JOURNAL OF EXPERIMENTAL ZOOLOGY 309A (2008)
A Journal of Integrative Biology
Genetic Characterization of Captive Cuban Crocodiles
(Crocodylus rhombifer) and Evidence of Hybridization
With the American Crocodile (Crocodylus acutus)
JEREMY P. WEAVER1, DAVID RODRIGUEZ1, MIRYAM VENEGAS-ANAYA1,
JOSÉ ROGELIO CEDEÑO-VÁZQUEZ2, MICHAEL R.J. FORSTNER3, AND
LLEWELLYN D. DENSMORE III1
1
Department of Biological Sciences, Texas Tech University, Lubbock, Texas
2
El Colegio de la Frontera Sur, Unidad Chetumal, Chetumal, Q. Roo, Me´xico
3
Department of Biology, Texas State University-San Marcos, San Marcos, Texas
ABSTRACT
There is a surprising lack of genetic data for the Cuban crocodile (Crocodylus
rhombifer), especially given its status as a critically endangered species. Samples from captive
individuals were used to genetically characterize this species in comparison with other New World
crocodilians. Partial mitochondrial sequence data were generated from cyt-b (843 bp) and the
tRNAPro- tRNAPhe-D-loop region (442 bp). Phylogenetic analyses were performed by generating
maximum parsimony, maximum likelihood, and Bayesian-based topologies. In addition, in an effort
to identify species-specific alleles, ten polymorphic microsatellite loci were genotyped. Distance and
model-based clustering analyses were performed on microsatellite data, in addition to a model-based
assignment of hybrid types. Both mitochondrial and nuclear markers identified two distinct
C. rhombifer genetic sub-clades (a and b); and microsatellite analyses revealed that most admixed
individuals were F2 hybrids between C. rhombifer-a and the American crocodile (C. acutus). All
individuals in the C. rhombifer-b group were morphologically identified as C. acutus and formed a
r 2008 Wiley-Liss, Inc.
distinct genetic assemblage. J. Exp. Zool. 309A, 2008.
How to cite this article: Weaver JP, Rodriguez D, Venegas-Anaya M, Cedeño-Vázquez JR,
Forstner MRJ, Densmore III LD. 2008. Genetic characterization of captive Cuban
crocodiles (Crocodylus rhombifer) and evidence of hybridization with the American
crocodile (Crocodylus acutus). J. Exp. Zool. 309A:[page range].
Cuban crocodiles (Crocodylus rhombifer) are
considered endangered (CITES Appendix, IUCN
Red List-EN) due to their limited distribution,
habitat loss, and the introduction of exotic animals
into their environment (Ross, ’98). The recent
encroachment of humans into C. rhombifer’s
territory has limited its distributional range to
about 186 square miles (300 km2) within Cienaga
de Zapata and Cienaga de Lanier in southwestern
Cuba (Fig. 1). Sub-fossils of C. rhombifer found in
Cuba are dated to the Pleistocene (Varona, ’66;
’84); whereas several sub-fossils found on Grand
Cayman (Morgan et al., ’93) and the Bahamas
(Franz et al., ’95) are from the Holocene, which
suggests it was historically found throughout most
of the Caribbean. It has been suggested that a
possible contributing factor to the decline
C. rhombifer over the past 25 years may have
been the introduction of the Brown caiman
r 2008 WILEY-LISS, INC.
(Caiman crocodilus fuscus) into Cienaga de Lanier
(Ross, ’98). Although recent reports by R. Soberón
(personal communication) have indicated that this
is not possible because the C. rhombifer population
on that island was already extirpated before
C. c. fuscus was introduced. C. rhombifer is also
naturally sympatric with the American crocodile
(C. acutus), but C. acutus has a more extensive
distribution that extends from North America into
South America and the Caribbean (Ramos et al.,
’94; Thorbjarnarson et al., 2006).
Grant sponsor: Howard Hughes Medical Institute Grant.
Correspondence to: Jeremy P. Weaver, Department of Biological
Sciences, Texas Tech University, Lubbock, TX 79409.
E-mail: jeremy.weaver@ttu.edu
Received 16 July 2007; Revised 14 February 2008; Accepted 4 May
2008
Published online in Wiley InterScience (www.interscience.wiley.
com). DOI: 10.1002/jez.471
2
WEAVER ET AL.
Fig. 1. Map of Cuba with localities of C. rhombifer-like haplotypes (a and b) found outside their present range (vertical
stripes).
The Cuban crocodile’s breeding season overlaps
with that of C. acutus by a few days in the month
of January (Varona, ’66). The variation in length
between these two animals is approximately 1.5 m
with adult C. acutus males reaching 5 m and
C. rhombifer reaching 3.5 m (Varona, ’66); thus,
making hybridization physically possible. There
have been several documented cases of hybridization between crocodiles in captive populations;
most pertinently between C. acutus and C.
rhombifer at the Laguna del Tesoro farm in Cuba
(Ross, ’98), and consequently these hybrids could
have been distributed to US zoological parks and
private collectors.
The ability to detect hybrids is essential in
identifying pure breeding populations for reintroductions into extirpated areas (Allendorf et al.,
2001; FitzSimmons et al., 2002). Despite the
critical status of wild C. rhombifer populations
there has been little or no genetic data published
or reported on this species; although, there is an
ongoing ecological study that will include some
genetic analyses (R. Ramos and O. Sanjur,
personal communication). Hybrid introgression
has been detected in some New World crocodilians
(Hekkala, 2004; Ray et al., 2004; Rodriguez, 2007;
Cedeño-Vázquez et al., 2008; Rodriguez et al.,
2008), and owing to C. rhombifer’s smaller
population numbers and its frequent sympatry
with C. acutus, the genetic integrity of this species
is at risk. It has been suggested that detecting
hybrids is less exhaustive when the two parental
crocodiles possess different karyotypes, but detection of hybridization between individuals with
J. Exp. Zool.
similar karyotypes requires more in-depth analyses (Chavananikul et al., ’94; FitzSimmons et al.,
2002). Due to the chromosomal and biochemical
similarity (Cohen and Gans, ’70; Densmore, ’83)
and the relatively recent divergence (Brochu,
2000) between C. rhombifer and C. acutus, detecting hybrids based on morphological characters
alone may be problematic. In this specific case, the
use of molecular markers is warranted.
Molecular markers have been used routinely to
characterize threatened species and populations
(Frankham et al., 2002), but genetic studies first
require a point of reference to accurately assess
species
assignments.
Mitochondrial
DNA
(mtDNA) is useful in constructing phylogenies
and haplotype networks. However, the maternal
inheritance of mtDNA limits our ability to detect
hybridization to cases where there is disagreement
between morphology and mtDNA assignments.
Therefore, nuclear DNA (nDNA) markers must
also be utilized to more accurately evaluate species
designations and determine hybrid types. Developing a pure breeding stock of C. rhombifer will be
essential in maintaining the genetic integrity of
the species, which is why any potential hybridization with other species can be a problem in captive
populations. The purpose of our study was to
genetically characterize all available samples of
captive Cuban crocodiles from US zoological
institutions and to describe C. rhombifer-like
haplotypes found in the Caribbean, Florida, and
Mexico, which may present a threat to the genetic
purity of other endemic crocodile species. This
work will provide the foundation for future genetic
3
GENETIC CHARACTERIZATION OF THE CUBAN CROCODILE
treatments of wild Cuban crocodile populations,
assist in further efforts to identify hybrids outside
of Cuba, and assess the utility of genetic methods
in detecting inter-specific admixture within other
captive populations.
MATERIALS AND METHODS
Samples
Wild C. rhombifer populations from Cuba could
not be sampled, but zoological specimens were
readily available. Whole blood or skin clips were
collected from captive C. rhombifer, wild caught
C. acutus, wild caught C. moreletii (Morelet’s
crocodile) and a captive C. intermedius (Orinoco
crocodile) (see Appendix). Blood was collected via
the caudal or dorsal sinus and stored in cell lysis
buffer (Gorzula et al., ’76; Bayliss, ’87), whereas
skin clips were stored in 95% ethanol. Both sets of
tissue were stored at
201C prior to DNA
isolation. Total genomic DNA was extracted using
the PureGene isolation kit (Gentra Systems,
Minneapolis, MN), electrophoresed on a 2%
agarose gel, and visualized with ethidium bromide
under UV light.
Genetic markers
Mitochondrial DNA
A partial cytochrome-b (cyt-b) fragment (843 bp)
was amplified from C. rhombifer, C. acutus,
C. moreletii, and C. intermedius using primers
crCYTBfor and crCYTBrev. Primers drL15459
(modified from Glenn et al., 2002) and CR2HA
(modified from Ray and Densmore, 2002) were
used to amplify the tRNAPro-tRNAPhe-D-loop
region (442 bp; Table 1) for only C. rhombifer
and C. acutus. Polymerase chain reactions (PCR)
were performed in 50 mL volumes using 0.50 mL of
total genomic DNA (tDNA) (50 ng/mL), 36.25 mL of
ddH20, 10 mL of buffer (0.3 M TRIS, 0.0175 M
MgCl2, and 0.075 M (NH4)2SO4), 2.0 mL of 2.5 mM
dNTPs, 0.50 mL (10 mM) of forward primer,
0.50 mL (10 mM) reverse of primer and 0.25 mL
(1.25 U) of Promega Taq polymerase (Promega
Corp., Madison, WI). Thermocycling conditions for
all primers consisted of an initial denaturation
step of 2 min at 941C, then 33 cycles of 30 sec at
941C, 1 min at 581C, and 45 sec at 721C; with a
final extension of 7 min at 721C. Unincorporated
dinucleotides and primers were removed from
PCR products using the Qiagen PCR purification
kit (Qiagen, Inc., Valencia, CA). Products were
cycle sequenced using Big Dye v3.1 dye terminator
TABLE 1. Primer sequences used to generate mtDNA fragments
Primer
crCYTBfor
crCYTBrev
CrCYTBintfor1
drL15459
CR2HA
1
Sequence
50
50
50
50
50
ATGACCCACCAACTACGAAAATC 30
CGAAGGGGTTTGATTAATAGGTT 30
TAGCAACTGCCTTCATAGGCTAC 30
AGGAAAGCGCTGGCCTTGTAA 30
GGGGCCACTAAAAACTGGGGGGA 30
Used only for sequencing.
(Applied Biosystems, Inc., Foster City, CA). Cycle
sequence products were purified by passing
through a G-50 Sephadex column (0.5 gm of
Sephadex/800 mL ddH2O), which was incubated
at room temperature for 30 min and centrifuged at
3,000 rpm for 2 min to construct the column. Dried
cycle sequence product was denatured in formamide and electrophoresed on an ABI 3100-Avant
genetic analyzer (Applied Biosystems, Inc., Foster
City, CA). Chromatograms were viewed and
trimmed using Sequencher 4.1.4 (Gene Codes
Corp. Ann Arbor, MI), and then aligned using
Clustal X (Thompson et al., ’97) and BioEdit 5.0.6
(Hall, ’99). All newly generated sequences were
accessioned
into
the
GenBank
database
(EU034541-EU034627). Three sequences obtained
from NCBI were used for comparison with our
cyt-b sequence data, the Estuarine crocodile
(C. porosus; AJ810453), the Siamese crocodile
(C. siamensis; DQ353946) and the Nile crocodile
(C. niloticus; AJ810452).
Microsatellites
Ten polymorphic microsatellite loci (Dever and
Densmore, 2001; FitzSimmons et al., 2001) were
also amplified for each individual (Table 2).
Primers were fluorescently labeled with WellRed
dyes (Beckman Coulter, Inc., Fullerton, CA) and
amplified using 12.5 mL PCR reactions, which
included 0.125 mL of tDNA, 9.06 mL of ddH2O,
2.50 mL buffer (0.3 M TRIS, 0.0175 M MgCl2, and
0.075 M (NH4)2SO4), 0.50 mL of 2.5 mM dNTPs
(10 mM), 0.13 mL forward primer (10 mM),
0.13 mL reverse primer (10 mM), and 0.0625 mL
(0.31 U) Promega Taq polymerase. PCR conditions
for all primers consisted of an initial denaturation
step of 2 min at 941C, then 33 cycles of 30 sec at
941C, 1 min at 581C or 621C (Table 2), and 45 sec at
721C; with a final extension of 5 min at 721C.
Fragments were sized based on a 400 bp size
J. Exp. Zool.
4
WEAVER ET AL.
TABLE 2. Microsatellite diversity values for each genetic cluster inferred using STRUCTURE (without admixed individuals)
C. rhombifer-I
Locus
C391
Cj16
Cj18
Cj20
Cj109
Cj119
Cj131
Cu5-123
Cud68
Cuj131
mean
C. rhombifer-II
C. acutus
AT (C1)1
N2
A3
HO 4
HE 5
N2
A3
HO 4
HE 5
N2
A3
HO 4
HE 5
58
62
58
62
62
58
58
58
58
58
19
22
21
22
20
21
21
22
18
20
1
4
3
4
3
4
6
3
3
2
3.3
0.00
0.82
0.57
0.64
0.80
0.43
0.91
0.32
0.67
0.05
0.52
0.00
0.65
0.64
0.66
0.63
0.50
0.66
0.35
0.48
0.05
0.46
7
7
7
7
7
7
7
7
7
7
5
3
3
5
4
3
2
5
3
4
3.7
1.00
0.57
0.57
0.71
0.29
0.43
0.86
0.71
0.86
0.43
0.64
0.78
0.58
0.65
0.73
0.50
0.69
0.53
0.81
0.67
0.71
0.66
14
14
14
14
14
14
14
14
14
14
5
7
4
4
6
3
2
5
2
2
4.0
0.57
0.86
0.64
0.50
0.50
0.50
0.14
0.50
0.36
0.29
0.49
0.68
0.78
0.60
0.56
0.82
0.62
0.25
0.48
0.52
0.35
0.57
1
Annealing temperature.
Number of individuals sampled per locus.
3
Number of average alleles per locus.
4
Observed heterozygosity.
5
Expected heterozygosity.
2
standard, using a CEQ8800 genetic analyzer and
software (Beckman Coulter, Inc., Fullerton, CA).
Using identical scoring methods, genotypes from
all C. rhombifer samples were compared with 14
pure C. acutus (see Appendix) identified by
Rodriguez (2007) and Rodriguez et al. (2008).
Data analysis
Sequence data
MODELTEST (Posada and Crandall, ’98) was
used to obtain the best-fit model of nucleotide
substitution for maximum likelihood (ML), and
MRMODELTEST (Nylander, 2004) was used to
obtain the best-fit model of evolution for Bayesian
inference (BI). Phylogenetic analyses of cyt-b
sequence data, under the ML and maximum
parsimony criteria, were performed in PAUP4.0b10
(Swofford, 2002). A starting tree was generated by
stepwise addition of taxa, with swapping performed
by utilizing the tree bisection reconnection algorithm. Node support was determined by bootstrapping topologies for 1,000 replications. Additionally, a
BI tree, with posterior probabilities, was constructed
using MrBayes v3.1.2 (Huelsenbeck and Ronquist,
2001). Four Markov chains were implemented for
1,000,000 iterations after an initial burn-in of
100,000 iterations. To obtain finer haplotype resolution, a neighbor-joining tree was constructed in
PAUP using uncorrected pairwise genetic distances
based on tRNAPro-tRNAPhe-D-loop sequence data.
J. Exp. Zool.
Node support was determined by bootstrapping the
resulting topology for 1,000 iterations.
Microsatellite data
The program POPULATIONS v1.2.28 (Langella, ’99) was employed to estimate Dc pairwise
genetic distances (Cavalli-Sforza and Edwards,
’67) to construct an exploratory neighbor-joining
tree for all C. rhombifer and C. acutus individuals,
which was visualized in TREEVIEW (Page, ’96).
The program STRUCTURE (Pritchard et al.,
2000) was used to determine assignment probabilities to specific genetic clusters by constraining
K to the number of clades suggested by the
neighbor-joining topology. We assumed that purebred individuals will have high assignment probabilities (40.97) to species genetic clusters,
whereas hybrids will have intermediate assignment probabilities (o0.97). CERVUS 3.0 (Marshall et al., ’98) was used to estimate measures of
microsatellite diversity. If hybrids were detected,
then NEWHYBRIDS (Anderson and Thompson,
2002) was used to implement a Bayesian-based
algorithm, which assigns individuals into six
genotypic classes. Genotype classes consist of two
parental groups (C. rhombifer and C. acutus), first
generation hybrids (F1), second generation hybrids (F2), F1 backcrosses to C. rhombifer and F2
backcrosses to C. acutus.
5
GENETIC CHARACTERIZATION OF THE CUBAN CROCODILE
RESULTS
Sequence-based analyses
For cyt-b sequence data, MODELTEST and
MRMODELTEST indicated the best model of
nucleotide substitution was GTR1G for both ML
and BI. The C. rhombifer species group fell within
the New World crocodilian clade that included
C. niloticus (Fig. 2). Only two haplotypes were
found using cyt-b sequence data (a and b), which
also corresponded to two haplotypes detected
using tRNAPro-tRNAPhe-D-loop sequences. Measures of percent uncorrected distances based on
cyt-b sequences (Table 3), between the C. rhombifer-a haplotype and C. acutus, and between the
C. rhombifer-b haplotype and C. acutus were both
estimated at 5.3%. Percent divergence between
the two C. rhombifer cyt-b haplotypes was
0.9%. Similarly, estimated divergences for
tRNAPro-tRNAPhe-D-loop sequences were 3.4%
between C. rhombifer-a haplotypes and C. acutus,
and 4.1% between C. rhombifer-b haplotypes and
C. acutus, respectively; whereas the genetic distance between C. rhombifer-a and C. rhombifer-b
haplotypes was estimated at 1.6% (data not
shown).
Microsatellite analyses
Fig. 2. (A) Bayesian consensus tree for cyt-b sequences,
showing values (BI/MP/ML) where 5 (1.00/100/100). (B)
Neighbor-joining tree based on tRNAPro-tRNAPhe-D-loop
sequences with bootstrap support values. Two distinct subclades were inferred using both mitochondrial sequences.
When all the samples were pooled, a matrix of
Dc distances returned three distinct clades on a
neighbor-joining tree (C. rhombifer-I, C. rhombifer-II, and C. acutus) with several individuals
clustering between clades (Fig. 3). Using the
neighbor-joining tree as a guide, K was constrained to three for model-based clustering
methods implemented in STRUCTURE; these
genetic clusters were named C. rhombifer-,
C. rhombifer-2, and C. acutus. Posterior assignment probabilities suggested that admixture was
primarily occurring between C. rhombifer-I and
C. acutus (Fig. 4), therefore, individuals with
intermediate probabilities were designated as
hybrids (Appendix). After the STRUCTURE results, NEWHYBRIDS was used to classify hybrid
types between C. rhombifer-I and C. acutus,
these individuals were mostly F2 hybrids (Fig. 5).
One individual, RC051, collected in Cancun,
TABLE 3. Uncorrected pairwise genetic distance values for cyt-b
(1) RC051 (2) MF348 (3) RC210 (4) LD175 (5) RC206 (6) AJ810452 (7) AJ810453 (8) DQ353946
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
C.
C.
C.
C.
C.
C.
C.
C.
rhombifer-a
rhombifer-b
acutus
intermedius
moreletii
niloticus
porosus
siamensis
–
0.009
0.053
0.057
0.052
0.079
0.104
0.121
–
0.053
0.055
0.052
0.077
0.107
0.121
–
0.015
0.045
0.056
0.093
0.107
–
0.046
0.059
0.093
0.107
–
0.061
0.098
0.110
–
0.114
0.123
–
0.021
–
J. Exp. Zool.
6
WEAVER ET AL.
Fig. 3. Exploratory neighbor-joining tree based on Dc distances showing microsatellite distances for C. rhombifer, C. acutus,
and possible C. acutus x rhombifer hybrids (H) constructed using POPULATIONS. A model-based analysis using STRUCTURE
confirmed the same species groupings and hybrid assignments (See Fig. 4). RC051 was intermediate between both C. rhombifer
microsatellite clusters, and MF271(?) was resolved as a possible hybrid between C. rhombifer and C. palustris.
Fig. 4. A barplot of posterior probability assignments (K constrained to 3) to species groups generated in STRUCTURE
based on microsatellite data and sorted by haplotype (see Fig. 2). C.r.-II 5 C. rhombifer genetic cluster II, C.r.-I 5 C. rhombifer
genetic cluster I, and C.a. 5 pure C. acutus genetic cluster. Inferred hybrids are designated by an H.
J. Exp. Zool.
GENETIC CHARACTERIZATION OF THE CUBAN CROCODILE
7
Fig. 5. A barplot of posterior probabilities for assignment to six genotype classes generated in NEWHYBRIDS (see text). The
plot is partitioned into inferred pure species groups or hybrid types, F1 5 first filial generation, F2 5 second filial generation,
C.a._Bx 5 backcross to C. acutus, and C.r.-I_Bx 5 backcross to C. rhombifer-I.
Mexico was morphologically identified as
C. acutus, carried a C. rhombifer-a haplotype
and exhibited evidence of admixture between
C. rhombifer-I and C. rhombifer-II. Captive specimen MF271, which also carried a C. rhombifer-a
haplotype, was resolved as a possible hybrid
between C. rhombifer and the Mugger crocodile
(C. palustris), and exhibited four unique
alleles (Cj16, 150 bp; Cj18, 207 bp; Cj131, 226 bp;
CUJ131, 193 bp) that were not found in either
C. rhombifer group or C. acutus. This individual
was removed from subsequent STRUCTURE
and NEWHYBRIDS analyses. A similar genetic
treatment of C. palustris will be needed to
accurately determine the actual paternity of this
specimen. After the removal of nine admixed
individuals, RC051 and MF271, we found that 20
alleles were C. rhombifer-a specific, 13 alleles were
C. rhombifer-b specific, and 17 alleles were specific
for C. acutus (Fig. 6).
DISCUSSION
Genetic status of C. rhombifer
Biotic homogenization due to anthropogenic
intervention can have serious evolutionary consequences on native species, such as changes in
their global distribution (Olden et al., 2004). We
have provided a genetic characterization of the
Cuban crocodile using captive specimens,
including some individuals with incongruent
morphological and mitochondrial assignments.
Phylogenetically, among New World crocodiles
three separate clades were inferred from cyt-b
sequence data (C. rhombifer-a and b, C. acutus–
C. intermedius, and C. moreletii), with C. niloticus
also grouping with the New World crocodiles
(Fig. 2). Within C. rhombifer there were
two distinct mitochondrial haplotype groups
(a and b), but it is possible that greater haplotype
diversity may be detected if a larger portion of the
mitochondrial genome is sampled. These two
haplotype groups were 0.9% divergent when
comparing cyt-b sequences and 1.6% divergent
when comparing D-loop sequences. If this pattern
remains consistent, as the Caribbean is more
extensively sampled for crocodiles, then C. rhombifer-b may represent a previously unidentified
lineage. Sampling the nuclear genome (using
microsatellites) we were able to detect both
C. rhombifer-a specific alleles and C. rhombifer-b
specific alleles, and thus two genetic clusters
(C. rhombifer-I and C. rhombifer-II) were also
inferred from model-based analyses of the microsatellite data. Taken together, these data may
actually reflect the genetic diversity within the
J. Exp. Zool.
8
WEAVER ET AL.
Fig. 6. Allele frequency distribution for three genetic clusters inferred by model-based clustering methods (see Fig. 3), after
inferred hybrids were removed.
wild population, as the C. rhombifer species clade
is consistently characterized as having two distinct
genetic sub-groups. One of these exhibits ‘‘typical’’ C. rhombifer morphology whereas the other
exhibits C. acutus morphology. We suspect that
many crocodiles that have been ‘‘morphologically’’
identified as C. acutus in Cuba may actually
belong to the C. rhombifer-b haplotype group. We
can only speculate that b haplotypes may have
been ancestrally present in Cuba, and that the
current C. rhombifer-b group could be the result of
past natural hybridization events. We cannot yet
J. Exp. Zool.
explain the evolutionary significance of C. rhombifer-b without reference samples from wild
populations in Cuba, but action should be taken
to identify these individuals and possibly remove
them from captive breeding programs.
Hybridization in crocodiles
In Vietnam, captive C. rhombifer and C. siamensis have been deliberately hybridized (Thang,
’94). Hybridization was also reported to have
occurred between captive C. rhombifer and
GENETIC CHARACTERIZATION OF THE CUBAN CROCODILE
C. acutus in breeding pens of the Laguna del
Tesoro farm in Cuba (Ross, ’98), as well as in the
wild (Ramos et al., ’94). Varona (’66) suggested
that in Cuba at least some admixture was taking
place between C. acutus and C. rhombifer, because
several specimens exhibited morphological characters typical of both species, but clearly outside
the normal range of variation found in C. acutus.
However, among our samples we found that only
MF271 and MF5267 exhibited anomalous morphology and that all F2 hybrids appeared to exhibit
C. rhombifer morphology. We found that hybridization events involving captive C. rhombifer were
invariably between the C. rhombifer-a group and
C. acutus. All hybrid individuals had C. rhombiferlike mtDNA, which suggests that in captivity
hybridization is typically between a female
C. rhombifer and a male C. acutus. These results
are consistent and congruent with those of
FitzSimmons et al. (2002).
Anthropogenic perturbation and natural migration events may pose a potential threat to the
genetic integrity of C. acutus populations in the
Caribbean, Florida, and Mesoamerica. Further,
given the large proportion of admixture between
C. acutus and C. moreletii (Cedeño-Vázquez et al.,
2008; Rodriguez et al., 2008), populations of true
crocodiles in Mexico may be threatened by an
additional source of hybrid introgression from
C. rhombifer. For example, RC051, which was
found in waterways near Cancun, Mexico, carried
a C. rhombifer-a haplotype and was morphologically identified as C. acutus; however, it was
ultimately designated as an admixture between
C. rhombifer-I and C. rhombifer-II by the nuclear
data.
Conservation and management
implications
Hybridization ultimately presents a management problem for New World crocodilians and
complicates the identification of species based on
morphology alone. A genetic evaluation (using
both mtDNA and nDNA) in conjunction with a
morphometric characterization can provide a
more accurate view of an individual’s ancestry
than either method alone. As an example, a
crocodile marked for reintroduction into Cat Tien
National Park (Vietnam) was genetically identified as a C. rhombifer siamensis hybrid and was
subsequently prevented from being released into
the wild population (FitzSimmons et al., 2002). In
order for reintroductions of native species to be
9
successful, only purebred individuals should be
released back into their native habitats (Allendorf
et al., 2001). Our work has provided an initial
genetic assessment of the critically endangered
Cuban crocodile. We hope these data can be used
to identify pure individuals for breeding stock,
which should be considered if repopulation of
extirpated areas (e.g. Cienaga de Lanier) is to take
place.
An evaluation of current Cuban crocodile stocks
is warranted to ensure a purebred captive breeding line, especially considering that out of seven
US captive stocks surveyed five had some level of
genetic admixture. Identification of hybrids using
morphology may be problematic given that only
two individuals studied were morphologically
anomalous. Additionally, thorough genetic assessments of wild Cuban and American crocodile
populations in the Caribbean are needed to
provide a better genetic reference for assignment
tests, to quantify the amount of potential genetic
admixture between genetically differentiated
groups, and to help clarify the evolutionary
implications of the C. rhombifer-b sub-clade. The
increase in ease and the concomitant decrease in
cost of generating and analyzing molecular genetic
data with statistical model-based analyses can be
practical and informative for both management
and conservation efforts. These methods are
especially important for the conservation of
endangered fauna with limited distributions, such
as the Cuban crocodile.
ACKNOWLEDGMENTS
We thank R. Bradley and J. Hanson for
assistance with sequencing, the TTU Core Lab
for assistance with genotyping, and O. Sanjur and
S. Mahecha at the Smithsonian Tropical Research
Institute for laboratory assistance. We thank S. K.
Davis and T. Guerra. In addition, we thank J.
McVay for his input on data analyses, S. McCracken for curatorial assistance, M. A. Mullen, D.
Hamilton, M. Vandewege, and P. Larsen for
editorial assistance; and J. Isom, D. Fabing, and
L. Durham for administrative assistance. This
research was supported in part by a Howard
Hughes Medical Institute grant through the
Undergraduate Science Education Program to
Texas Tech University. This research was
developed under the following permits: SIM/A4-07,
2007/KY/000107,
MX31441
(NRA:
CFSTB2300411) and PRT 2-2996.
J. Exp. Zool.
10
WEAVER ET AL.
APPENDIX
Samples used in this study, source of samples and group assignments for each data set.
Sample #
Morphology
Source
MtDNA1
LD042
LD068
LD069
LD099
LD100
LD123
LD125
LD145
LD147
LD148
LD150
LD154
LD178
LD179
LD182
LD183
LD197
LD198
LD221
LD305
LD324
MF267
MF268
MF269
MF270
MF271
MF272
MF274
MF277
MF1744
MF1745
MF5267
RC051
GC001
MF347
MF348
PM008
PM031
PM037
SP014
PM020
PM021
RC013
RC052
RC106
RC109
RC114
RC117
RC130
RC132
RC136
RC140
RC141
RC210
RC023
LD175
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
Anomalous
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
C. rhombifer
Anomalous
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. acutus
C. intermedius
St. Augustine Zoo
St. Augustine Zoo
St. Augustine Zoo
St. Augustine Zoo
St. Augustine Zoo
Gladys Porter Zoo
Private property
Private property
Private property
St. Augustine Zoo
Jumbo Lair
Jumbo Lair
St. Augustine Zoo
St. Augustine Zoo
Toledo Zoo
Toledo Zoo
Bronx Zoo
Bronx Zoo
Bronx Zoo
Private property
Bronx Zoo
St. Augustine Zoo
St. Augustine Zoo
Bronx Zoo
Bronx Zoo
Bronx Zoo
Toledo Zoo
Toledo Zoo
Busch Gardens
Gladys Porter Zoo
Gladys Porter Zoo
Jupiter, FL
Cancun, Mexico
Grand Cayman Island
Jamaica
Jamaica
Busch Gardens
Imperial River, FL
Snapper Creek Canal, FL
Private property
North Key Largo, FL
North Key Largo, FL
Yucatan, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Quintana Roo, Mexico
Yucatan, Mexico
Quintana Roo, Mexico
Private property
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
b
b
b
b
b
b
b
–
–
–
–
–
–
–
–
–
–
–
–
–
C. acutus
C. acutus
C. intermedius
J. Exp. Zool.
MtDNA2
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
b
b
b
b
b
b
b
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
C.
–
acutus5
acutus5
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
acutus6
Msat3
Msat4
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
C.r.-1
Hyb
C.r.-1
C.r.-1
Hyb
Hyb
Hyb
Hyb
Hyb
C.r.-1
C.r.-1
Hyb
Hyb
C.r.-2
C.r.-2
C.r.-2
C.r.-2
C.r.-2
C.r.-2
C.r.-2
C.r.-2
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
–
–
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
Hyb
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
C.r.-I
Hyb
C.r.-I
C.r.-I
Hyb
C.r.-I
C.r.-I
Hyb
Hyb
–
Hyb
Hyb
C.r.-I
C.r.-I
Hyb
Hyb
C.r.-II
C.r.-II
C.r.-II
C.r.-II
C.r.-II
C.r.-II
C.r.-II
C.r.-II
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
C.a.
–
–
11
GENETIC CHARACTERIZATION OF THE CUBAN CROCODILE
Sample #
RC206
RC207
Morphology
C. moreletii
C. moreletii
Source
Quintana Roo, Mexico
Yucatan, Mexico
MtDNA1
C. moreletii
C. moreletii
MtDNA2
C. moreletii
C. moreletii
Msat3
–
–
Msat4
–
–
Wild caught individuals.
1
Based cyt-b sequence data (see Fig. 2A).
Based on tRNAPro- tRNAPhe-Dloop sequence data (see Fig. 2B).
3
Distance-based analysis of microsatellite data (see Fig. 3).
4
Model-based analysis of microsatellite data (see Fig. 4).
5
Rodriguez (2007).
6
Cedeño-Vázquez et al. (2008).
2
LITERATURE CITED
Allendorf FW, Leary RF, Spruell P, Wenburg JK. 2001. The
problems with hybrids: setting conservation guidelines.
Trends Ecol Evol 16:613–622.
Anderson EC, Thompson EA. 2002. A model-based method for
identifying species hybrids using multilocus genetic data.
Genetics 160:1217–1229.
Bayliss P. 1987. Survey methods and monitoring within
crocodile management programmes. In: Webb GJ, Manolis
SC, Whithead PJ, editors. Wildlife management: crocodiles
and alligators. Sydney, Australia: Surrey Beatty and Sons.
p 157–175.
Brochu CA. 2000. Phylogenetic relationships and divergence
timing of Crocodylus based on morphology and the fossil
record. Copeia 2000:657–673.
Cavalli-Sforza LL, Edwards AWF. 1967. Phylogenetic analysis: models and estimation procedures. Am J Hum Genet
19:233–257.
Cedeño-Vázquez JR, Rodriguez D, Calmé S, Ross JP,
Densmore III LD, Thorbjarnarson JB. 2008. Hybridization
between Crocodylus acutus and Crocodylus moreletii in the
Yucatan Peninsula: I. Evidence from mitochondrial DNA
and Morphology. J Exp Zool 309A.
Chavananikul V, Wattanodorn S, Youngprapakorn P. 1994.
Karyotypes of 5 species of crocodile kept in Samutprakan
crocodile farm and zoo. In: Crocodiles, Proceedings of the
12th Working Meeting of the Crocodile Specialist Group,
IUCN—The World Conservation Union, Gland, Switzerland. p 58–62.
Cohen MM, Gans C. 1970. The chromosomes of the order
crocodilia. Cytogenetics 9:81–105.
Densmore LD. 1983. Biochemical and immunological systematics of the order crocodilia. In: Hecht MK, Wallace B,
Prance GT, editors. Evolutionary biology. Vol. 16. New
York: Plenum Publ Co. p 397–465.
Dever JA, Densmore III LD 2001. Microsatellites in Morelet’s
crocodile (Crocodylus moreletii) and their utility in addressing crocodilian population genetics questions. J Herpetol
35:541–544.
FitzSimmons NN, Tanksley S, Forstner MRJ, Louis EE,
Daglish R, Gratten J, Davis S. 2001. Microsatellite markers
for Crocodylus: new genetic tools for population genetics,
mating system studies and forensics. In: Grigg GC,
Seebacher F, Franklin CE, editors. Crocodilian biology
and evolution. Chipping Norton: Surrey Beatty & Sons.
FitzSimmons NN, Buchan JC, Lam PV, Polet G, Hung TT,
Thang NQ, Gratten J. 2002. Identification of purebred
Crocodylus siamensis for reintroduction in Vietnam. J Exp
Zool 294:373–381.
Frankham R, Ballou JD, Briscoe DA. 2002. Introduction to
conservation genetics. Cambridge: Cambridge University
Press.
Franz R, Morgan GS, Albury N, Buckner SD. 1995.
Fossil skeleton of a cuban crocodile (Crocodylus rhombifer)
from a blue hole on Abaco Bahamas. Caribbean J Sci
31:149–152.
Glenn TC, Staton JL, Vu AT, Davis LM, Bremer JRA,
Rhodes WE, Brisbin Jr. IL, Sawyer RH. 2002. Low
mitochondrial DNA variation among American alligators
and a novel non-coding region in crocodilians. J Exp Zool
294:312–324.
Gorzula S, Arocha-Pinango CL, Salazar C. 1976. A method of
obtaining blood by vein puncture from large reptiles.
Copeia 1976:838–839.
Hall TA. 1999. BioEdit: a user-friendly biological sequence
alignment editor and analysis program for Windows 95/98/
NT. Nucleic Acids Symp Ser 41:95–98.
Hekkala ER. 2004. Conservation genetics at the species
boundary: case studies from African and Caribbean
crocodiles (Genus: Crocodylus). PhD Dissertation. Columbia University, New York, NY.
Huelsenbeck JP, Ronquist FR. 2001. MRBAYES: Bayesian
inference of phylogenetic trees. Bioinformatics 17:754–755.
Langella. 1999. Populations v1.2.28: Available at http://
www.cnrs-gif.fr/pge/bioinfo/populations.
Marshall TC, Slate J, Kruuk LEB, Pemberton JM. 1998.
Statistical confidence for likelihood-based paternity inference in natural populations. Mol Ecol 7:639–655.
Morgan GS, Richard F, Crombie RI. 1993. The Cuban
crocodile, Crocodylus rhombifer, from late quaternary
fossil deposits on Grand Cayman. Caribbean J Sci
29:153–156.
Nylander JAA. 2004. MrModeltest 2.1: Program distributed by
the author. Evolutionary Biology Centre, Uppsala.
Olden JD, Poff NL, Douglas MR, Douglas ME, Fausch KD.
2004. Ecological and evolutionary consequences of biotic
homogenization. Trends Ecol Evol 19:18–24.
Page RDM. 1996. TreeView: an application to display
phylogenetic trees on personal computers. Comput Appl
Biosci 12:357–358.
Posada D, Crandall KA. 1998. MODELTEST: testing
the model of DNA substitution. Bioinformatics 14:
817–818.
Pritchard JK, Stephens M, Donnelly P. 2000. Inference of
population structure using multilocus genotype data.
Genetics 155:945–959.
Ramos R, de Buffrenil V, Ross JP. 1994. Current status of the
Cuban crocodile, Crocodylus rhombifer, in the wild. In:
Crocodiles, Proceedings of the 12th Working Meeting of the
J. Exp. Zool.
12
WEAVER ET AL.
Crocodile Specialist Group, IUCN—The World Conservation Union, Gland, Switzerland. p 113–140.
Ray DA, Densmore L. 2002. The crocodilian mitochondrial
control region: general structure, conserved sequences, and
evolutionary implications. J Exp Zool 294:334–345.
Ray DA, Dever JA, Platt SG, Rainwater TR, Finger AG,
McMurry ST, Batzer MA, Barr B, Stafford PJ,
McKnight J, Densmore LD. 2004. Low levels of
nucleotide diversity in Crocodylus moreletii and evidence
of hybridization with C. acutus. Conserv Genet 5:
449–462.
Rodriguez D. 2007. Crocodilian evolution, systematics and
population genetics: recovery and ecological interactions of
the American crocodile (Crocodylus acutus). PhD Dissertation. Texas Tech University, Lubbock, Texas.
Rodriguez D, Cedeño-Vázquez JR, Forstner MRJ, Densmore
III LD. 2008. Hybridization between Crocodylus acutus and
Crocodylus moreletii in the Yucatan Peninsula: II. Evidence from microsatellites. J Exp Zool 309A.
Ross JP. 1998. Crocodiles status survey and conservation
action plan. Gainesville, FL: Crocodile Specialist Group.
96p.
J. Exp. Zool.
Swofford DL. 2002. PAUP v4.0b10: phylogenetic analysis
using parsimony ( and other methods). Version 4.0b10.
Sunderland, MA: Sinauer.
Thang NQ. 1994. The status of Crocodylus rhombifer in the
Socialist Republic of Vietnam. In: Crocodiles, Proceedings
of the 12th Working Meeting of the Crocodile Specialist
Group, IUCN—World Conservation Union. Gland, Switzerland. p 141–151.
Thompson J, Gibson T, Plewniak F, Jeanmougin F, Higgins D.
1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Res 25:4876–4882.
Thorbjarnarson J, Mazzotti F, Sanderson E, Buitrago F,
Lazcano M, Minkowski K, Muniz M, Ponce P, Sigler L,
Soberon R, Trelancia AM, Velasco A. 2006. Regional
habitat conservation priorities for the American crocodile.
Biol Conserv 128:25–36.
Varona LS. 1966. Notas sobre los crocodı́lidos de Cuba y
descripción de una nueva especie del Pleistoceno. Poeyana
16:1–21.
Varona LS. 1984. Los cocodrilos fosiles De Cuba (reptilia:
crocodylidae). Caribbean J Sci 20:13–18.