Biological Journal of the Linnean Society, 2010, 101, 961–976.

With 6 figures

Parsimony analysis of endemicity as a

panbiogeographical tool: an analysis of Caribbean
plant taxa

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

AMPARO ECHEVERRY and JUAN J. MORRONE*
Museo de Zoología ‘Alfonso L. Herrera’, Departamento de Biología Evolutiva, Facultad de Ciencias,
Universidad Nacional Autónoma de México (UNAM), Apartado Postal 70-399, 04510 Mexico DF,
Mexico

Received 21 March 2010; revised 8 June 2010; accepted for publication 8 June 2010 bij_1535 961..976

To demonstrate that parsimony analysis of endemicity (PAE) can be a method implementing the panbiogeographic
approach, we analyzed two data matrices of 40/38 biogeographic provinces ¥ 148 plant species from the Caribbean
subregion of the Neotropical region, one where taxa are represented by individual tracks and the other where taxa
are represented by single sample localities. We obtained six generalized tracks resulted from the PAE of the
areas ¥ individual tracks matrix, and one generalized track from the PAE of the areas ¥ single sample localities
matrix, with the latter nested within the former tracks. The results obtained show that PAE works as a
panbiogeographical tool if it is based on an areas ¥ individual tracks matrix. When performed in this way, PAE
retrieves spatial information that is lost when it is based on an areas ¥ single sample localities matrix, raising
doubts regarding the conclusions derived from this latter type of analysis. © 2010 The Linnean Society of London,
Biological Journal of the Linnean Society, 2010, 101, 961–976.

ADDITIONAL KEYWORDS: evolutionary biogeography – primary biogeographic homology – systematic

biogeography – vicariance.

INTRODUCTION units: Cracraft, 1991; Da Silva & Oren, 1996;

Morrone, 1998; to identify areas of endemism:
Parsimony analysis of endemicity (PAE) was proposed
Morrone, 1994; Watanabe, 1998; Ippi & Flores, 2001;
by Rosen & Smith (1988) in a paleontological context,
to classify biogeographic provinces: Morrone et al.,
involving the distributions of taxa from sampling
1999; Espinosa-Organista et al., 2000; to detect gen-
localities in different geological horizons. PAE was
eralized tracks: Luna-Vega et al., 2000; Morrone &
originally applied to determine relations among sam-
Márquez, 2001).
pling localities, through the analysis of a presence/
As a panbiogeographical tool the history of PAE is
absence distributional data taxa matrix, in which
somewhat tangled. Craw (1989b) described PAE as an
taxa are taken as the characters (i.e. in a direct
area cladogram method and compatibility track
analogy with phylogenetic systematics) to be analyzed
analysis as a track method, considering that both
using a parsimony algorithm. This method has been
were methods of quantitative panbiogeography.
since then applied to different geographic units (e.g.
Morrone & Crisci (1995) and Crisci, Katinas &
localities: Rosen, 1988; previously delimited areas of
Posadas (2003) treated PAE as an independent
endemism: Craw, 1988a, Cracraft, 1991; quadrats:
approach, not being part of panbiogeography, consid-
Morrone, 1994) and for reaching different goals (e.g.
ering only connectivity and incidence matrices (Page,
to determine relations among different geographic
1987) and track compatibility analysis (Craw, 1988a,
b, 1989a, b) as quantitative panbiogeographic proce-
dures. Smith (1992: 265) stated that ‘In effect PAE is
*Corresponding author. a method of nesting panbiogeographical tracks into a
E-mail: juanmorrone2001@yahoo.com.mx hierarchical scheme’. Craw, Grehan & Heads (1999)

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976 961
962 A. ECHEVERRY and J. J. MORRONE

included PAE among other quantitative panbiogeo- next to its northern and southern limits, were con-
graphic methods. Harold & Mooi (1994) and Morrone sidered (Fig. 1). These biogeographic provinces are
(2001) argued that areas of endemism are equivalent assumed to represent natural units or areas of
to generalized tracks, where ‘areas of endemism are endemism based on previous panbiogeographic and
fundamentally historical entities, not distributional cladistic analyses, integrated into a single model
ones, and their definition should take history into of hierarchically arranged biogeographic areas
account’ (Harold & Mooi, 1994: 262). Grehan (2003) (Morrone, 2004a, 2006). The use of pre-defined areas
considered PAE as one of the analytical methods in of endemism, however, must be taken as an opera-
panbiogeographical analyses. Morrone (2004a, 2005, tional shortcut because the results obtained in the
2009) considered that panbiogeography, including cladogram must be always confirmed by going back to

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

PAE as a quantitative method, allows the identifica- the map and drawing the individual tracks that
tion of primary biogeographical homology, which rep- support a given area relationship, aiming to ensure
resents a conjecture on a common biogeographical that such a relationship is not an artefact of the
history and can be used to identify biotic components analysis.
under a panbiogeographical approach. Nihei (2006)
aimed to demonstrate that the original formulation
of PAE provides an insightful method for inferring TAXA
historical patterns, meaning, by original PAE, the We analyzed 148 vascular plant species, belonging to
dynamic approach that led to an analysis of the biota the families Araceae, Arecaceae, Aristolochiaceae,
throughout space and time, as opposed to the static Aspleniaceae, Bromeliaceae, Dryopteridaceae, Lau-
approach that excludes the temporal component raceae, Leguminosae, Meliaceae, Orchidaceae, Piper-
basing the analysis on a single geological horizon. aceae, Poaceae, Rubiaceae, and Urticaceae. These
Garzón-Orduña, Miranda-Esquivel & Donato (2008: taxa are distributed in the area of study and repre-
910) concluded: ‘we stress that PAE can be used in the sent a good sample of the biological diversity of the
same way as a panbiogeographic approach, to define area. It includes palms, ferns, epiphytes, shrubs,
a primary biogeographical homology (Morrone, 2001), trees, and grasses. Distributional data were taken
but its use as a test or assessment of biogeographical from published monographs and revisionary studies
history is misleading’. (Pfeifer, 1966; Croat, 1975, 1983, 1997; Chase, 1986;
The objectives of the present study are three-fold: Zuloaga, 1987; Zuloaga & Sendulsky, 1988; Taylor,
(1) to set the position of PAE within the taxonomy of 1989, 1992, 1994, 1997; Moran, 1991; Murakami &
evolutionary biogeographic approaches; (2) to provide Moran, 1993; Sousa, 1993; van der Werff, 1993, 2002;
a step-by-step description of its algorithm to make it Henderson, Galeano & Bernal, 1995; Salzman &
conceptually and methodologically consistent with the Judd, 1995; Franco & Berg, 1997).
principles of panbiogeography, as originally formal-
ized by Croizat (1958, 1964); and (3) to compare the
results from PAE when it is performed based on a ALGORITHM
matrix of areas ¥ single sample localities versus a Léon Croizat’s panbiogeography (1958, 1964) was a
matrix of areas ¥ individual tracks, aiming to deter- pioneering attempt to explore the potential of a
mine how much is gained in terms of spatial infor- graphical approach to biogeography (Craw et al.,
mation. We intend to show that PAE works as a 1999). Croizat’s method was to plot distributions of
panbiogeographical tool allowing the generation of organisms on maps and connect the disjunct distri-
testable historical hypotheses of area relationships. bution areas or collection localities with lines he
termed tracks. Individual tracks for unrelated groups
of organisms were then superimposed and, if they
coincided (if they show substantial topological agree-
MATERIAL AND METHODS
ment), the resulting summary line was termed a
AREA generalized or standard track. A generalized track
The area analyzed essentially corresponds to that was interpreted by Croizat as indicating the preexis-
which Morrone (2004b, 2006) defined as the Carib- tence of an ancestral biota that had subsequently
bean subregion, based on distributional biogeographic become fragmented as a result of tectonic and/or
analyses of plant and animal taxa. The Caribbean climatic change (Morrone, 2009).
subregion extends through southern Mexico, Central When PAE has been used to obtain generalized
America, Greater and Lesser Antilles, and north- tracks (Luna-Vega et al., 2001; Corona, Toledo &
western South America. As operational geographical Morrone, 2007), the matrices analyzed were often
units for the analysis, biogeographic provinces based on areas ¥ taxa, with ‘taxa’ representing single
belonging to this subregion, as well as those situated sample localities. We postulate here that the

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
 PAE AND PANBIOGEOGRAPHY 963

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

Figure 1. Provinces analyzed. Cal, California; BCa, Baja California; Son, Sonora; MPl, Mexican Plateau; Tam, Tamauli-
pas; Smw, Sierra madre occidental; Sme, Sierra madre oriental; Tvb, Transmexican volcanic belt; Bal, Balsas basin; Sms,
Sierra madre del sur; *MPc, Mexican Pacific coast; *MGu, Mexican Gulf; *YPe, Yucatan Peninsula; Flo, Florida Peninsula;
*Bah, Bahamas; *Cub, Cuba; *Cay, Cayman Islands; *Jam, Jamaica; *His, Hispaniola; *PRi, Puerto Rico; *LAn, Lesser
Antilles; *T&T, Trinidad & Tobago; *Chi, Chiapas; *ECa, Eastern Central America; *WPI, Western Panamanian Isthmus;
*Cho, Choco; *Mar, Maracaibo; *Vco, Venezuelan coast; *Mag, Magdalena; *Vel, Venezuelan Llanos; *Cau, Cauca; *WEc,
Western Ecuador; *AEc, Arid Ecuador; *TuP, Tumbes Piura; *Gal, Galapagos Islands; NoA, North Andean paramo; Nap,
Napo; Ime, Imeri; Guy, Guyana; Ror, Roraima; Var, Varzea; Mad, Madeira; Uca, Ucayali. Provinces marked with an
asterisk (*) correspond to the Caribbean subregion.

interpretation of the resulting cladograms as gener- spanning tree, either manually or using any of the
alized tracks may not be conceptually and method- software available (Rojas-Parra, 2007; Liria, 2008;
ologically consistent because they are not based on Cavalcanti, 2009).
individual tracks. To use PAE as a panbiogeographic 2. Construct an r ¥ c matrix, where the rows (r) rep-
tool, it should take into account the basic panbiogeo- resent geographic units and the columns (c) rep-
graphic concept of the individual track. This would resent individual tracks (Morrone, 2004a, b). Each
allow incorporation of the spatial component a priori, matrix entry is ‘1’ when an individual track is
namely in the construction of the matrix before per- present in or crosses a given geographical unit and
forming the parsimony analysis. ‘0’ if it is absent or does not cross it. A ‘?’ code may
Because the panbiogeographic approach allows be included in case of doubtful occurrence in some
composite areas or nodes to be identified where single geographical unit. Include a hypothetical area
areas are involved in different generalized tracks, coded 0 for all columns in the data matrix to root
PAE with progressive character elimination (Luna- the resulting cladogram(s).
Vega et al., 2000; García-Barros et al., 2002) should be 3. Perform a parsimony analysis on this data matrix
implemented. This modification of PAE means that, using any of the available programs (e.g. NONA:
once the most parsimonious cladograms have been Goloboff, 1998; PAUP: Swofford, 2001). If more than
obtained, the individual tracks supporting the differ- one most parsimonious cladogram is found, con-
ent clades are removed and the reduced matrix is struct a consensus tree.
analyzed to search for alternative clades supported by 4. Take as a generalized track any clade supported
other individual tracks. by at least two synapomorphic individual tracks
The algorithm implementing PAE as a panbiogeo- shared by given geographic units.
graphic method comprises six steps (Fig. 2): 5. Draw the supporting individual tracks to connect
the geographic units identified as part of the same
1. Construct individual tracks for different taxa, con- generalized track, and to determine the overlap-
necting the occurrence localities by a minimal ping sectors in it. If some areas result in the

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
964 A. ECHEVERRY and J. J. MORRONE

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

Figure 2. Algorithm comprising the required steps to make parsimony analysis of endemicity (PAE) an appropriate
panbiogeographic tool. A, B, C, D, Individual tracks for different taxa; E, distribution areas ¥ individual tracks matrix;
F, cladogram produced by PAE showing one generalized track; G, generalized tracks identified and node.

overlap of two or more generalized tracks, identify (Figs 4, 5, 6) were drawn using Photoshop 7.0 (Adobe
them as nodes. Systems Inc.).
6. Remove from the matrix the synapomorphic indi-
vidual tracks supporting the clades obtained pre-
viously and repeat steps 3–5 until no more
RESULTS
individual tracks support any clade.
The PAE of the geographic units ¥ individual tracks
To undertake PAE, the algorithm described above produced 100 cladograms, with 617 steps, a consis-
was employed. Individual tracks for each species were tency index of 0.23 and a retention index of 0.65. The
constructed using TRAZOS (Rojas-Parra, 2007) as an strict consensus cladogram had 674 steps, a consis-
extension of ARCVIEW GIS 3.2. Some individual tracks tency index of 0.21 and a retention index of 0.61.
are shown in Figure 3. The data matrix of 40 geo- Six generalized tracks were identified (Figs 5F, 6K;
graphic units ¥ 148 individual tracks is presented in Table 1). The generalized track identified from the
the Appendix (Table A1). An alternative data matrix geographic units ¥ single sample localities matrix, as
of the same 148 species but composed by 38 geo- well as the species that support it, are included as a
graphic units ¥ single sample localities, which was nested track in the hierarchical scheme of the PAE
constructed to test the behaviour of individual tracks based on the geographic units ¥ individual tracks
in PAE, is presented in the Appendix (Table A2). The matrix.
cladistic analyses were carried out using NONA When the synapomorphic individual tracks (sup-
(Goloboff, 1998) through WINCLADA 1.00.08 (Nixon, porting the six generalized tracks obtained) were
1999), with the heuristic search option (tree bisection removed from the geographic units ¥ individual
and reconnection, 100 replications). Resulting tracks tracks matrix, and the reduced matrix was analyzed

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
 PAE AND PANBIOGEOGRAPHY 965

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

Figure 3. Four individual tracks out of 148 plant species included in this study. A, Ocotea aurantiodora. B, Philodendron
ligulatum. C, Anthurium andicola. D, Anthurium chiapasense.

Figure 4. Species distribution (A, B, C) supporting the generalized track (D) identified by PAE from the areas ¥ single
sample localities matrix. A, Philodendron ligulatum. B, Inga goldmanii. C, Inga mucuna. D, generalized track.

to search for alternative clades supported by other The PAE for the single sample localities matrix
individual tracks, 200 most parsimonious cladograms, produced 50 most parsimonious cladograms, with
with 591 steps, a consistency index of 0.22 and a 635 steps, a consistency index of 0.23 and a reten-
retention index of 0.64 were produced. The strict tion index of 0.53. The strict consensus cladogram
consensus cladogram had 692 steps, a consistency had 775 steps, a consistency index of 0.19 and a
index of 0.19 and a retention index of 0.56. For this retention index of 0.40. One clade supported by
dataset, however, no additional generalized tracks three species defined one generalized track (Fig. 4D,
were found. Table 1).

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
966 A. ECHEVERRY and J. J. MORRONE

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

Figure 5. Species distribution (A, B, C, D, E) supporting the generalized tracks (F) identified by PAE from the
areas ¥ individual tracks matrix. A, Aristolochia tentaculata. B, Anthurium andicola. C, Anthurium chiapasense. D,
Aristolochia bullata. E, Inga flexuosa. F, two generalized tracks (with the thickest line representing the track nested
within the other).

When the synapomorphic individual tracks (sup- they have resulted from the same major historical
porting the different clades) were removed from the events, such as tectonics or climate change (Rosen &
single sample localities matrix, and the reduced Smith, 1988). The possibility that the same pattern
matrix was analyzed to search for alternative gener- has arisen by chance becomes negligibly small as
alized tracks supported by other individual tracks, congruence increases (Nelson & Platnick, 1981:
150 most parsimonious cladograms, with 630 steps, a 503). Long before PAE or cladistic biogeography
consistency index of 0.23 and a retention index of 0.53 methods were developed, panbiogeography (Croizat,
were produced. The strict consensus cladogram had 1958, 1964) focused on the role of locality and place
774 steps, a consistency index of 0.18 and a retention in the history of life. Its goal is to recover the
index of 0.39. No additional generalized tracks were importance of the geographical spatial information
found. Nonetheless, it is worth noting that two as a direct subject of analysis in biogeography.
species (Anthurium andicola and Ocotea auran- In panbiogeography, comparisons are spatial and
tiodora) supporting two different clades (Tvb, Sms, homologous distributions can be identified through
MPc, MGu, Chi) and (Mag, Cau, Cho, ECa, WPI) were the track method (Craw et al., 1999).
retrieved by the first run of PAE from the geographic In the panbiogeographical implementation of PAE
units ¥ individual tracks matrix, along with addi- discussed in the present study, the spatial or geo-
tional supporting species, as can be seen in graphical information is highlighted by the use of
Figures 5A, B, C, D, E and 6A, B, C, D, E, F, G, H, I, individual tracks (instead of single sample localities),
J and Table 1. as the source from which generalized tracks is recog-
nized. In other words, when performed in this way,
PAE captures the spatial information in the same
DISCUSSION
manner that panbiogeography does, although it pre-
PAE is a pattern-oriented method that uses a cla- sents it in the hierarchy implicit in the cladograms.
distic algorithm to analyze geographical patterns of Generalized tracks represent congruent distributional
distributions. As in cladistic biogeography, the rec- patterns and therefore provide a hypothesis of
ognition of congruent patterns of distribution among primary biogeographic homology, which refers to a
different taxonomic groups is in itself evidence that conjecture on a common biogeographic history by

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
 PAE AND PANBIOGEOGRAPHY 967

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

Figure 6. Species distribution (A, B, C, D, E, F, G, H, I, J) supporting the generalized tracks (K, L) identified by PAE
from the areas ¥ individual tracks matrix. A, Inga alba. B, Inga sertulifera. C, Anthurium consobrinum. D, Inga sierrae.
E, Ocotea aurantiodora. F, Philodendron ligulatum. G, Inga goldmanii. H, Inga mucuna. I, Anthurium clidemioides. J,
Inga latipes. K, four generalized tracks (with the relative thickness representing the three tracks nested within the other).

virtue of their ancestors occupying the same terns. As Rosen & Smith (1988: 289) stated
paleogeographic sector and being subjected to the ‘. . . although PAE seems, in the end, to settle for
same geological, geomorphological, and climatic quantity rather than quality of data, the resultant
changes (Craw et al., 1999; Morrone, 2001, 2009). cladograms do provide the best corroborated hierar-
Homology, being essentially a comparative concept, chy of sample localities using available data on
where individual homology statements (distributions shared endemic taxa, and they therefore provide
of individual taxa) interact with each other, the quan- working hypotheses to be tested further’. This coin-
tity of the data analyzed plays an important role, and cides with the evolutionary biogeographic approach
it is desirable to have as many distributional data as (Morrone, 2009), where historical hypotheses of
possible to obtain better supported historical pat- area relationships generated by panbiogeographical

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
968 A. ECHEVERRY and J. J. MORRONE

Table 1. Comparison of parsimony analysis of endemicity (PAE) performance from the geographic units x single sample
localities matrix versus from the geographic units x individual tracks matrix

PAE geographic units ¥

single localities matrix PAE geographic units ¥ individual tracks matrix

Number of provinces involved 4 22

in the resulting cladograms
Total number of generalized 1 6
tracks identified
Total number of species 3 15

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

supporting generalized tracks
Number of nested generalized 0 4
tracks
Species supporting each Philodendron ligulatum, Aristolochia tentaculata, Anthurium andicola, and
generalized track and Inga goldmanii, Anthurium chiapasense (Bal, Tvb, Sms, MPc, MGu,
the provinces grouped by it and I. mucuna Chi)
(Mag, Cho, ECa, WPI) Aristolochia bullata and Inga flexuosa (Sms, MPc, MGu,
Chi)*
Inga alba and Inga sertulifera (Uca, Ror, Ime, Var, Mad,
Guy, Vel, Mar, Vco, Nap, NoA, Cau, Mag, Cho, ECa,
WPI)
Anthurium consobrinum and Inga sierrae
(NoA, Cau, Mag, Cho, ECa, WPI)*
Ocotea aurantiodora, Philodendron ligulatum, Inga
goldmanii, and Inga mucuna (Cau, Mag, Cho, ECa,
WPI)*
Anthurium clidemioides and Inga latipes
(Mag, Cho, ECa, WPI)*

*Represents the four generalized nested tracks. The first one is nested within the track shown in Fig. 5F and the last
three are nested within the track shown in Fig. 6K.

methods might be tested using cladistic biogeographi- For example, Myers (1991: 24) suggested the use of
cal methods in a further step within the same analy- different taxonomic levels to ‘allow an interpretation
sis. This also clarifies the aim of PAE, namely to help of the history of space occupancy by lineages through
identify patterns (generalized tracks, congruent dis- geological time’. Cracraft (1991) independently
tributional data, primary biogeographic homology). applied this principle to impose some hierarchical
Causal explanations lie beyond its scope and there- structure to the raw data. Morrone & Márquez (2001)
fore is not a weakness per se (contra Garzón-Orduña argued that the inclusion of supraspecific monophyl-
et al., 2008). etic groups, along with their specific taxa, would
Criticisms of PAE as a historical method are based introduce cladistic information into the analysis.
on two facts. First, it does not use phylogenetic infor- Porzecanski & Cracraft (2005) formally proposed a
mation to infer area relationships (Humphries, 1989; derived method, named cladistic analysis of distribu-
Humphries & Parenti, 1999) and, second, it uses only tions and endemism.
distributional data to infer either area relationships Criticizing PAE for its use of distributional data as
(Nihei, 2006; Garzón-Orduña et al., 2008) or to iden- the only information to infer area relationships is
tify areas of endemism (Santos, 2005). Both criticisms fundamentally based on misunderstandings of the
have in common the idea that only phylogenetic infor- primary literature. For example, Rosen & Smith
mation allows the incorporation of a temporal dimen- (1988) noted that a direct reading of a PAE cladogram
sion to the analysis of area relationships (Nelson & (from any given time plane) provides a hypothesis
Platnick, 1981; Wiley, 1987, 1988; Nelson & Ladiges, about the history of biotas and areas irrespective of the
1991, 1996; Page, 1994). Although it is hardly debat- nature of the historical events evolved. As they stated
able that well supported phylogenies are a useful tool ‘. . . since the nature of barriers and events cannot be
to infer area relationships, when lacking phylogenetic specified from PAE alone, or in fact from any other
information, some studies have proposed indirect analysis of presence/absence distributional data (such
methods to incorporate a temporal dimension to PAE. as cladistic biogeography or cluster analysis), and

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
 PAE AND PANBIOGEOGRAPHY 969

since histories inferred from these methods cannot be method retrieves spatial information more efficiently
used as rigorous tests of geological events, or vice than PAE performed from an areas ¥ single sample
versa, the main role of biogeographical analysis lies in localities matrix, and presents it in a hierarchical
making test-like comparisons with independently scheme, making possible the generation of historical
derived histories’ (Rosen & Smith, 1988: 286). This hypotheses of area relationships into an area-
contrasts with Garzón-Orduña et al. (2008: 904) who fragmentation sequence frame. It is desirable to make
pointed out that ‘. . . its use has now been extended to further analyses of distributional data to explore
reconstruct “area cladograms” and infer vicariance the potential of using both PAE performances
events’. Rosen & Smith (1988: 286) stated clearly that (areas ¥ single sample localities matrix versus
‘. . . taken on their own, the PAE cladograms only areas ¥ individual tracks matrix) as representing dif-

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

provide historical hypothesis about biotas as repre- ferent spatio/temporal scenarios and present and past
sented at particular sample localities. Nevertheless, distributions, respectively. The distributional change
divergences in cladograms can be read as hypothetical through time, implicit in the concept that defines a
vicariance events, and these imply that barriers to generalized track (as indicating the pre-existence of an
biotic exchange might have appeared somewhere in ancestral biota that had subsequently become frag-
regions which lie between pairs of sister sample locali- mented by tectonic and/or climatic change), could be
ties on each side of a divergence’. Another misinter- viewed therefore as a two time-planes approach, in
pretation made by Garzón-Orduña et al. (2008) when conjunction with the historical interpretation of biotas
citing Rosen (1988) refers to the fact that PAE only of a single horizon, as outlined by Rosen & Smith
uses distributional data, whereas cladistic biogeogra- (1988).
phy uses ‘taxonomic characters to fingerprint areas’. Further comparisons of PAE (as a panbiogeo-
Rosen (1988: 457) highlighted this as one of the graphical method) should be made with methods
differences between PAE and cladistic biogeography, that explicitly evaluate the congruence between dis-
without any additional statement about the superior- tributions (such as the optimality criterion method
ity of any of them to produce area cladograms, as of Szumik & Goloboff, 2004) and with PAE with
implied by Garzón-Orduña et al. (2008). Additionally models of potential distributional areas (Escalante,
citing Rosen (1988: 457), these authors state that ‘as Szumik & Morrone, 2009). Because these previous
PAE only uses contemporary distributions, the studies reported that PAE with models of distribu-
meaning and scope of the results are not clear when tional areas performed better than PAE with point
there are taxa from different strata at the same records, as a result of point data underestimating
locality’. Rosen (1988: 457) pointed out that ‘. . . cladis- the real distributional areas, it would be interesting
tic biogeography can use related taxa from any geo- to compare the performance of PAE from an
logical horizon, but PAE currently uses sets of areas ¥ individual tracks matrix versus PAE with
contemporaneous taxa. Until the implications of using models of potential distributional areas. The
taxa as analogues of characters in systematics are segment connecting two single sample localities
worked out, the meaning and scope of using taxa from (points data) in an individual track represents a
different horizons within a single sample locality in spatiotemporal projection of a given distribution,
PAE will not be known’ (emphasis added). Contempo- and therefore could be viewed as representing a
raneous taxa do not mean extant taxa; instead, it potential distribution.
means taxa occurring in the same time plane, what-
ever the time plane is; so, as noted previously, each
time plane cladogram produce by PAE from any given
ACKNOWLEDGEMENTS
time plane provide a hypothesis about the history of
biotas and areas, irrespective of the nature of the We thank Michael Heads, Tania Escalante, John
historical events evolved. When it is possible to Grehan, and two anonymous reviewers for providing
produce several PAEs from distributional data of dif- useful comments to improve the manuscript. A.
ferent time planes, then ‘. . . this multiple time plane Echeverry thanks CONACyT for economic support
approach introduces a synoptic approach in conjunc- through doctoral grant 93841 and the Posgrado en
tion with the outright historical interpretation of Ciencias Biológicas (UNAM).
biotas of a single horizon already outlined’ (Rosen &
Smith, 1988: 285).
When applying PAE as implemented in the present
study, two things come together: the insightful panbio-
REFERENCES
geographical concepts and the methodological strength Cavalcanti MJ. 2009. Croizat – a software package for
of parsimony analysis. The results obtained in the quantitative analysis in panbiogeography. Available at
present study show that PAE as a panbiogeographical http://croizat.sourceforge.net

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
970 A. ECHEVERRY and J. J. MORRONE

Chase MW. 1986. A monograph of Leochilus (Orchidaceae). hacia una síntesis de su conocimiento, Vol. II. Mexico DF:
Systematic Botany Monographs 14: 1–97. Facultad de Ciencias, UNAM, 61–94.
Corona AM, Toledo VH, Morrone JJ. 2007. Does the Franco P, Berg CC. 1997. Distributional patterns of Cecro-
Trans-mexican Volcanic Belt represent a natural biogeo- pia (Cecropiaceae): a panbiogeographic analysis. Caldasia
graphical unit? An analysis of the distributional patterns of 19: 285–296.
Coleoptera. Journal of Biogeography 34: 1008–1015. García-Barros E, Gurrea P, Luciáñez MJ, Cano JM,
Cracraft J. 1991. Patterns of diversification within continen- Munguira ML, Moreno JC, Sainz H, Sainz MJ, Simón
tal biotas: hierarchical congruence among the areas of ende- JC. 2002. Parsimony analysis ofendemicity and its applica-
mism of Australian vertebrates. Australian Systematic tion to animal and plant geographical distributions in the
Botany 4: 211–227. Ibero-Balearic region (western Mediterranean). Journal of
Craw RC. 1988a. Continuing the synthesis between panbio- Biogeography 29: 109–124.

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

geography, phylogenetic systematics and geology as illus- Garzón-Orduña IJ, Miranda-Esquivel DR, Donato M.
trated by empirical studies on the biogeography of New 2008. Parsimony analysis of endemicity describes but does
Zealand and the Chatham Islands. Systematic Zoology 37: not explain: an illustrated critique. Journal of Biogeography
291–310. 35: 903–913.
Craw RC. 1988b. Panbiogeography: method and synthesis in Goloboff P. 1998. NONA, Version 2.0. Available at http://
biogeography. In: Myers AA, Giller PS, eds. Analytical bio- www.cladistics.com/about_nona.htm
geography: an integrated approach to the study of animal Grehan J. 2003. Panbiogeografía y la geografía de la vida.
and plant distributions. London: Chapman and Hall, 405– In: Llorente J, Morrone JJ, eds. Introducción a la bio-
435. geografía en Latinoamérica: teorías, conceptos, métodos y
Craw RC. 1989a. New Zealand biogeography: a panbiogeo- aplicaciones. Mexico DF: Facultad de Ciencias, UNAM,
graphic approach. New Zealand Journal of Zoology 16: 181–195.
527–547. Harold AS, Mooi RD. 1994. Areas of endemism:
Craw RC. 1989b. Quantitative panbiogeography: introduc- definition and recognition criteria. Systematic Biology 43:
tion to methods. New Zealand Journal of Zoology 16: 485– 261–266.
494. Henderson A, Galeano G, Bernal R. 1995. Field guide to
Craw RC, Grehan JR, Heads MJ. 1999. Panbiogeography: the palms of the Americas. Princeton, NJ: Princeton Uni-
tracking the history of life. New York, NY: Oxford University versity Press.
Press. Humphries CJ. 1989. Any advance on assumption 2?
Crisci JV, Katinas L, Posadas P. 2003. Historical biogeog- Journal of Biogeography 16: 101–102.
raphy: an introduction. Cambridge, MA: Harvard Univer- Humphries CJ, Parenti LR. 1999. Cladistic biogeography,
sity Press. 2nd edn. Oxford Biogeography Series 12, Oxford: Oxford
Croat TB. 1975. A reconsideration of Trichilia cipo (A. Juss.) University Press.
C. DC. (Meliaceae). Annals of the Missouri Botanical Ippi S, Flores V. 2001. Las tortugas neotropicales y
Garden 62: 491–496. sus áreas de endemismo. Acta Zoológica Mexicana 84:
Croat TB. 1983. A revision of the genus Anthurium (Araceae) 49–63.
of Mexico and Central America. Part I: Mexico and Middle Liria J. 2008. Sistemas de información geográfica y análisis
America. Annals of the Missouri Botanical Garden 70: 211– espaciales: un método combinado para realizar estudios
416. panbiogeográficos. Revista Mexicana de Biodiversidad 79:
Croat TB. 1997. A revision of Philodendron subgenus Philo- 281–284.
dendron (Araceae) for Mexico and Central America. Annals Luna-Vega I, Alcántara Ayala O, Morrone JJ, Espinosa
of the Missouri Botanical Garden 84: 311–704. Organista D. 2000. Track analysis and conservation pri-
Croizat L. 1958. Panbiogeography. Published by the author, orities in the cloud forests of Hidalgo, México. Diversity and
Caracas. Distributions 6: 137–143.
Croizat L. 1964. Space, time, form: the biological synthesis. Luna-Vega I, Morrone JJ, Alcántara Ayala O, Espinosa
Published by the author, Caracas. Organista D. 2001. Biogeographical affinities among Neo-
Da Silva JMC, Oren DC. 1996. Application of parsimony tropical cloud forests. Plant Systematics and Evolution 228:
analysis of endemicity in Amazonian biogeography: an 229–239.
example with primates. Biological Journal of the Linnean Moran RC. 1991. Monograph of the Neotropical fern genus
Society 39: 427–437. Stigmatopteris (Dryopteridaceae). Annals of the Missouri
Escalante T, Szumik C, Morrone JJ. 2009. Areas of ende- Botanical Garden 78: 857–914.
mism of Mexican mammals: reanalysis applying the opti- Morrone JJ. 1994. On the identification of areas of ende-
mality criterion. Biological Journal of the Linnean Society mism. Systematic Biology 43: 438–441.
98: 468–478. Morrone JJ. 1998. On Udvardy’s insulantarctica province: a
Espinosa-Organista D, Morrone JJ, Aguilar C, Llorente test from the weevils (Coleoptera: Curculionoidea). Journal
J. 2000. Regionalización biogeográfica de México: provincias of Biogeography 25: 947–955.
bióticas. In: Llorente J, González E, Papavero N, eds. Biodi- Morrone JJ. 2001. Homology, biogeography and areas of
versidad, taxonomía y biogeografía de artrópodos de México: endemism. Diversity and Distributions 7: 297–300.

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
 PAE AND PANBIOGEOGRAPHY 971

Morrone JJ. 2004a. Homología biogeográfica: las coordena- Rojas-Parra CA. 2007. Una herramienta automatizada para
das espaciales de la vida. Mexico DF: Cuadernos del Insti- realizar análisis panbiogeográficos. Biogeografía 1: 31–
tuto de Biología 37, Instituto de Biología, UNAM. 33.
Morrone JJ. 2004b. Panbiogeografía, componentes bióticos y Rosen BR. 1988. From fossils to earth history: applied his-
zonas de transición. Revista Brasileira de Entomologia 48: torical biogeography. In: Myers AA, Giller PS, eds. Analyti-
149–162. cal biogeography: an integrated approach to the study of
Morrone JJ. 2005. Cladistic biogeography: identity and animal and plant distributions. London: Chapman and
place. Journal of Biogeography 32: 1281–1284. Hall, 437–481.
Morrone JJ. 2006. Biogeographic areas and transition zones Rosen BR, Smith AB. 1988. Tectonics from fossils?:
of Latin America and the Caribbean Islands based on pan- analysis of reef-coral and sea-urchin distributions from
biogeographic and cladistic analyses of the entomofauna. late Cretaceous to Recent, using a new method. In:

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

Annual Review of Entomology 51: 467–494. Audley-Charles MG, Hallam A, eds. Gondwana and tethys.
Morrone JJ. 2009. Evolutionary biogeography: an integrative London: Geological Society Special Publication no. 37, 275–
approach with case studies. New York, NY: Columbia Uni- 306.
versity Press. Salzman VT, Judd WS. 1995. A revision of the Greater
Morrone JJ, Crisci JV. 1995. Historical biogeography: Antillean species of Bactris (Bactridinae: Arecaceae). Brit-
introduction to methods. Annual Review of Ecology and tonia 47: 345–371.
Systematics 26: 373–401. Santos CMD. 2005. Parsimony analysis of endemicity:
Morrone JJ, Márquez J. 2001. Halffter’s Mexican Transi- time for an epitaph? Journal of Biogeography 32: 1284–
tion Zone, beetle generalized tracks, and geographical 1286.
homology. Journal of Biogeography 28: 635–650. Smith AB. 1992. Echinoid distribution in the Cenomanian:
Morrone JJ, Espinosa-Organista D, Aguilar Zúñiga C, an analytical study in biogeography. Palaeogeography,
Llorente-Bousquets J. 1999. Preliminary classification of Palaeoclimatology, Palaeoecology 92: 263–276.
the Mexican biogeographic provinces: a parsimony analysis Sousa M. 1993. El género Inga (Leguminosae: Mimosoideae)
of endemicity based on plant, insect, and bird taxa. South- del sur de México y Centroamérica: estudio previo para la
western Naturalist 44: 508–515. Flora Mesoamericana. Annals of the Missouri Botanical
Murakami N, Moran RC. 1993. Monograph of the Neotro- Garden 80: 223–269.
pical species of Asplenium sect. Hymenasplenium (Aspleni- Swofford DL. 2001. PAUP*: phylogenetic analysis using par-
aceae). Annals of the Missouri Botanical Garden 80: 1–38. simony. Vers. 4.0b10. Sunderland, MA: Sinauer.
Myers AA. 1991. How did Hawaii accumulate its biota? A test Szumik C, Goloboff P. 2004. Areas of endemism: an
from the Amphipoda. Global Ecology and Biogeography improved optimality criterion. Systematic Biology 53: 968–
Letters 1: 24–29. 977.
Nelson G, Ladiges PY. 1991. Three-area statements: stan- Taylor CM. 1989. Revision of Palicourea (Rubiaceae)
dard assumptions for biogeographic analysis. Systematic in Mexico and Central America. Systematic Botany
Zoology 40: 470–485. Monographs 26: 1–102.
Nelson G, Ladiges PY. 1996. Parology in cladistic biogeog- Taylor CM. 1992. Revision of Cosmibuena (Rubiaceae).
raphy and analysis of parology-free subtrees. American Annals of the Missouri Botanical Garden 79: 886–900.
Museum Novitates 3167: 1–58. Taylor CM. 1994. Revision of Hillia (Rubiaceae). Annals of
Nelson G, Platnick N. 1981. Systematics and biogeography: the Missouri Botanical Garden 81: 571–609.
cladistics and vicariance. New York, NY: Columbia Univer- Taylor CM. 1997. Conspectus of the genus Palicourea (Rubi-
sity Press. aceae: Psychotrieae) with the description of some new
Nihei SS. 2006. Misconceptions about parsimony analysis of species from Ecuador and Colombia. Annals of the Missouri
endemicity. Journal of Biogeography 33: 2099–2106. Botanical Garden 84: 224–262.
Nixon KC. 1999. About Winclada. Ithaca, NY. Available at Watanabe K. 1998. Parsimony analysis of the distribution
http://www.cladistics.com/about_winc.htm patterns of Japanese primary freshwater fishes, and
Page RDM. 1987. Graphs and generalized tracks: quantify- its application to the distribution of the bagrid catfishes.
ing Croizat’s panbiogeography. Systematic Zoology 36: 1–17. Ichthyological Research 45: 259–270.
Page RDM. 1994. Maps between trees and cladistic analysis van der Werff H. 1993. A revision of the genus Pleurothy-
of historical associations among genes, organisms, and rium (Lauraceae). Annals of the Missouri Botanical Garden
areas. Systematic Biology 43: 58–77. 80: 39–118.
Pfeifer HW. 1966. Revision of the North and Central Ameri- van der Werff H. 2002. A synopsis of Ocotea (Lauraceae) in
can hexandrous species of Aristolochia (Aristolochiaceae). Central America and southern Mexico. Annals of the Mis-
Annals of the Missouri Botanical Garden 53: 115–196. souri Botanical Garden 89: 429–451.
Porzecanski AL, Cracraft J. 2005. Cladistic analysis of Wiley EO. 1987. Methods in vicariance biogeography. In:
distributions and endemism (CADE): using raw distribu- Hovenkamp P, Gittenberger E, Hennipman E, de Jong R,
tions of birds to unravel the biogeography of the South Roos MC, Sluys R, Zandee M, eds. Systematics and evolu-
American aridlands. Journal of Biogeography 32: 261– tion: a matter of diversity. Utrecht: Utrecht University,
265. 283–306.

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
972 A. ECHEVERRY and J. J. MORRONE

Wiley EO. 1988. Parsimony analysis and vicariance biogeog- Zuloaga FO, Sendulsky T. 1988. A revision of Panicum
raphy. Systematic Zoology 37: 271–290. subgenus Phanopyrum section Stolonifera (Poaceae: Pan-
Zuloaga FO. 1987. A revision of Panicum subgenus Panicum iceae). Annals of the Missouri Botanical Garden 75: 420–
section Rudgeana (Poaceae: Paniceae). Annals of the Mis- 455.
souri Botanical Garden 74: 463–478.

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
 APPENDIX
Table A1. Data matrix of geographic units ¥ individual tracks

root
0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Son
0000000000000000000000000000000000000000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
MPl
0000001101000110000000001001000001000000000000000110000010000000000000000000000000010000000000000011000000000000000000000000000000000000000000000000
Smw
0000001000000000000100000000000000000000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Sme
0000001101001110000000011001000000001000000000000110000010000000000000000000000000010000000000000011000000000000000000000000000000000000000000000000
Tvb
0000011100001010000100011011000001000100100011000100001011100010001000001100100000000100000000001111010010000000000000000000000000000000010000000000
Bal
0000001000001010000100001001000000000100100001000100001010110010001000001100100000000000000000001001010010000000000000000000000000000000010000000000
Sms
0000111100000011000101111011001001000100100011000110001111111010001010001100100000010100000010001111010010010000000000001100000010000000010000000011
MPc
1000111111110011001111111001010011000100100011100110001101111010001110001101100000000000000010001010010010111000000100001100010011000000010010100011
MGu
1100111111111111110110111111001101101000100011001111001111000010000110000011110000011101000010000111011000110010000100101100000000000000000000100011
YPe
0000001111111001000010100001101000100000100000000010000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Chi
1100111111110111011111111101011011101100100011000111001111011010001111000111100000011101001010001011011010111010000100101100000011000000111010100011
ECa
1111111111110111111111111111111111111111101101111111111111111111011101000011011100011111111101111010111111111111011111111110110011000000111111101111
WPI
1010111010110001010011011100011101010001101111011111001011100011000000000011011000011010111100111000010100111111010111111100110010000000100101000111
Bah
0000000000100000000000000000100000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Cub
1000001000111100101110100100101010100000000000001010000010000010000000010000000000000000000000000000000000000000000000001000000000000000000000000001
Cay
0000000000000000100000000000100000000000000000000010000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Jam
1000000101100110101110110000101100100000000000000100000001000000000000000010000000000000000000000000000000000000000000000000000000000000000000000001

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
His
0000010101101111101000110000001010100000000000000100000000000000000000110000000000000000000000000000010000000000000000000000000000000000000000000001
PRi
0000001100000111101000110000001000100000000000000100000000000000000000110000000000000000000000000000010000000010000000000010000000000000000000000001
LAn
0000001000000100101110100000000000000000000000000100000000000000000000110000000000000000000100000000010000000000000000000010000000000000000000000001
Cho
PAE AND PANBIOGEOGRAPHY

1110011001110011111111111110011101111111001101111111111111101011110000000011011100011110111101110010111111111110111111011010110000010001100101111001
Mar
1000011110110011011101110110011111100010010100001110000000001010000100000000010000000110010100000000000000110110000100001111010000001000100001000000
973

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

 974

Table A1. Continued

Vco
1000001110110100011100110010000011101000000100000100000010001000000100000000000000000000001000010000000000011010000010011110000000001000100001011001
T&T
0000000000011000100010110010000101000000000100001100000010001000000100000000000000000000000100010000010000010000000000000100000000000000000000000001
Mag
0010011100110101011111110110011101111101010000101101100111101011000000000011011000001110011100110000111011111110101011011010010000100001100101011001
Vel
0010001100100101011110110110011101111010000100101110000011000000100000000000010000011110011100010000011000111100100110011111010000101000100001011001
Cau
0111011100100011011111111110001001111101011011111101110011101011110000000011000111111000111101110000011111111111111111011010111000010011100101001101
GaI
0000001000000000000000000000000000000000000000000000000000000000000000000001000000000000000000000000000000000000000000000000000000000000000000000000
WEc
A. ECHEVERRY and J. J. MORRONE

0101001000110001010110010000000001000100000010010101001001101111010000000001000011001100011100000000011000001011010000011000100001000011100000001101
AEc
0100001000000001010000000000000001000000000000010000001001101010010000000000000001001100010000000000011000000010000000001000000000000000000000001000
TuP
0100000100000001000000010000000001000000000000010000000000000000000000000000010000000100001100010000001000010000010000000000000001010000000000001000
Nap
0101011000100001011101110100001101010001011000111100010010001100100000000010010000010010111101010000011000111110111110011001111100110110100001001001
Ime
0110001000100000100000110100001100000000011100101100000001000000100000000010010000010010000100010000011000101000011110010000000000110000000000000001
Guy
0000001100100000001000110110001100000010001100101100000001000000000000000010010000010110011100010000011000011100110010010000010000110000000000011001
Ror
0000001000000000100000110100000100000000001100101100000000000000000000000010010000010010001100000000001000011000100110010000010000110000000000001001
Var
0000011000000000110000110100001100000001011100101100000001000000000000000010010000000010111101010000011000111100011110010000001000010110100000001001
Uca
0100001000000000010000110100000100000001000000001000000000001000000000000010010000000010010101010000011000000000010010011001011100110000000001000001
Mad
0000001000000000100000110100001100000001010100101100000001000000000000000000010000000010011101010000010000111000001110010000000000010100100000001001
NoA
0111011100100111011111111110010101011011011011111111010011101101100000000010010110111110111101010000011000111111111111011111111100111010100001011101
Flo
0000000000000000000000000000000000100000000010000000000010000010000000000000000000000000000000000000000000000000000000000000000000000000000000000000

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019
 Table A2. Data matrix of geographic units ¥ single sample localities

root
0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Son
0000000000000000000000000000000000000000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
MPl
0000001101000110000000001000000001000000000000000110000010000000000000000000000000010000000000000011000000000000000000000000000000000000000000000000
Smw
0000000000000000000000000000000000000000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Sme
0000001000000100000000011001000000000000000000000100000010000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Tvb
0000011100001010000000001010000000000100100011000100000011100010001000001100100000000000000000001010010010000000000000000000000000000000000000000000
Bal
0000001000001010000000001001000000000000100001000000001010010000000000000000000000000000000000001001000000000000000000000000000000000000000000000000
Sms
0000111100000011000100111001001001000100000011000110000111001010000000000100000000010100000010001111010000010000000000001100000010000000010000000011
MPc
1000111111110011000111111001010010000000100011100110001101010010000110001101100000000000000000001000010000010000000000001100010001000000010010100011
MGu
1100111111111111110110111111001101101000000011001111001111000010000110000011110000011100000010000111011000110010000100101100000000000000000000100001
YPe
0000001111110001000010100001101000000000100000000010000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Chi
1000011111110111011111111101011011101100100011000111001111010010001100000110100000011101001010001011011010011000000100001100000011000000010010100001
ECa
1111111011110011111111111111111101111111101101111111111111110111011001000011011100011011111101111010111111111111011111111110110011000000111111101111
WPI
1010111010110001010011011100011101000001101101011111000011100011000000000011011000011010011100111000010100111111010110111100010010000000100100000011
Bah
0000000000000000000000000000100000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Cub
1000001000011100101110100100101010000000000000001010000000000000000000010000000000000000000000000000000000000000000000001000000000000000000000000001
Cay
0000000000000000000000000000000000000000000000000010000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Jam
0000000101100110101110110000101100100000000000000100000001000000000000000010000000000000000000000000000000000000000000000000000000000000000000000001
His
0000010101100111101000110000001010100000000000000100000000000000000000110000000000000000000000000000010000000000000000000000000000000000000000000001

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
PRi
0000001100000111101000110000001000100000000000000100000000000000000000000000000000000000000000000000010000000010000000000010000000000000000000000001
LAn
0000001000000100100110100000000000000000000000000100000000000000000000110000000000000000000100000000010000000000000000000010000000000000000000000001
Cho
0110001001100011010010111110000001000010001100111111111011101011110000000011010000011000010101110010001101111110111110011000000000010000100000111001
PAE AND PANBIOGEOGRAPHY

Mar
1000001000100010011101110110011111100010010100001110000000001010000100000000010000000000010100000000000000010100000100001111010000001000100001000000
975

Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019

 976

Table A2. Continued

Vco
1000001010110100011000110010000011101000000100000100000010001000000100000000000000000000001000010000000000011000000010011010000000001000100001011001
T&T
0000000000011000100010110010000101000000000100000100000010001000000100000000000000000000000000010000010000010000000000000100000000000000000000000001
Mag
0010011100100101011111110110010101010101010000101100100111101011000000000011011000001000011100110000111011011110101011011010010000100001000101001000
Vel
0000001100100001001100110110011101111000000100101110000010000000000000000000010000001110000100000000010000111000000100011111000000101000100000000001
Cau
0101001000100011011101110010001001111100011000011101010011101011110000000011000001011000011100000000011110011010011001000000111000000011100101001101
WEc
A. ECHEVERRY and J. J. MORRONE

0101001000010001010110010000000001000100000000010101000000101011000000000001000011000000011100000000011000001011010000011000100000000011100000001001
AEc
0100001000000000010000000000000001000000000000000000001000100010010000000000000001001000010000000000011000000010000000001000000000000000000000001000
TuP
0100000100000001000000010000000000000000000000010000000000000000000000000000010000000100001100010000000000010000010000000000000001010000000000001000
Nap
0101001000100001011101110100001101010001001000111100010000001100000000000010010000010010111101010000011000110010110010010001110000110100100001001001
Ime
0110000000100000000000110100001100000000001000101000000000000000100000000010010000010010000100010000011000001000000100010000000000010000000000000001
Guy
0000001100100000001000110100001100000010001100101100000001000000000000000000010000000110000100010000011000001100110010010000010000100000000000011001
Ror
0000001000000000100000110100000100000000001100101100000000000000000000000010010000010010001000000000001000011000100010010000010000110000000000001000
Var
0000001000000000010000110000000000000001001000001100000000000000000000000000010000000010111101010000011000001100010000010000001000010010100000000001
Uca
0100001000000000010000010100000000000001000000001000000000001000000000000010010000000000010100000000011000000000010000010001000100100000000000000001
Mad
0000001000000000100000100000001100000001010000101100000000000000000000000000010000000010000101000000010000111000001110010000000000000100100000001001
NoA
0100011100100011011101011110000101010001000011011101000011101101000000000000010110111000111100010000011000001011110111010000011100110010100001011101

© 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 961–976
Downloaded from https://academic.oup.com/biolinnean/article-abstract/101/4/961/2450582 by guest on 21 May 2019