Académique Documents
Professionnel Documents
Culture Documents
Abstract
A cladistic biogeographical analysis was undertaken to identify the main events in the biotic diversification of the terrestrial
Neotropical biota. For the 36 animal and plant taxa analysed, a component 9 area matrix was constructed, associating geographical data only with informative nodes, and it was analysed under implied weights using the software TNT. The general
area cladogram obtained shows that the Neotropical region constitutes a monophyletic unit, with a first split separating the
Antilles and a second one dividing the continental areas into a north-western and a south-eastern component. Within the northwestern component the areas split following the sequence northern Amazonia, south-western Amazonia, north-western South
America, and Mesoamerica. Within the south-eastern component the areas split following the sequence south-eastern Amazonia,
Chaco, and Parana. The three main components are treated as subregions: Antillean, Amazonian (northern Amazonian, southwestern Amazonian, Mesoamerican, and north-western South American dominions), and Chacoan (south-eastern Amazonian,
Chacoan, and Parana dominions). Dispersal and vicariant events postulated to explain these pattens might have occurred during
the Cretaceous, when the Caribbean plate collided with the Americas, a combination of eustatic sea-level changes and tectonic
deformations of the continental platform exposed large parts of South America to episodes of marine transgressions, and the
Andean uplift reconfigured the Amazonian area. Tertiary and Quaternary events are assumed to have later induced the diversification within these large biogeographical units.
The Willi Hennig Society 2013.
203
Fig. 1. The Neotropical region, with the Mexican and South American transition zones represented by dashed lines. The Neotropical
region sensu stricto does not include these transition zones, whereas
the Neotropical region sensu lato encompasses them.
nian. For South America, several authors have recognized two subregions: one named Guyano-Brazilian
or Brazilian, and the other named Andean-Patagonian, Patagonian, Argentinean, Chilean, or Austral
(Cabrera and Yepes, 1940; Ringuelet, 1961, 1975; Fittkau, 1969; Hershkovitz, 1969; Kuschel, 1969; Sick,
1969; Smith, 1983; Rivas-Martnez and Navarro,
1994; Almir
on et al., 1997; Kreft and Jetz, 2010;
Proches and Ramdhani, 2012). This main division
within South America has been also evidenced by
ecogeographical (Bailey, 1998) and macroecological
analyses (Ruggiero et al., 1998; Ruggiero and
Ezcurra, 2003). Morrone (2001) proposed a classification of the Neotropical region, based on panbiogeographical analyses, dividing it into four subregions:
Caribbean, Amazonian, Chacoan, and Parana. Morrone and Coscar
on (1996) undertook a parsimony
analysis of endemicity based on South American Peiratinae (Heteroptera: Reduviidae), hypothesizing that
the gradual development of an open vegetation diagonalcomprising the Chaco, Cerrado, and Caatinga
separated a former forest into a north-western part
(north-western South America and Amazonian forest)
and a south-eastern part (Parana and Atlantic
204
forests). This hypothesis was corroborated by a cladistic biogeographical analysis (Morrone and Coscar
on, 1998). Other cladistic biogeographical analyses
(Amorim, 2001; Nihei and Carvalho, 2007; Sigrist
and Carvalho, 2009; Pires and Marinoni, 2010;
Echeverry and Morrone, in press) have questioned
the naturalness of the Amazonian and Caribbean
subregions of Morrone (2006).
My objective is to analyse the relationships of different areas of the Neotropical region to test its naturalness, revise its regionalization, and identify the main
events in the biotic diversification of the terrestrial
Neotropical biota.
Fig. 2. Units of the analysis. (a) Nearctic region; (b) Mexican Transition Zone; (c) Mesoamerica; (d) Antilles; (e) north-western South
America; (f) northern Amazonia; (g) south-eastern Amazonia; (h)
south-western Amazonia; (i) Chaco; (j) Parana; (k) South American
Transition Zone; (l) Andean region.
Taxa
Thirty-six taxa (Table 1) were analysed. They
include genera and species groups of insects, spiders,
vertebrates, and plants, which were chosen because
phylogenetic hypotheses were available for them and
their species are distributed in the areas analysed
herein. Although there are other potential taxa with
published phylogenetic analyses that could be considered, some of them lack information on the geographical distribution of the species analysed and others
include only a limited sample of the terminal taxa, as
occurs in most molecular analyses. Some of the taxa
analysed herein have been considered in previous cladistic biogeographical studies (Morrone and Coscar
on,
1998; Nihei and Carvalho, 2007; Sigrist and Carvalho,
2009; Pires and Marinoni, 2010; Echeverry and Morrone, in press).
Analysis
A cladistic biogeographical analysis is based on a
correspondence between phylogenetic relationships
205
Paralogy-free subtrees
Reference(s)
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
4a
4b
4c,d
4eg
4h,i
4j
Miller (2001)
Chani-Posse (2006)
Morrone (2002b), Vanin and Gaiger (2005)
Mermudes (2005)
Mermudes and Napp (2006)
Morrone and Cuevas (2002)
Fig.
Fig.
Fig.
Fig.
4k
4l
4m
4n
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
4o
4p
4q,r
4s,t
4u,v
4w
Coscar
on (1989), Morrone and Coscar
on (1998)
Albertson and Dietrich (2005)
Morrone and Coscar
on (1998)
Paula et al. (2007)
Fortes and Grazia (2005)
Bernardes et al. (2009)
Fig. 4xz, aa
Fig. 4bb
Fig. 4ccff
Fig. 4ggii
Fig. 4jj
Agnarsson (2005)
Fig. 4kk
Fig. 4ll
Fig. 4mm
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
4nn
4oo
4pp
4qq
4rr
4ss
4tt
Fig.
Fig.
Fig.
Fig.
Fig.
4uu
4vvxx
4yy, zz
4aaa
4bbb, ccc
To obtain the general area cladogram(s), a parsimony analysis of paralogy-free subtrees (Nelson and
Ladiges, 1996; Contreras-Medina et al., 2007; Le
on
Paniagua and Morrone, 2009; Morrone, 2009) was
undertaken. This method consists of four steps
(Fig. 3):
1 For each taxonarea cladogram, areas duplicated
or redundant in the descendants of a node are eliminated (Fig. 3a), so that geographical paralogy is eliminated or reduced significantly, and data are associated
only with informative nodes.
2 Components are identified on each of the paralogy-free subtrees obtained (Fig. 3b).
206
(a)
(b)
(c)
(d)
Fig. 3. Steps of the cladistic biogeographical analysis. (a) Original taxonarea cladograms; (b) paralogy-free subtrees derived from the taxon
area cladograms, with informative components marked on them; (c) data matrix of areas 9 components; (d) general area cladogram obtained.
Results
Two most parsimonious cladograms were obtained
analysing the matrix under equal weights. Implied
weights calculated with different constants of concavity (k = 130) consistently produced one of them
(Fig. 5). This general area cladogram shows that the
Neotropical areas in the strict sense constitute a monophyletic unit, whereas the Mexican and South American transition zones are the sister areas to the Nearctic
and Andean regions, respectively. Within the Neotropical region in the strict sense, a first dichotomy separates the Antilles from the continental areas, which in
a second dichotomy are arranged into a north-western
and a south-eastern component. Within the north-western component the areas split according to the
sequence northern Amazonia, south-western Amazonia, north-western South America, and Mesoamerica.
Within the south-eastern component the areas split following the sequence south-eastern Amazonia, Chaco,
and Parana. According to these results, the Caribbean
and Amazonian subregions as formerly recognized by
Morrone (2006) under a panbiogeographical framework do not represent natural areas: within the former, the Antilles do not group with Mesoamerica and
north-western South America; and within the latter,
south-eastern Amazonia is closer to Chaco and Parana
than to the remaining Amazonian areas.
Based on the general area cladogram, the following
regionalization is proposed:
Neotropical region
Antillean subregion
Amazonian subregion
Northern Amazonian dominion
South-western Amazonian dominion
207
(a)
(d)
(g)
(f)
(e)
(c)
(b)
(l)
(i)
(h)
(j)
(k)
(m)
(o)
(s)
(q)
(p)
(r)
(n)
(v)
(u)
(x)
(y)
(w)
(z)
(aa)
(dd)
(bb)
(cc)
(ee)
(ff)
(gg)
(ll)
(hh)
(ii)
(mm)
(kk)
(jj)
(rr)
(pp)
(qq)
(oo)
(nn)
(ww)
(uu)
(tt)
(vv)
(ss)
(yy)
(zz)
(xx)
(aaa)
(ccc)
(bbb)
Fig. 4. Paralogy-free subtrees analysed. (a) Agaporomorphus; (b) Chroaptomus; (c,d) Entimini; (eg) Hypselotropis; (h, i) Ptychoderes; (j) Rhinostomus; (k) Coenopsia; (l) Polietina; (m) Pseudoptilolepis; (n) Sepedonea; (o) Eidmannia; (p) Nicomia; (q,r) Rasahus; (s,t) Rhodnius; (u,v) Serdia;
(w) Thoreyella and related genera; (xz, aa) Amorbia; (bb) Charis gynaea species group; (ccff) Cliniodes; (ggii) Abracrini; (jj) Anelosinus ethicus
species group; (kk) Amazona ochrocephala complex; (ll) Chlorospingus ophthalmicus complex; (mm) Pionopsitta; (nn) Alouatta; (oo) Ateles; (pp)
Caluromys; (qq) Marmosa murina; (rr) Metachirus nudicaudatus; (ss) Oryzomys megacephalus species group; (tt) Rhipidomys; (uu) Exostema; (vv
xx) Hillia; (yy, zz) Prosopis; (aaa) Sabal; (bbb, ccc) Stigmatopteris. Areas as in Fig. 2.
Mesoamerican dominion
North-western South American dominion
Chacoan subregion
South-eastern Amazonian dominion
Chacoan dominion
Parana dominion
Some of the events implied by the general area cladogram (Fig. 6) can be associated with known tectonic
and geological information available:
1 The oldest event corresponds to the former
connection between the North American and South
American landmasses during the Early Jurassic to
208
Table 2
Data matrix, where files correspond to the areas analysed and columns to the components identified in the paralogy-free subtrees of Fig. 4
root
00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
Nearctic region
00000000000000000000001000011111011000000000000000000000000000001001000000000000000011101000000000000000000000000000000000000000000000000000011111000000000
Mexican Transition Zone
00111000000000000000001000011111010000000000000000000000000000101000000000000000000000000000000000000000000000000000000000000000001111111000000000000000000
Mesoamerica
00111010000000010100101111111110011111100000001000000110010000110001111110001111011011111100000000100101000100000110100000000000001111111001100111111111111
Antilles
00000000000000000000001111100000000000000000000000000000000000001100001000001100010100000000000000000000000000000000000000000000000000000101100111101000010
North-western South America
00110000000110000110001100011100011100110100101111000111010000001111111111111000010111001111010100110111110111110110100000000000000001110111000111111111111
Northern Amazonia
00000011010111000000111110011000000000110110110011111001100000001111100011101110000000000001100010111001101111100100000100101111001001110100000000001110000
South-eastern Amazonia
10000000000001111111111111000000011100000000000010000000000000001110000000000000000000000001100000000001101110001111011111001111110000000100000000000000000
South-western Amazonia
11100111011101100111111111011000000000000011000000011000011000001111101011111110000011010001110100111111110111110111000110111001101001000111000110001111000
Chaco
11000011100000011111001111010000111110111000000011110000000111001110000011111111100000001111111011000000000111000000010000001110001110000000011110000000000
Parana
11100111111000011111110111110000111111111011111111111001100111111110001111001111111010010001111011000000000110001000011111111101110001100100000000001100000
South American Transition Zone
00000000000000000000000000000000000000000000000000011100011100000000000000000000000000000000000000000000000000000000000000000000001100000000010100000000000
Andean region
00000000000000000000000000000000000000000000000000000000000110000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
209
(a)
(b)
(d)
(g)
(e)
(c)
(f)
(h)
Fig. 6. Sequence of events implied by the general area cladogram. (a) Geodispersal of the Neotropical biota from South to North America; (b)
vicariance of the Antilles and the rest of the Neotropical region; (c) vicariance of the north-western and south-eastern continental components;
(d) vicariance of northern Amazonia and the rest of the north-western component; (e) vicariance of south-western Amazonia and north-western
South AmericaMesoamerica; (f) vicariance of north-western South America and Mesoamerica; (g) vicariance of southern Amazonia and ChacoParana; (h) vicariance of Chaco and Parana.
Early Cretaceous (190148 Ma). It allowed the geodispersal of the Neotropical biota to North America and
would explain the presence of ancient Neotropical elements in northern Mexico (Halffter, 1987; Morrone,
2005b).
210
leading edge of the Caribbean plate reconnected temporarily with North and South America at 125
100 Ma. This connection began to be severed in the
Late Cretaceous (80 Ma) and finished in the Miocene
to Middle Pliocene (Pitman et al., 1993). The continental mainland constituting the rest of the Neotropical region shares extensive hydrological connections
and there is evidence that it constituted a superbasin
that has persisted in relative isolation since at least the
Late Cretaceous (Albert et al., 2011).
3 The vicariance between the north-western and
south-eastern continental components of the Neotropical region might have begun with the formation of a
lake along the Amazon, Madeira, and Mamore rivers,
in the Late Cretaceous; and finished with the formation of a wide epicontinental sea by water invasion
through the north, east, and south portal seaways, in
the Miocene (Amorim, 2001; Frailey, 2002; Nihei and
Carvalho, 2004, 2007).
4 The vicariance between northern Amazonia and
the remaining areas of the north-western continental
component can be linked to the Romeral Fault Zone
and/or the final uplift of the northern Andes. The
Romeral Fault, of Cretaceous age, is an active and
continuous fault system almost 700 km long that comprises three or four parallel regional faults, which form
the boundary between autochthonous continental
rocks to the east and accreted oceanic arc rocks
related to Caribbean terranes in the west (Kennan and
Pindell, 2009; Heads, 2012; Echeverry and Morrone,
in press). The uplift of the Andes began in the Cretaceous, but has been more conspicuous since the Miocene, 237 Ma, and finished in the Pliocene (Lundberg
et al., 1998; Kennan, 2000; Cortes-Ortiz et al., 2003;
Garzione et al., 2008; Hoorn et al., 2010).
5 The vicariance between south-western Amazonia
and north-western South AmericaMesoamerica can
be linked to the formation of an epicontinental sea by
water invasion through the Maracaibo and Amazon
basins in the Pliocene (Rodrguez-Olarte et al., 2011).
6 The vicariance between Chaco and Parana can be
linked to the connection of the Parnaba and Parana
basins in the Palaeocene (Amorim, 2001; Nihei and
Carvalho, 2004).
Discussion
Previous analyses have suggested that the Amazonian and Caribbean subregions might not be natural
areas. Amorim and Pires (1996) considered that
Amazonia consisted of two non-related areas: northwestern Amazonia and south-eastern Amazonia. Amorim (2001) also considered that the Amazon forest did
not correspond to a natural biogeographical area,
being geographically delimited, based on the Amazon
Andean uplift reconfigured the Amazonian area (Garzione et al., 2008; Hoorn et al., 2010). Tertiary and
Quaternary events are assumed to have induced the
diversification within these large biogeographical units
(Bush, 1994; Werneck, 2011). The vicariance between
the Antilles and the rest of the Neotropical region,
and the vicariance between the north-western and
south-eastern continental components are coincident
with previous cladistic biogeographical analyses (Amorim and Pires, 1996; Amorim, 2001; Nihei and Carvalho, 2004). The dispersal of the Neotropical biota
from South to North Americaa prerequisite for the
first vicariant eventhas been supported by several
authors, working on different taxa, who have previously recognized two cenocrons: Old Southern or
Ancient Neotropical that dispersed between the Cretaceous and the Palaeocene; and a younger one that
dispersed between the Pliocene and Pleistocene, after
the establishment of the Panama Isthmus (Rosen,
1976; Gentry, 1982; Savage, 1982; Bussing, 1985;
Halffter, 1987; Morrone, 2005b). It is interesting to
note that the vicariance between north-western South
America and Mesoamerica occurs more recently than
detected in some previous analyses (Amorim and Pires,
1996; Amorim, 2001; Camargo and Pedro, 2003),
where it was considered as an older vicariant event. In
the present analysis this event has a younger age, as
suggested by Camargo (1996) and Camargo and Moure (1996).
Several authors (Bush, 1994; Bates et al., 1998;
Marks et al., 2002; Costa, 2003; Nihei and Carvalho,
2007) have concluded that it is not possible to postulate a single hypothesis explaining the current distribution of the Neotropical terrestrial biota. Pires and
Marinoni (2010) have even suggested that in the
absence of temporal information (inferred from molecular phylogenies) one cannot be sure that a general
pattern is due to a common history of biotic diversification. Additionally, it might be argued that taxa
showing so different dispersal means may not show a
clear, congruent pattern. I think, however, that biogeographical regionalizations are useful as general reference models (Ribichich, 2002). Their heuristic value
should be explored by examining the geographical distribution of other plant and animal taxa. Molecular
phylogenetic analyses that have been calibrated may
allow us to disprove the proposed sequence of vicariant events and help to clarify the time frame of the
events that led to the biotic diversification of the Neotropics.
Acknowledgments
I thank Amparo Echeverry and Dalton de Souza
Amorim for interesting discussions on the biogeogra-
211
212
213
214