excel-
lent opportunity to test if the polymorphism it shows is
interrupted by morphological gaps that could represent
species discontinuities, or if its diversity represents environ-
mentally induced morphological variation within a single
species. Additionally, as expected in plants with extensive
areas of distribution, this study could provide an opportu-
nity to observe evolution in progress and understand the
origin of systematic diversity.
Therefore, the main objective of this study is to clarify whether
Quinchamalium is composed of more than one taxon or only
a single polymorphic species. If the first scenario is true, we
aim to find diagnostic characters that could be use to diag-
nose species or subspecies. If the second scenario is true, we
attempt to identify biologically relevant climatic or geographic
factors that may be influencing morphological variation.
To accomplish these objectives and to provide support for
a taxonomic revision of the group, the present study evalu-
ates the morphological variation existing in the genus. As
with species complexes in many genera of angiosperms, we
addressed this problem using multivariate analyses.
Materials And Methods
Taxon Sampling for Species Delimitation Approximately 750 her-
barium specimens from BAB, BCRU, CONC, CORD, LP, MO and SI
were examined to identify useful morphological traits for species delimi-
tation. Many of these specimens could not be included in a morphomet-
ric study because of their inappropriate phenological stage (e.g. young
emerging inflorescence that had not yet expanded, late fruiting speci-
mens, or missing key floral parts), so adequate complete specimens were
selected from 117 localities distributed throughout the geographic range
of Quinchamalium,i.e.fromnorthernPeru(ca.7.1°S)tosouthernArgentina
(ca. 50.4° S), from western Chile (ca. 73.35° W) to western Argentina
(ca. 69.57° W), and from sea level to 3,800 m a.s.l., including the Bolivian
Fig. 1. Morphological variation in plants of Quinchamalium. A. Bariloche, Argentina. B. Paposo, Chile. C. Portillo, Chile. D. Huasco, Chile. Photo-
graphs by C. Calviño and R. Vidal Russell.
Puna. These specimens were chosen to be representative of the whole
morphological and ecological variation found in herbarium material.
Traits MeasurementsUsing previous analyses on the genus, descrip-
tions of the taxa found in the literature, and personal observations, a list
of potentially useful distinguishing traits was created. In the past, most
of the original species descriptions in Quinchamalium were based upon
morphological characters such as the shape of the leaves (Miers 1880)
and the size of the flowers (Presl 1849; Philippi 1857). More recently,
other characters have been used to distinguish species such as size of
the floral bracts (Dawson 1944), and length of the free filament (Navas
1976). Therefore, in total, we used 17 morphological and floral traits that
were scored on specimens usually pressed during anthesis. Measure-
ments are based on rehydrated herbarium material, using hot water and
detergent. The 17 variables were (Table 1; abbreviation given between
parentheses): stem length (ST), leaf length (LFL), leaf width (LFW), leaf
length: leaf width ratio as an estimator of leaf shape (LFSH), corolla
length (CL), corolla width (CW), corolla lobe length (LOBL), corolla lobe
width (LOBW), calyculus length (CAL), calyculus teeth length (CTL),
total filament length (FIL), free filament length (FRFIL), nectary length
(NL), anther length (AL), anther width (AW), style length (SL), and corolla
length: style length ratio as an estimator of floral morph (FM). Because of
considerable diversity in definitions of floral parts by different authors,
floral characters measured are illustrated in Fig. 2.
Heterostylous species have two (distylous) or three (tristylous) floral
morphs that differ reciprocally in the placement of their anthers and
stigmas (Darwin 1877; Barrett 1992). Rivero et al. (1987) have reported
distyly in the genus among populations of Quinchamalium of Volcán
Casablanca in Chile. Therefore, to study the association between hetero-
styly and morphological, geographic and climatic variables, we calcu-
lated corolla length: style length ratio (floral morph) and implemented
it as one of the 17 morphological quantitative traits used in the multi-
variate analyses (FM, Table 1).
Putative Species Delimitation Because taxa limits are unclear and
no key is available at present that includes the total number of spe-
cies currently recognized, a preliminary hypothesis was tested. For this,
two putative species were defined after a two-step procedure. First, every
quantitative trait distribution was examined to identify potential vari-
ables without continuous variation and with gaps to differentiate among
species. Only one variable (corolla tube length) was found with a
bi-modal distribution. Wilcoxon tests were made to detect significant
differences between these two putative species for the rest of the mor-
phological variables. Second, in an exploratory principal components
analysis (PCA) including all the continuous variables, the character that
most contributed to the first axis was corolla length. Therefore, two puta-
tive species were proposed: plants with corolla lengths 9 mm in length
(Q. chilense Molina), and those with corolla lengths < 9 mm in length
(Q. parviflorum Phil.). To apply the oldest available species names
within the genus to these two species, the types and protologues of all
published names were consulted.
Morphometric Analyses To detect homogeneous groups among speci-
mens and extract the variables that best diagnose these groups, PCA was
performed. Variable contribution (loading values) to the PCA axes were
interpreted as significant when 0.6. Before conducting this analysis, all
measurements were standardized. In addition, discriminant analysis (DA)
of quantitative morphological variables was used to discriminate between
the two pre-classified putative species. Prior to DA, the data matrix was
modified in several ways. Because some quantitative characters were not
normally distributed, they were log
transformed before the analyses.
Traits that strongly correlated with other traits as determined by having a
Pearsons correlation coefficient r > |0.7| were identified, and one of the
traits of the correlating pair was excluded from the analysis (ST, LFSH,
LOBL, CTL, FIL, FRFIL, AL, AW, SL, and FM). The trait corolla length
(CL) was excluded because it was used previously in the identification of
the two putative species. Finally, a subset of six continuous morphological
variables was selected to conduct the DA (LFL, LFW, CW, LOBW, CAL,
NL). In this study we considered a posteriori values of misclassification
higher than 15% as indicator of low discriminant power. This threshold is
commonly used in taxonomic studies (e.g. Lehnebach 2011). The PCA and
the DA were carried out using the program JMP 9.0 (SAS Institute Inc.
Cary, North Carolina).
Geographic and Climatic VariablesTo determine the geographic
range and climatic characteristics of the two putative species, collection
data from herbarium specimens were examined, latitude and longitude
Table 1. Summary of comparison of morphometric variables in two putative species of Quinchamalium. Significance levels: *p< 0.05; **p< 0.01;
***p< 0.001.
N° Characters Q. chilense (N = 80) Q. parviflorum (N = 37) χ
Vegetative characters
1 Stem length (ST) 20.28 ± 0.8 (7.540) 10.59 ± 0.79 (2.528) 34.34***
2 Leaf length (LFL) 17.79 ± 0.52 (8.433) 17.22 ± 1.27 (7.647.4) 0.92
3 Leaf width (LFW) 1.08 ± 0.03 (0.371.88) 0.85 ± 0.03 (0.361.78) 13.86***
4 Leaf length: leaf width ratio (leaf shape) (LFSH) 18.01 ± 0.7 (643) 22.78 ± 1.7 (8.4965) 4.92*
Floral characters
5 Corolla length (CL) 11.41 ± 0.16 (918) 6.62 ± 0.13 (58) 76.35***
6 Corolla width (CW) 0.9 ± 0.01 (0.431.29) 0.73 ± 0.01 (0.571) 30.67***
7 Corolla lobe length (LOBL) 3.89 ± 0.06 (2.25.39) 2.54 ± 0.08 (1.664.3) 60.86***
8 Corolla lobe width (LOBW) 1.12 ± 0.01 (0.691.64) 0.76 ± 0.02 (0.51.36) 58.08***
9 Calyculus length (CAL) 2.8 ± 0.04 (1.763.9) 2.5 ± 0.05 (1.873.23) 17.46***
10 Calyculus teeth length (CTL) 1 ± 0.02 (0.22) 0.68 ± 0.02 (0.321.3) 40.20***
11 Total filament length (FIL) 7.54 ± 0.18 (3.811) 4.77 ± 0.13 (37) 59.19***
12 Free filament length (FRFIL) 0.52 ± 0.03 (01.75) 0.36 ± 0.03 (0.11) 1.02
13 Nectary length (NL) 0.54 ± 0.02 (0.181.14) 0.49 ± 0.02 (0.21) 2.01
14 Anther length (AL) 1.48 ± 0.03 (0.822.77) 0.82 ± 0.04 (0.181.54) 56.52***
15 Anther width (AW) 0.37 ± 0.007 (0.230.65) 0.26 ± 0.01 (0.140.44) 35.93***
16 Style length (SL) 8.62 ± 0.22 (3.817.5) 5.2 ± 0.20 (39.3) 43.83***
17 Corolla length: style length ratio (floral morph) (FM) 1.42 ± 0.04 (0.93.25) 1.32 ± 0.03 (0.751.81) 0.45
Fig. 2. Floral parts of Quinchamalium chilense included in morpho-
logical analyses. Left, closed flower. Right, open flower. CL: Corolla tube
length. CW: Corolla tube width. LOBL: Corolla lobe length. LOBW:
Corolla lobe width. CAL: Calyculus length. CTL: Calyculus teeth length.
FIL: Filament length. FRFIL: Free filament length. NL: Nectary length.
AL: Anther length. AW: Anther width. SL: Style length. Redrawn by
C. Ezcurra from illustrations in Pilger (1935) and Rivero et al. (1987).
coordinates determined, and the distributions of each species plotted.
For each specimens locality, environmental information represented by
19 bioclimatic variables (Table 3) derived from monthly temperature and
rainfall values with a spatial resolution of 2.5 arc-minutes (5 km
) were
downloaded from the WorldClim 1.4 database (http://www.worldclim
.org/current; Hijmans et al. 2005). These environmental variables were
extracted for each locality using QGIS 1.8.0-Lisboa (© 20022011 QGIS
Development Team).
Association of Morphology with Climatic and Geographic Variation
We first measured the degree of association of the morphological vari-
ables with the climatic and geographic data, and among each type of
data (Spearman correlation coefficients). Because of high correlations
among climatic variables, we performed a PCA to generate two principal
axes that were not correlated and accounted for the majority of variance
of the climatic data. All variables were standardized (transformed into
standard-deviation units) prior to performing the PCA. All individual
morphological variables and the first two axes scores of the morphologi-
cal PCA were regressed onto the first two axes scores of the climatic
PCA. These analyses were performed to illustrate patterns of covariance
between morphological traits and climate across the entire geographic
range of Quinchamalium.
Putative Species DelimitationNone of the morphologi-
cal variables explored with specimen frequency histograms
showed gaps among groups of specimens, and corolla length
was the only morphological character that showed a bi-modal
distribution (Fig. 3). Two high frequency peaks resulted at
two corolla length values (6 mm and 11 mm). The geographic
distribution of the putative species, Quinchamalium chilense
and Q. parviflorum is mapped in Fig. 4.
The mean values of corolla lengths were significantly dif-
ferent (Wilcoxon = 76.35, p< 0.0001), and separated two
groups. Of the 17 morphological variables, 13 were also sig-
nificantly different between putative species (Table 1).
Morphology: Principal Component AnalysisThe first prin-
cipal component (PC1) accounted 41.9% of the morphologi-
cal variation, the second (PC2), 15.14%, and the third (PC3),
only 8% of the total variance of the data. A scatterplot of the
scores of principal components 1 and 2 (Fig. 5) partially sep-
arates two groups. Component 1 provides an incomplete
separation between Q. chilense and Q. parviflorum. Individ-
uals of Q. parviflorum (triangles) are dispersed entirely within
the left side and have negative values for PC1 (-50). Indi-
viduals of Q. chilense (circles) are mainly positioned within
the right side of the plane in a range of values of PC1 (-26).
Fourteen of the 17 variables employed in the PCA have
relatively high loadings (absolute values >0.6) on at least
one of the first two components (Table 2). The first PC
expressed mainly variation in plant and flower part size, e.g.
stem length (ST), corolla length (CL), corolla lobe length
(LOBL), anther length (AL), corolla lobe width (LOBW),
corolla width (CW), and style length (SL; see Table 2 for
all loadings), and tended to separate the individuals of the
two putative species (Fig. 5). The second PC expressed mainly
variation in leaf shape (LFSH), floral morph (FM), and free
filament length (FRFIL; see Table 2 for all loadings). How-
ever, in this case, this component did not separate the two
putative species (Fig. 5).
Morphology: Discriminant AnalysesIn the discriminant
analysis (DA), the function was significant (Function1: Wilks
Lambda = 0.511, F = 17.2, p< 0.0001). Overall, there is clear
overlap in the symbols of the two groups, but no overlap in
the 95% confidence limits around group centroids (Fig. not
shown). The DA of the two groups classified 83% of the speci-
mens correctly into the groups to which they were a priori
assigned to test the morpho-species hypothesis. Sixteen of the
80 individuals classified a priori as Q. chilense were misclas-
sified, and four of the 38 individuals identified a priori as
Q. parviflorum were misclassified.
Climate: Principal Component Analyses Climatic data
were well summarized by the first two principal compo-
nents of PCA, which together accounted for 67% of the vari-
ation. Thirteen of the 19 variables employed in the PCA had
relatively high loadings (absolute values >0.6) on one of the
first two components (Table 3). The first PC of climate was
dominated by temperature, with highest contributions of
mean temperature of coldest quarter (BIO11), annual mean
temperature (BIO1), mean temperature of wettest quarter
Fig. 3. Frequency histogram of corolla lengths of 117 plants of
Quinchamalium in 1 mm intervals.
Fig. 4. Geographic distribution of 117 specimens studied of two puta-
tive species of Quinchamalium on southwestern South America (Perú,
Bolivia, Chile, Argentina, Paraguay and Uruguay). Circles represent
Q. chilense and triangles represent Q. parviflorum.
(BIO8), minimum temperature of coldest period (BIO6),
and mean temperature of warmest period (BIO10; see
Table 3 for all loadings). However, the PC2 of climate was
dominated by precipitation factors: annual (BIO12), wet-
test quarter (BIO16), driest quarter (BIO17), wettest month
(BIO13), coldest quarter (BIO19), and driest month (BIO14;
see Table 3 for loadings).
Morphological Variation, Geography and Climate All
composite and some of the individual morphological vari-
ables were best predicted by climatic variables (Fig. 6A, B).
The second principal axis of climate (dominated by precipi-
tation; Table 3) was the best predictor of the first morpho-
logical principal axis (dominated by corolla length; Fig. 6A).
In contrast, this second principal axis of climate was not a
good predictor of the second morphological principal com-
ponent (results not shown). The first axis of climate (domi-
nated by temperature; Table 3) was the best predictor of the
second morphological principal component (dominated by
leaf shape and floral morph; Fig. 6B). However, it was not a
good predictor of the first morphological principal compo-
nent (results not shown).
Floral morph was calculated as corolla length: style length
ratio. Following Rivero et al. (1987), we interpreted flowers
with small FM values as pin flowers (with relatively long
styles), and flowers with high values of FM as thrum flowers
(with relatively short styles). Therefore thrum flowers and
wider leaves were more common in cooler sites while plants
with pin flowers and narrower leaves were more predomi-
nant in warmer sites (Fig. 6B). In the same way, in general
plants from wetter sites had larger flowers and were taller
while plants of drier sites had smaller flowers and were
smaller (Fig. 6A).
The degree of association between morphological vari-
ables and geographic data was dominated by the correla-
tion of latitude with floral morph; in fact the most significant
correlation among variables was between FM and latitude
(Spearmans=0.31, p= 0.0006). Despite this, the two
putative species with different floral sizes (Q. chilense and
Q. parviflorum) are sympatric nearly the entire latitudinal
and longitudinal distribution of Quinchamalium (Fig. 4).
This study presents the results of numerical analyses of
morphological variation of vegetative and floral characters
in Quinchamalium, a taxonomically understudied genus from
the central and southern Andes. The existence of two puta-
tive species characterized by different corolla lengths was
investigated. Although several other correlated traits showed
significant differences between the two putative species, the
Table 2. Results of the principal component analysis (PCA) of the
17 morphological variables measured for 117 Quinchamalium specimens.
The first two PCA axes accounted for 57% of the variance in the data.
Values shown are the loadings of each variable on each of the first two
PCA axes, with the highest contributing variables (>0.6) indicated in bold.
Variables Description Axis1 Axis2
ST Stem length 0.651 0.178
LFL Leaf length 0.288 0.575
LFSH Leaf shape 0.159 0.776
LFW Leaf width 0.493 0.437
CL Corolla tube length 0.938 0.078
CW Corolla tube width 0.777 0.110
LOBL Corolla lobe length 0.885 0.130
LOBW Corolla lobe width 0.810 0.298
CAL Calyculus length 0.583 0.066
CTL Calyculus teeth length 0.691 0.278
FIL Filament length 0.749 0.086
FRFIL Free filament length 0.471 0.639
NL Nectary length 0.244 0.226
AL Anther length 0.858 0.026
AW Anther width 0.700 0.172
SL Style length 0.734 0.447
FM Floral morph 0.036 0.713
Fig. 5. Scatterplot of the first two principal components (PC1 and
PC2) of the principal component analysis (PCA) of the 117 specimens of
Quinchamalium studied based on 17 quantitative morphological vari-
ables. The variance explained associated with each PC is provided in
parentheses. Symbols indicate the two putative species: Q. chilense cir-
cles, and Q. parviflorum triangles.
Table 3. Results of the principal component analysis of the 19 cli-
matic variables of 117 specimens of Quinchamalium studied. The first two
principal axes accounted for 67% of the variance in the data. Values
shown are the loadings of each variable on each of the first two axes.
Variables Description Axis 1 Axis 2
BIO1 Annual mean temperature 0.932 0.343
BIO2 Mean diurnal temperature range 0.219 0.365
BIO3 Isothermality 0.342 0.307
BIO4 Temperature seasonality 0.587 0.138
BIO5 Maximum temperature of warmest period 0.695 0.489
BIO6 Minimum temperature of coldest period 0.872 0.462
BIO7 Temperature annual range 0.566 0.139
BIO8 Mean temperature of wettest quarter 0.923 0.164
BIO9 Mean temperature of driest quarter 0.718 0.494
BIO10 Mean temperature of warmest quarter 0.834 0.438
BIO11 Mean temperature of coldest quarter 0.948 0.284
BIO12 Annual precipitation 0.393 0.871
BIO13 Precipitation of wettest month 0.357 0.792
BIO14 Precipitation of driest month 0.454 0.783
BIO15 Precipitation seasonality 0.419 0.442
BIO16 Precipitation of wettest quarter 0.343 0.816
BIO17 Precipitation of driest quarter 0.470 0.794
BIO18 Precipitation of warmest quarter 0.012 0.358
BIO19 Precipitation of coldest quarter 0.446 0.790
multivariate analyses and the geographic distribution did
not support them as different.
The results of PCA showed a partial separation of these
two species, Q. chilense being generally characterized by larger
flowers and taller plants, and Q. parviflorum by smaller
flowers and shorter plants. However, multivariate analyses
of quantitative characters were not useful for discriminat-
ing clearly among these two putative morpho-species, as
evidenced by the overlap of their groups shown in the PCA
scatterplot (Fig. 5) and the relatively low success rate (83%
in the classification matrix) in the DA of the two putative
species. In conclusion, and based on the practical criterion
of morpho-species, our data do not support the existence
of these two hypothesized species, inferring that only one
single polymorphic species can be identified in the genus
(Q. chilense; Lopez Laphitz et al. 2015).
Other important characters in relation to the morphological
variability of the species were leaf shape and floral morph as
shown by significant loadings in PC2, but these traits also
did not support species separation. In the past, several authors
(Presl 1849; Philippi 1857; Miers 1880; Dawson 1944; Navas
1976) already used leaf shape and floral size in taxonomic
treatments of Quinchamalium, but they did not include numeri-
cal analyses. Therefore, our findings are the first ones that
use these characters in quantitative multivariate analyses to
determine if morpho-species exist. As additional evidence,
the distribution map (Fig. 4) clearly shows that there are no
geographic distinctions between the hypothesized morpho-
species (Q. chilense and Q. parviflorum).
Considering the recognition of a single polymorphic and
extended Q. chilense, geographic and climatic patterns asso-
ciated with this morphological variation were studied. A
geographical perspective is particularly important for spe-
cies with broad distributions because they are likely to experi-
ence a wide variety of environmental conditions (e.g. Hodgins
and Barrett 2008; Nicola et al. 2014). Significant associa-
tions between climate and morphology were observed for
Quinchamalium chilense, e.g. sites with lower precipitations
tended to have shorter plants with smaller flowers, and
sites with higher temperatures had plants with relatively
narrower leaves and pin flowers with relatively long styles.
These types of differences suggest the effect of aridity gradi-
ents resulting from the combination of higher temperatures
and reduced precipitations (e.g. Ezcurra et al. 1997; Quiroga
et al. 2002). Larger plant and flower size associated with
higher rainfall and more favorable climatic conditions for
growth have been also observed in other plant groups
(Hodgins and Barrett 2008).
By altering traits through phenotypic plasticity in response
to changes in environmental conditions, plants are able to
respond adaptively to a range of environments and thus use
a wider range of habitats than would be possible if all traits
were genetically fixed (Sultan 2000). But also, widespread
species are likely to face a variety of conditions throughout
their ranges that through natural selection can result in
genotypic differentiation (Gaston 2003). Based on our results
we cannot distinguish between these two alternatives for
Q. chilense. The large morphological variation of this spe-
cies resulting from heterogeneous environmental conditions
could be the consequence of either genetic differences as
inferred in the herb Cerastium arvense (Caryophyllaceae;
Quiroga et al. 2002) or in the tree species Nothofagus pumilio
(Nothofagaceae: Premoli 2003; Premoli and Brewer 2007;
Premoli et al. 2007), or phenotypic plasticity as described in
Ceanothus (Rhamnaceae; Pugnaire et al. 2006) and Psychotria
(Rubiaceae; Valladares et al. 2000). A common-garden experi-
ment could differentiate between the two, but until now,
cultivation of Q. chilense has not been successful; seeds failed
to germinate and transplanted plants died after repeated
attempts (Lopez Laphitz, pers. obs.).
Another important result of this study is the relation
between floral morph and temperature. First, we observed
Fig. 6. Relationship between plant morphology and climate. A. The first morphological axis dominated by flower size and stem length (see Table 2
for all loadings) as a function of climate. Symbols are as in Fig. 4. The line is the best-fit linear regression. Slope = 0.29, r
= 0.07, p= 0.0037. In A, the
predictor variable (x-axis) is the second principal axis of the PCA of climate, which is dominated by precipitation (see Table 3 for all loadings). B. The
second morphological principal axis dominated by leaf shape and floral morph (see Table 2 for all loadings) as a function of climate. The line is
the best-fit linear regression. Slope = 0.11, r
= 0.037, p= 0.036. In B, the predictor variable (x-axis) is the first principal axis of PCA of climate, which
is dominated by temperature. (see Table 3 for all loadings).
that geographic latitude was highly correlated with floral
morph; the proportion of thrum floral morphs increasing to
the south and pin morphs increasing to the north. At the
same time, individuals with thrum flowers were not found to
inhabit latitudes lower than 28° S (northern Chile: Huasco).
Hence, northern regions (Peru, Bolivia and northern Chile)
have exclusively pin flowers (long styles), while both morphs
appear to be co-inhabitants of the southern portion of the
distribution range of the species. It has been observed that
similarly, many style-dimorphic Narcissus (Amaryllidaceae)
display great variation in floral morph ratio, from dimorphic
populations (pin: thrum or pin-biased) to pin-monomorphic
populations in different geographic areas (Hodgins and Barrett
2008; Pérez-Barrales et al. 2014). This variation is frequently
associated with shifts in pollinators, and the absence of one
morph seems to be a derived condition (Hodgins and Barrett
2008; Pérez-Barrales et al. 2014). Secondly, as shown by the
regression between the second axis of the morphological PCA
and the first axis of the climatic PCA (Fig. 6A), at higher
temperatures the corolla: style length ratio decreases, with
a tendency to produce only pin flowers. Different patterns of
flower morphology and floral-morph ratio have been found
when plants are under selection by different functional
groups of pollinators and in different climates (Chalcoff
et al. 2008; Hodgins and Barrett 2008; Pérez-Barrales et al.
2014). Thus, temperature could be affecting floral morph
indirectly by determining different pollinator assemblages
in colder or warmer areas. Therefore, the different morph
frequencies could be the result of specialization to a frac-
tion of the environmental heterogeneity (abiotic or biotic)
through selection processes, resulting in specialized taxa
or ecotypes.
In conclusion, Quinchamalium chilense,asingle,polymor-
phic, widespread species, shows a continuum of morpho-
logical variation because of either genotypic variation or
phenotypic plasticity in response to environmental factors
associated with latitude and climate. Considering the uni-
fied species concept (de Queiroz 2007), we cannot discard
that Q. chilense may be showing different ecotypes or is in an
early stage of species differentiation, with our results only
presenting one view of this divergence. Future research com-
bining molecular data could give support to the recognition
of monophyletic groups with geographic, climatic and/or
morphological coherence within this variable species.
Acknowledgments. The authors thank A. Sersic and E. Urtubey
for suggesting improvements during the first authors production of
her doctoral thesis, and to two anonymous reviewers and J. Smith, editor-
in-chief of Systematic Botany, for valuable comments. We are grateful to the
curators of the following herbaria for access to collections: BAB, BCRU,
CONC, CORD, LP, MO, and SI. The first author thanks the Argentina
Research Council (CONICET) for a Doctoral fellowship. This research
was made possible through grants from the following scientific institu-
tions in Argentina: ANPCyT PICT 2007-1047, and 2011-1036, CONICET
PIP 0282, and UNComahue PIN B149.
Literature Cited
Barrett, S. C. H. 1992. Evolution and function of heterostyly. Ed. 1. Berlin,
Germany: Springer Verlag.
Brako, L. and J. L. Zarucchi. 1993. Catalogue of the Flowering Plants
and Gymnosperms of Peru. Monographs in Systematic Botany of the
Missouri Botanical Garden 45: 11286. Saint Louis, Missouri: Missouri
Botanical Garden.
Chalcoff, V., C. Ezcurra, and M. Aizen. 2008. Uncoupled geographical
variation between leaves and flowers in a SouthAndean Proteaceae.
Annals of Botany 102: 7991.
Cronquist, A. 1978. Once again, what is a species? Pp. 320 in BioSystematics
in Agriculture, ed. J. A. Romberger. New Jersey: Allaheld & Osmun.
Darwin, C. 1877. The different forms of flowers on plants of the same
species. London: John Murray.
Dawson, G. 1944. Las santaláceas argentinas. Revista del Museo La Plata 6:
580. (Nueva Serie).
Dawson, G. 1984. Santalaceae. Pp. 3042 in Flora Patagónica. ed. M. N.
Correa. Buenos Aires: Colección Científica del Instituto de Tecnología
Agropecuaria vol. 8 (4a).
Ezcurra, C., A. Ruggiero, and J. V. Crisci. 1997. Phylogeny of Chuquiraga
sect. Acanthophyllae (Asteraceae) and the evolution of its leaf mor-
phology in relation to climate. Systematic Botany 22: 113.
Gaston, J. K. 2003. The structure and dynamics of geographic ranges. Ed. 1.
New York: Oxford University Press.
Hijmans, R. J., S. E. Cameron, J. L. Parra, P. G. Jones, and A. Jarvis. 2005.
Very high resolution interpolated climate surfaces for global land
areas. International Journal of Climatology 25: 19651978.
Hodgins, K. A. and S. C. Barrett. 2008. Geographic variation in floral
morphology and stylemorph ratios in a sexually polymorphic daf-
fodil. American Journal of Botany 95: 185195.
Jørgensen, P. M., M. H. Nee, and S. G. Beck. 2014. Catálogo de las
plantas vasculares de Bolivia. Monographs in Systematic Botany of
the Missouri Botanical Garden 127: 11744. Saint Louis, Missouri:
Missouri Botanical Garden Press.
Lehnebach, C. 2011. Re-evaluating species limits in Uncinia angustifolia,
U. caespitosa s. str., U. rupestris, U. zotovii and U. viridis (Cyperaceae).
Australian Systematic Botany 24: 405420.
Lopez Laphitz, R. M., Y. Ma, and J. C. Semple. 2011. A multivariate
study of Solidago subsect. Junceae and a new species in South America
(Asteraceae: Astereae). Novon: A Journal for Botanical Nomenclature
21: 219225.
Lopez Laphitz, R. M., C. Ezcurra, and R. Vidal-Russell. 2015. Revisión
taxonómica del género sudamericano Quinchamalium (Schoepfiaceae).
Boletín de la Sociedad Argentina de Botánica 50: 235246.
Mascó, M., I. Noy-Meir, and A. N. Sérsic. 2004. Geographic variation in
flower color patterns within Calceolaria uniflora in southern Patagonia.
Plant Systematics and Evolution 244: 7791.
Miers, J. 1880. On the Schoepfieae and Cervantesieae, distinct tribes of the
Styracaceae. Journal of the Linnean Society - Botany 17: 6887.
Nagahama, N., A. M. Anton, and G. A. Norrmann. 2014. Taxa delimita-
tion in the Andropogon lateralis complex (Poaceae) in southern South
America based on morphometrical analysis. Systematic Botany 39:
Navas, L. E. 1976. FloradelacuencaSantiagodeChileTomo 2: 40. Santiago,
Chile: Ediciones de la Universidad de Chile.
Nicola, M. V., L. A. Johnson, and R. Pozner. 2014. Geographic variation
among closely related, highly variable species with a wide distribution
range: the South Andean-Patagonian Nassauvia subgenus Strongyloma
(Asteraceae, Nassauvieae). Systematic Botany 39: 331348.
Philippi, R. A. 1857. Ueber die chilenischen Formen von Quinchamalium.
Botanische Zeitung 15: 745749.
Pilger, R. 1935. Santalaceae. Pp. 5291 in Die natürlichen Pflanzenfamilien
Vol. 16b. eds. A. Engler and K. Prantl. Leipzig: Engelmann.
Pérez-Barrales, R., V. I. Simón-Porcar, R. Santos-Gally, and J. Arroyo.
2014. Phenotypic integration in style dimorphic daffodils (Narcissus,
Amaryllidaceae) with different pollinators. Philosophical Transactions of
the Royal Society of London. Series B, Biological Sciences 369: 20130258.
Premoli, A. 2003. Isozyme polymorphisms provide evidence of clinal
variation with elevation in Nothofagus pumilio. The Journal of Heredity
94: 218226.
Premoli, A. and C. Brewer. 2007. Environmental v. genetically driven
variation in ecophysiological traits of Nothofagus pumilio from con-
trasting elevations. Australian Journal of Botany 55: 585591.
Premoli, A. C., E. Raffaele, and P. Mathiasen. 2007. Morphological and
phenological differences in Nothofagus pumilio from contrasting eleva-
tions: Evidence from a common garden. Austral Ecology 32: 515523.
Presl, C. 1849. Epimeliae Botaniae. Pp. 246. Pragae: Amadei Haase.
Pugnaire, F. I., F. S. Chapin, and T. M. Harding. 2006. Evolutionary
changes in correlations among functional traits in Ceanothus in
response to Mediterranean conditions. Web Ecology 6: 1726.
de Queiroz, K. 2007. Species concepts and species delimitation. Systematic
Biology 56: 879886.
Quiroga, P., A. Premoli, and C. Ezcurra. 2002. Morphological and iso-
zyme variation in Cerastium arvense (Caryophyllaceae) in the south-
ern Andes. Canadian Journal of Botany 80: 786795.
Rivero, M., M. Kalin Arroyo, and A. Humana. 1987. An unusual kind of
distily on Quinchmalium chilense in volcan Casablanca, Southern
Chile. American Journal of Botany 74: 313320.
Sokal, R. R. and P. H. A. Sneath. 1963. Principles of numerical taxonomy.
San Francisco: W. H. Freeman.
Stuessy, T. F. 1990. Plant taxonomy: The systematic evaluation of com-
parative data. New York: Columbia University Press.
Sultan, S. E. 2000. Phenotypic plasticity for plant development, function
and life history. Trends in Plant Science 5: 537542.
Ulibarri, E. A. 1994. Santalaceae Pp. 289302. In: Flora de San Juan,
República Argentina Vol. I, ed. R. Kiesling. Buenos Aires, Argentina:
Vázquez Mazzini Editores.
Valladares, F., S. J. Wright, E. Lasso, K. Kitajima, and R. W. Pearcy. 2000.
Plastic phenotypic response to light of 16 congeneric shrubs from
a Panamanian rainforest. Ecology 81: 19251936.
Wheeler, Q. D. and R. Meier. 2000. Species concepts and phylogenetic theory:
A debate. New York: Columbia University Press.
Wiens, J. J. and M. Servedio. 2000. Species delimitation in systematics:
inferring diagnostic differences between species. Proceedings Biological
Sciences 267: 631636.
Zuloaga, F., O. Morrone, and M. Belgrano. 2009. Catálogo de las plantas
vasculares del Cono Sur (Argentina, sur de Brasil, Chile, Paraguay y
Uruguay). Monographs in Systematic Botany of the Missouri Botanical Garden
107: 13486. Saint Louis, Missouri: Missouri Botanical Garden Press.
Appendix 1: Currently recognized Quinchamalium species after Brako and Zarucchi (1993), Zuloaga et al. (2009), and Jørgensen et al. (2014).
Nr. Species Author Year of
description Distribution
1Q. andinum Philippi 1857 Chile (Región V, VIII, X)
2Q. bracteosum Philippi 1857 Chile (Región X)
3Q. brevistaminatum Pilger 1930 Perú (Dep. Lima)
4Q. carnosum Philippi 1857 Chile (Región II, III)
5Q. chilense Lamarck 1782 Argentina (Chubut, Jujuy, Mendoza, Neuquén, Río Negro, Santa Cruz)
Chile (Región I, II, III, IV, V, VI, VII, VIII, IX, X, XI, RME)
Bolivia (Deps. Cochabamba, La Paz, Oruro, Potosí)
6Q. elongatum Pilger 1930 Perú (Dep. Ancash)
7Q. excrescens Philippi 1857 Chile (Región II, III, IV, V)
8Q. fructiculosum Steudel 18211824 Chile (Región VIII)
9Q. hopii Pilger 1930 Perú (Dep. Arequipa)
10 Q. linarioides Philippi 1857 Chile (Región VII)
11 Q. linifolium Meyen 1878 Chile (Región VII, VIII)
12 Q. litorale Philippi 1864 Chile (Región V)
13 Q. lomae Pilger 1930 Perú (Dep. Arequipa)
14 Q. parviflorum Philippi 1857 Chile (Región IV, VII, RME)
15 Q. pratense Philippi 1857 Chile (Región VIII)
16 Q. procumbens Ruiz y Pavón 1799 Perú (Deps. Ancash, Apurímac, Arequipa, Cajamarca, Cuzco, Lima,
La Libertad, Moquegua, Tacna).
17 Q. purpureum Philippi 1857 Chile (Región VII)
18 Q. tarapacanum Philippi 1891 Chile (Región I)
19 Q. thesioides var thesioides Philippi 1860 Chile (Región II)
20 Q. thesioides var flaccidum Philippi 1860 Chile (Región II)
21 Q. stuebelii Hieronymus 1896 Perú (Dep. Puno)
Bolivia (Dep. Oruro)
... Por lo tanto, el objetivo principal de este trabajo es resolver la taxonomía de este género sudamericano teniendo en cuenta el reciente estudio de análisis morfométrico multivariado en donde se resolvió la existencia de una sola especie altamente polimórfica en el mismo: Q. chilense Lam. (Lopez Laphitz et al., 2015). ...
... Para las medidas incluidas en la descripción de la especie, se usaron caracteres vegetativos y reproductivos de un subconjunto de 156 de las 750 plantas observadas. El rango de los valores mostrados en la descripción de la especie representa la variabilidad observada en esta muestra de toda la distribución del género (Lopez Laphitz et al., 2015). Para el mapa de distribución se ubicaron todos los especímenes medidos de los cuales la información de localización estaba disponible. ...
... A partir de la información bibliográfica, las observaciones de ejemplares de herbario, las observaciones a nivel poblacional en el campo, los estudios de los patrones de distribución geográfica y los análisis estadísticos multivariados de caracteres morfológicos (Lopez Laphitz et al., 2015), así como también de las observaciones de las imágenes de todos los ejemplares tipo, este estudio reconoce la existencia de una sola especie: Q. chilense Molina. Como resultado de la revisión nomenclatural, 28 nombres de especies y/o de taxones infraespecíficos son reducidos a la sinonimia de Q. chilense por primera vez, como se señala en la lista sinonímica a continuación con la abreviatura nov. ...
Full-text available
The genus Quinchamalium comprises hemiparasitic herbs endemic to South America. Its distribution expands through the Andes from northern Peru to southern Patagonia. The vague morphological limits and the existence of scarce diagnostic characters for the correct identification of the currently accepted species, implies the need for a revision of the taxonomy of the genus. The morphological and biogeographic information along with recently published morphometric analyses indicate the existence of a single polymorphic species: Q. chilense. As a result, a taxonomic treatment of the monospecific genus Quinchamalium is here presented. Valid name, synonyms, description of vegetative and reproductive morphology, and geographical distribution is provided for Q. chilense. In this study, 28 names of species or infraspecific taxa are reduced to synonymy of Q. chilense for the first time. In addition, 13 names are lectotypified and one is neotypified.
... Specifying criteria or methods to delimit species have been much discussed amongst taxonomists and systematists in the past. In plant systematics, the traditional and most used criterion to identify species has been morphology (Stuessy, 1990), although in many cases it has not been a conclusive criterion (e.g., Lopez Laphitz, Ezcurra, & Vidal-Russell, 2015a;Lopez Laphitz & Semple, 2015). More recently, molecular phylogenetics has decreased the uncertainty when defining species limits (e.g. ...
... Nowadays, DNA-based approaches such as barcoding and the identification of species with a threshold measure of divergence between taxa have increased their role in the recognition of diversity, and the genetic distances between homologous genes have been evaluated to identify species (CBOL Plant Working Group, 2009;Fazekas et al., 2009;Yao et al., 2010). In particular, in many taxa where the identification of species is difficult or impossible based on morphological characters alone (Bickford et al., 2007;Lopez Laphitz et al., 2015a;Lopez Laphitz & Semple, 2015), molecular assessments are a powerful tool used for systematics (Gagnon, Hughes, Lewis, & Bruneau, 2015;Ma, Zhao, Wang, Long, & He, 2015;Shepherd, Thiele, Sampson, Coates, & Byrne, 2015). ...
... Previous multivariate evaluation (principal component and discriminant analyses) of 17 morphological and 19 climatic variables emphasized the importance of some variables to characterize the variation within Q. chilense (Lopez Laphitz et al., 2015a). The mean, standard deviation, and range values for these variables were calculated for each lineage. ...
The integration of different characters (e.g. morphological, ecological, and molecular) is now recognized as important in species delimitation. In particular, genetic distances between homologous genes have been suggested as one of the main tools to identify species, especially in the case of cryptic species. Quinchamalium is morphologically variable and occupies a diverse set of biomes across its distribution in the Southern Andes. Recent work based on morphology has synonymized the entire genus as a single morphospecies, Quinchamalium chilense. This widely distributed taxon presents the opportunity to find potential cryptic species. The main objective of this study was to test the existence of cryptic species, based mainly on phylogenetic gene trees, genetic distances, and geographic patterns of haplotypes from molecular markers of the nuclear (ITS) and chloroplast (trnL-F) genomes, considering climatic and morphological characteristics. The ITS phylogeny and corresponding haplotype network resulted in three lineages with strong genetic differentiation and distinct geographic patterns. These lineages were informally named Desert, Matorral, and Mountain, based on their geographic distribution in different biomes. The trnL-F chloroplast phylogeny did not distinguish Desert from Matorral, and the haplotype network showed overlap between these last two lineages. Overall, we hypothesize the existence of two cryptic species within Quinchamalium chilense (Mountain and Matorral–Desert) that correspond to genetic, climatic, and morphological differences.
... It is known that geographic variability in plant morphology can be a function of genetic changes and phenotypic plasticity (Quiroga et al. 2002;Premoli et al. 2007) in response to local environmental conditions. For this reason, the study and interpretation of geographic variability patterns on morphological characters can contribute to the resolution of taxonomic problems (Lopez Laphitz et al. 2015). In this context, the objectives of this research were to study the patterns of morphological variability of teosinte, the climatic variability in the areas where it is distributed, and its phylogenetic relationships based on morphological characters, to clarify the relationships among all taxa of teosinte and to evaluate the current classifications of the genus Zea. ...
... Climatic variation studies have been frequently used to separate different plant groups (Lopez Laphitz et al. 2015), as in the case reported by Ruiz Corral et al. (2008), who classified 42 maize races in climatic groups. Environmental differences between teosinte groups seem to be critical factors determining local population distributions. ...
The wild species of the genus Zea comprise nine different taxa, including annual and perennial, diploid and tetraploid species, distributed from northern Mexico to Costa Rica. Due to the great variability existing among species, the aim of this work was to evaluate the morphological and climatic variability patterns and phylogenetic relationships among species, based on morphological characters. Two hundred and seventy-six populations of known species and races were included; 15 individuals per population were planted. From each individual, 19 morphological and physiological traits were measured, and at each collection site 23 climatic descriptors were obtained. Both data sets were analyzed with multivariate analysis. For the phylogenetic analysis 17 morphological variables were transformed to character states using GAP-WEIGHTING. With multivariate analyses, it was possible to determine the existence of well-defined groups. Species, and races of different subspecies, were found to be clearly delimited by geographic regions, climatic characteristics, altitudes, and distinct morphological features. The phylogenetic relationships found are congruent with those identified in similar studies based on molecular data. Morphological data on their own cannot give enough information to infer divergence times. In this context, multivariate and phylogenetic analyses with morphological data are taxonomically of great importance for identifying species and subspecies of genus Zea. At species level, the study has added elements to distinguish between Zea diploperennis from the states of Nayarit and Jalisco, between Zea perennis from Michoacán and Jalisco, and between Zea luxurians from Oaxaca and Guatemala.
... We used these measures to test the degree of vestigialization and developmental instability of floral traits between 22 monomorphic and 26 polymorphic populations widely distributed up to the northern range edge in Michigan, Ontario, and Quebec. Temperature during floral development can negatively affect flower production and morphology (Hodgins and Barrett 2008;Laphitz et al. 2015). Extensive field observation of D. verticillatus populations from across the range of the species suggests that bouts of cold temperature may inhibit flower production and reduce flower size and developmental stability. ...
... Anticipating that temperature during flower development would proximately influence the production, size, and developmental stability of flowers (Sawhney 1983;Hodgins and Barrett 2008;Laphitz et al. 2015), we statistically controlled for the effect of August mean temperature in all analyses. Instead of a general effect of temperature, we detected frequent interactions between temperature and both population type and morph structure. ...
Premise of research. Transitions from sexual to asexual reproduction are predicted to be accompanied by the reduction and developmental disintegration of sexual traits that no longer maintain fitness. While such transitions are common in plants, the evolutionary fate of sexual traits in derived asexual populations remains largely unknown, especially in species where the loss of sex is not confounded with a change in ploidy. In the tristylous wetland plant Decodon verticillatus, populations polymorphic for style length are sexual, while monomorphic populations are typically asexual under field conditions. Methodology. We compared ploidy, per-ramet flower production, flower size, and floral developmental stability between 22 monomorphic asexual populations and 26 polymorphic sexual populations distributed across the northern half of the species range in the Great Lakes region of North America. Pivotal results. Flow cytometry revealed that all populations were diploid—except for one monomorphic population that exhibited nuclear DNA content consistent with triploidy, which is known to cause sexual sterility. After accounting for the potential effects of temperature during flowering on floral development, we found that ramets in asexual populations produced ~50% (although not quite significantly) fewer flowers than those in sexual populations. Flowers varied widely in size among populations but were not smaller in asexual populations. Developmental stability of flowers was not lower in asexual populations than in sexual populations, although among monomorphic populations stability was lowest in populations for which confidence in exclusive asexuality was highest. Conclusions. Sex has been lost repeatedly across the northern range edge of D. verticillatus via at least two genetic pathways. Yet investment in and developmental stability of sexual traits is not consistently, and rarely significantly, reduced. The observation that no asexual genotype of D. verticillatus is widespread suggests that asexual populations do not persist long—perhaps only long enough to exhibit the very first stages of sexual vestigialization.
... Morphological variation has been related to geographic and climatic gradients in several widespread taxa of the southern Andes and Patagonia (e.g., Embothrium, Chalcoff, Ezcurra, & Aizen, 2008;idem, Souto, Premoli, & Reich, 2009;Nassauvia, Nicola et al., 2014;Quinchamalium, Lopez Laphitz, Ezcurra, & Vidal-Russell, 2015), so it seems important to explore this relationship in M. spinosum. Studies that determine the environmental characteristics (geographic and climatic) that are related to morphological variation within species can evidence the drivers of phenotypic differentiation that result in genetic adaptation or express in phenotypic plasticity (Chalcoff et al., 2008;Herrera, 2005;Lopez Laphitz et al., 2015;Paiaro, Oliva, Cocucci, & S ersic, 2012;Sultan, 2000). ...
... vary within these species, as well. Additional evidence of the two being the same entity is that 40% of the specimens originally considered as M. echinus were classified a posteriori by DA as M. spinosum, which is considerably higher than the 15% threshold often used in taxonomic studies (e.g., Lehnebach, 2011;Lopez Laphitz et al., 2015). Besides, the climatic PCA presented here shows an overlap in climatic preferences between both entities; both are sympatric and found in a wide array of semiarid climates in the Andes of central Chile and bordering areas of Argentina, and in Patagonia. ...
Full-text available
Delimiting species is an important, but frequently difficult aspect of systematics that should be addressed using data from multiple sources. Here we combine morphometric analyses and environmental characteristics to delimit species in the South Andean and Patagonian taxonomically difficult species-group composed by Mulinum spinosum, M. echinus and M. leptacanthum (Apiaceae-Azorelloideae). Molecular phylogenies have shown that these three species form part of a polytomy together with other Mulinum species, and therefore these data are not useful for their delimitation. We include measurements of 25 morphometric variables from 163 herbarium specimens and perform univariate and multivariate principal component analysis (PCA) and discriminant analysis (DA) to establish the limits amongst the three mostly sympatric, morphologically similar, and phylogenetically unresolved species. We also use 19 bioclimatic and three geographic variables from localities of the specimens to infer environmental characteristics of the taxa and test their relation with morphological variation. Morphological evidence supports the inclusion of M. echinus within the morphologically and climatically variable M. spinosum, and rejects its recognition as a distinct taxon at any rank. On the contrary, M. leptacanthum is considered a morphologically distinct species, generally restricted to high altitude areas of the southern Andes with a cooler and wetter climate. Within the widespread M. spinosum, environmental gradients of precipitation and temperature relate to morphological gradients (e.g., in leaf and inflorescence sizes, in leaf acicularity, and in fruit-wing width). These last results showed that the large morphological variation in vegetative and reproductive characteristics of this species that grows in arid and semiarid habitats are related to regional climatic gradients that have probably been important in the evolution of this species' plasticity, diversification, and differentiation.
... These studies showed a parallel divergence in the phylogeny and distribution of species of Veronica L. (Lin et al., 2016) and Picea L. worldwide; similar findings were reported for the genera of Gentianaceae in Brazil (Struwe et al., 2011). Links between plant functional traits and bioclimatic gradients have been established in the Mediterranean region (Díaz Barradas et al., 1999;Lopez Laphitz et al., 2016;Thuiller et al., 2004), whilst other studies demonstrated that environmental factors played an active role in the divergence of intra-and interspecific variation (Anacker and Strauss, 2014;Huang et al., 2015). ...
The species from the tribe Stipeae Dumort. dominate Eurasian grasslands including Anatolian steppes. In Turkey, the genera of the Stipeae show distinct patterns of regional distribution, leading to a combination of dispersal and adaptive radiation. However, the importance of climatic gradients to regional distribution of these genera of the Stipeae has not yet been examined, and the aim of our study was to address this need. We retrieved data on the presence of three genera of the Stipeae, and morphological traits of its member species in Anatolia from available literature. We also acquired data on bioclimatic variables from the WorldClim database. We used multivariate methods (PCA) and correlation analysis to describe the major climatic gradients within the distribution area of the Stipeae in Anatolia. Links between the major climatic gradients and the divergence of the genera were examined using the method of Spatial Evolutionary and Ecological Vicariance Analysis (SEEVA). Possible relationships between the genera of the Stipeae and climatic and geographic variables were also examined by means of Canonical Correspondence Analysis (CCA). Our results suggest that elevation and temperature are the key factors for the ecological divergence of the genera in the tribe Stipeae in Turkey.
... Populations within a widely distributed species can adapt differently in different environments, leading to varieties or ecotypes that are genetically distinct. A recent study in the closely related genus Quinchamalium showed that all accepted species were actually one polymorphic species, thereby requiring that 28 names be synonymized to Q. chilense Molina (Lopez Laphitz et al. 2015aLaphitz et al. , 2015b). These authors found no morphological gaps between different Quinchamalium individuals sampled across the entire distribution. ...
... Since the emergence of numerical taxonomy (Sneath & Sokal, 1973), multivariate and statistical analyses have become a powerful tool in delimiting species and assessing morphological variation and an increasing number of taxonomic works use complementary studies such as morphological, phylogenetic, cytogenetics, phylogeography, and population genetics as standard methodologies in taxonomy (Nordström & Hedrén, 2009;Akhavan et al., 2015;Gale et al., 2015;Ali et al., 2016;Arbizu et al., 2016;Banasiak et al., 2016). Nevertheless, in the taxonomic literature there are few examples that deserve special attention with regard to morphological variation promoted by environmental factors and its taxonomic implications; these investigations demonstrated that phenotypic variation could hinder species delimitation (Ellison et al., 2004;Ložiene, 2006;Cavallero et al., 2011;Nicola, Johnson & Pozner, 2014;Scrivanti et al., 2014;Lopez Laphitz, Ezcurra & Vidal-Russell, 2015). In a group of species with restricted distributions, in which specialization to narrow and distinct climatic/ environmental envelopes has been demonstrated to be the main force leading to speciation, the specialization of a newly discovered population to a climatic/environmental condition distinct from all known species in the group might be a suitable argument to advocate its species status (Padial et al., 2010). ...
Full-text available
Senna series Aphyllae includes xeromorphic shrubs and subshrubs that occur in three different biogeographic subregions in arid and semi-arid habitats of southern South America. The series provides a good opportunity to understand better the relationship among geographical, climatic, and morphological variation in different taxa. Moreover, in this group the specific and varietal delimitation is still problematic due to the high morphological variation within and among taxa. Statistical analyses of climatic and morphological data and geographical distribution were used to understand the patterns of morphological variation among 394 individuals and to clarify the taxonomic delimitation of entities that belong to series Aphyllae. Senna acanthoclada and S. nudicaulis were segregated from each other and from the remaining taxa; the three recognized varieties of the S. aphylla complex were delimited; S. spiniflora was well-delimited and S. crassiramea and S. rigidicaulis were overlapping; S. pachyrrhiza was not differentiated from S. aphylla. Ecological niche modelling showed several areas of contact and a large overlap of suitable conditions for several species. The results of this work revealed that most morphological variability is associated with different environmental conditions. This phenotypic plasticity may be caused by the presence of different environments with climatic factors in the South American Transition Zone, Chaco, and northern region of Central Patagonia province.
Full-text available
Pleistocene glacial periods have had a major influence on the geographical patterns of genetic structure of species in tropical montane regions. However, their effect on morphological differentiation among populations of cloud forest plants remains virtually unexplored. Here, we address this question by testing whether geographical patterns of morphological variation in Ocotea psychotrioides can be explained by the intensity of climate change occurring during 130,000 years. For this, we measured vegetative and reproductive traits for 96 individuals from 36 localities registered across the species’ distribution range. Species distribution models and multivariate statistics were used to investigate geographical patterns of morphological variation and test their association with current and past climatic conditions. Leaf size and pubescence in O. psychotrioides showed a latitudinal pattern of clinal variation that does not fit the geographical gradient of increasing leaf size towards lower latitudes observed globally among plants. Instead, the observed clinal variation conforms to a pattern of increasing leaf size towards higher latitudes. However, our analyses showed weak to non-significant association between morphology and current climate. Interestingly, our analyses showed that predicted shifts in the distribution range of O. psychotrioides during the last 130,000 years were accompanied by significant changes in climatic conditions, particularly temperature seasonality and precipitation. Accordingly, climatic instability showed a better fit to the observed patterns of leaf size and pubescence variation than current climate conditions. These results suggest that climatic instability during the Pleistocene glacial periods might play a key role in promoting morphological differentiation among populations of cloud forest plants.
Full-text available
A new South American species of Solidago L. (Asteraceae, Astereae) from Argentina, Chile, and Bolivia was identified within the subsection Junceae (Rydb.) G. L. Nesom. Using multivariate analyses on a matrix of 50 characteristics for 79 specimens, the distinctiveness of a South American S. missouriensis-like taxon was tested and determined to be statistically different from four morphologically similar North American species: S. gattingeri Chapm. ex A. Gray, S. juncea Aiton, S. missouriensis Nutt., and S. pinetorum Small. Therefore, a new South American species is proposed: S. argentinensis Lopez Laph. & Semple.
Full-text available
Phylogenetic relationships among the 11 species and subspecies of the Andean-Patagonian Chuquiraga sect. Acanthophyllae were resolved by parsimony cladistic analysis using 22 morphological characters. The comparative method was used to test whether a reduction in leaf width occurred in species due to an adaptation to warmer desert climates. Mean values of annual precipitation, January (summer) temperature and July (winter) temperature were estimated for each taxon. Independent comparisons for leaf width and climatic variables were calculated at each node of the cladogram and a regression analysis of leaf variation versus climatic variation was performed. The probable ancestral geographic area for the group was determined using Bremer's method. Results of these analyses suggest that marked involution and reduction in leaf width occurred twice independently in the evolution of the group. Reduction of leaf width was correlated with an increase in temperature. The Puna, Patagonia and the High Andes have the highest probability of having been part of the ancestral area of this section, which currently also extends to the Monte Prepuna and Chilean Desert. This study suggests a relatively recent climatic effect on the evolution of leaf morphology.
Full-text available
Abstract— In the tribe Andropogoneae, morphological variation is remarkable, mainly in the inflorescence on the pair of spikelets which are the core elements of the inflorescence. The genus Andropogon includes the Andropogon lateralis complex, which is distributed primarily in South and Central America, comprising approximately twelve taxa and inter-specific hybrids. The aim of this study was to assess morphological variation in the A. lateralis complex through morphometric analyses of specimens from natural populations. For this purpose, univariate ANOVA, as well as principal component analysis and discriminant analysis of 19 morphological variables of synflorescences were performed, revealing differences between species and interspecific hybrids. The selected diagnostic traits of species and hybrids based on quantitative characters of the synflorescences provided a valuable tool for taxonomic studies in the genus. The results obtained made it possible to generate the first identification key that includes both species and hybrids of the A. lateralis complex for South America.
A century of research on heterostylous plants has passed since the publication of Charles Darwin's book "The Different Forms of Flowers on Plants of the Same Species" in 1877 summarizing his extensive observations and experiments on these complex breeding systems involving genetic polymorphisms of floral sex organs. Since then heterostylous plants have provided a rich source of material for evolutionary biologists and today they represent one of the classic research paradigms for approaches to the study of evolution and adaptation. The present book is the first modern and comprehensive accont of the subject. In 10 chapters it is concerned with the evolution, genetics, development, morphology, and adaptive significance of heterostyly. Broad syntheses of research on heterostyly as well as new theoretical ideas and experimental data are included.
Different pollinators can exert different selective pressures on floral traits, depending on how they fit with flowers, which should be reflected in the patterns of variation and covariation of traits. Surprisingly, empirical evidence in support of this view is scarce. Here, we have studied whether the variation observed in floral phenotypic integration and covariation of traits in Narcissus species is associated with different groups of pollinators. Phenotypic integration was studied in two style dimorphic species, both with dimorphic populations mostly visited by long-tongued pollinators (close fit with flowers), and monomorphic populations visited by short-tongued insects (loose fit). For N. papyraceus, the patterns of variation and correlation among traits involved in different functions (attraction and fit with pollinators, transfer of pollen) were compared within and between population types. The genetic diversity of populations was also studied to control for possible effects on phenotypic variation. In both species, populations with long-tongued pollinators displayed greater phenotypic integration than those with short-tongued pollinators. Also, the correlations among traits involved in the same function were stronger than across functions. Furthermore, traits involved in the transfer of pollen were consistently more correlated and less variable than traits involved in the attraction of insects, and these differences were larger in dimorphic than monomorphic populations. In addition, population genetic parameters did not correlate with phenotypic integration or variation. Altogether, our results support current views of the role of pollinators in the evolution of floral integration.