New dates and new rates for divergence
across the Isthmus of Panama
Nancy Knowlton
1*
and Lee A. Weigt
2
{
1
SmithsonianTropical Research Institute, Apartado 2072, Balboa, Republic of Panama
2
Field Museum of Natural History, Roosevelt Road at Lake Shore Drive, Chicago, IL 60605, USA
Sister species separated by the Isthmus of Panama have been widely used to estimate rates of molecular
evolution. These estimates are based on the assumption that geographic isolation occurred nearly simulta-
neously for most taxa, when connections between the Caribbean and eastern Paci?c closed
approximately three million years ago. Here we show that this assumption is invalid for the only genus
for which many taxa and multiple genetic markers have been analysed. Patterns of divergence exhibited
by allozymes and the mitochondrial COI gene are highly concordant for 15 pairs of snapping shrimp in
the genus Alpheus, indicating that they provide a reasonable basis for estimating time since cessation of
gene ?ow. The extent of genetic divergence between pairs of sister species varied over fourfold. Sister
species from mangrove environments showed the least divergence, as would be expected if these were
among the last habitats to be divided. Using this pair yields a rate of sequence divergence of 1.4% per
one million years, with implied times of separation for the 15 pairs of 3^18 million years ago. Many past
studies may have overestimated rates of molecular evolution because they sampled pairs that were sepa-
rated well before ?nal closure of the Isthmus.
Keywords: allozymes; Alpheus; COI; molecular clock; mtDNA; Panama
1. INTRODUCTION
The Isthmus of Panama represents a complete, relatively
recent, and well-dated barrier across what was once a
large neotropical marine environment (Farrell et al. 1995;
Coates & Obando 1996). Sister species separated by the
Isthmus have thus provided an important tool for
estimating rates and patterns of molecular evolution
for many marine groups (Lessios 1979, 1998; Vawter et al.
1980; Martin et al. 1992; Knowlton et al. 1993;
Bermingham & Lessios 1993; Collins 1996; Bermingham
et al. 1997; Sturmbauer et al. 1996; Hart et al. 1997;
Schubart et al. 1998).
Studies to test the constancy and estimate the rate of
molecular evolution are based on the assumption that
most transisthmian sister-species pairs were separated at
roughly the same time by ?nal closure of the connection
between the Caribbean and the eastern Paci?c (Collins
1996), approximately three million years ago (Ma)
(Coates & Obando 1996). Recent molecular and fossil
studies suggest that this assumption may be invalid
(Knowlton et al. 1993; Jackson et al. 1993), but the signi?-
cance of these concerns remains unclear (Cunningham &
Collins 1994). The continuing use of the Isthmus as a
basis for estimating rates of molecular evolution (see, for
example, Burton & Lee 1994; Sturmbauer et al. 1996;
Hart et al. 1997; Chenoweth et al. 1998; Metz et al. 1998;
Schubart et al. 1998) makes resolution of the timing of
transisthmian divergences an important task.
Here we report allozyme and mtDNA divergences for
15 transisthmian sister-species pairs in the snapping
shrimp genus Alpheus. All are shallow-water forms, but
two pairs are restricted to mangroves, which were likely
to be the last marine habitats separated by the rising
Isthmus. This transisthmian molecular data set is unique,
in that no other genus analysed to date contains more
than two sister-species pairs (Bermingham et al. 1997;
Schubart et al. 1998), and with few exceptions (Knowlton
et al. 1993; Bermingham & Lessios 1993; Metz et al. 1998),
independent measures of genetic divergence (such as
those based on mitochondrial and nuclear genes or gene
products) are not available. Multiple taxa and indepen-
dent measures of genetic divergence are essential, because
a single gene might misrepresent the species tree, and a
single species might misrepresent the general biogeo-
graphic history of a region.
2. MATERIALS AND METHODS
(a) Taxa
The taxonomic literature suggests approximately 20 possible
transisthmian sister-species pairs in the genus Alpheus (Kim &
Abele 1988). In some cases these are currently recognized as
distinct at the species level, and in other cases not. This inconsis-
tency is a taxonomic artefact; all should be recognized as
distinct species because of ?xed genetic di?erences and repro-
ductive incompatibilities (Knowlton et al. 1993; this paper).
Shrimp were initially identi?ed by using the keys published
by Chace (1972) and Kim & Abele (1988). In the course of our
Proc. R. Soc. Lond. B (1998) 265, 2257^2263 2257 & 1998 US Government
Received 29 June 1998 Accepted 28 August 1998
*
Author for correspondence (knowlton@naos.si.edu).
{Present address: CMMID, Virginia Tech, 1410 Prices Fork Road,
Blacksburg,VA 24061-0342, USA.
collections we discovered several additional, undescribed
species that resembled taxa with close transisthmian relatives;
these sibling species were also included (see Knowlton & Mills
1992; Knowlton et al. 1993). All shrimp were collected from the
Caribbean and Paci?c coasts of Panama, typically from the
intertidal zone and in no case from depths of more than 20m.
In total we collected over 2000 individuals from a variety of
habitats.
Correct identi?cation of transisthmian sister species is
essential for estimating rates of molecular evolution, and can be
a problem in highly diverse groups like Alpheus (Lessios 1998).
Many transisthmian sister-species pairs closely resemble each
other in morphology and colour pattern, and do not resemble
other Alpheus from the eastern Paci?c or western Atlantic. This
allows one to assign sister-species pairs reliably without resorting
to molecular data. In several cases two sibling species occur in
2258 N. Knowlton and L. A.Weigt Molecular divergence across the Isthmus of Panama
Proc. R. Soc. Lond. B (1998)
Figure 1. COI-based maximum
likelihood approximation
(Puzzle 4.0) (Strimmer et al. 1997)
phylogeny of Alpheus taxa with
possible transisthmian sister species.
Morphological (see Knowlton et al.
1993) and molecular assignments
of sister species were entirely
consistent. Likelihood approxima-
tion reliability estimates are shown
in bold to the left of the 15 trans-
isthmian pairs. Branch lengths are
shown above each Caribbean (C)
and Paci?c (P) pair member.
one or both oceans, but in these cases sister species are also
readily identi?ed by means of behavioural tests or subtle
di?erences in morphology and colour (Knowlton et al. 1993).
However, we excluded potential transisthmian pairs involving
the Caribbean A. armillatus and A. heterochaelis complexes
because we were not con?dent that sister species could be
assigned or collected owing to the large number of sibling
species in these groups.
(b) Analyses
After identi?cation, living shrimp were frozen in liquid N
2
,
or placed directly in a 780 ?C ultracold freezer for storage
before processing. For all pairs of taxa, we characterized at least
11 and typically 15^16 allozyme loci by using conventional
starch gel electrophoresis, and sequenced 564 base pairs (bp) of
the mtDNA cytochrome oxidase I (COI) gene (GenBank
accession numbers U02002^U02018, AF097858^AF097873)
(see Knowlton et al. (1993) for details). We used the same
methods as in our earlier analyses, except that cycle-sequencing
reactions were performed with a dRhodamine kit (PE-ABI) or
a Thermosequenase kit (Amersham) at 10 ml volume (modi?ed
to include halfTERM (Genpak)), following manufacturers'
instructions, and run on an ABI377 automated DNA sequencer.
Sequences were trimmed and aligned by means of Sequencher
(v3.0, Genecodes), and analysed with the Puzzle 4.0 (Strimmer
et al. 1997) maximum likelihood approximation.
3. RESULTS
As before (Knowlton et al. 1993), transisthmian sister
species identi?ed by traditional morphological criteria
(Chace 1972; Kim & Abele 1988; Knowlton & Mills
1992) were con?rmed by phylogenies based on the mito-
chondrial COI gene (?gure 1). Doubling the sample size
from seven to 15 pairs strengthened our earlier conclusion
that both measures of genetic divergence (allozymes and
mtDNA) varied widely across the pairs of taxa in a
concordant fashion (?gure 2).
The most parsimonious interpretation of these results is
that time of isolation varied widely among these pairs of
Molecular divergence across the Isthmus of Panama N. Knowlton and L. A.Weigt 2259
Proc. R. Soc. Lond. B (1998)
Figure 2. Relation between corrected per cent sequence divergence and allozyme di?erentiation (Nei's D) for 15 independent
pairs of transisthmian sister species (?gure 1). Maximum likelihood divergence estimates were corrected for among-site rate
variation by means of a four-category gamma distribution. Just below the top axis are approximate dates of divergence based
on COI data, with the assumption that the most similar pair diverged at approximately 3 Ma. Various biological (Collins 1989;
Cunningham & Collins 1994) and geological (table 5 of Farrell et al. (1995)) events related to the rise of the Isthmus, whose
timing has been independently estimated based on the geological record, are positioned above the top axis to show their
relationship to the timing of shrimp divergences based on molecular rate calibrations. The two most distant pairs appear to
be the survivors of lineages that diverged about the time the Tethyan connection between Atlantic and Paci?c was severed.
taxa. The magnitude of variability in mtDNA divergence
between pairs is inconsistent with simultaneous isolation
and clock-like divergence based on an analysis of the
nucleotide changes shown in ?gure 1: the statistic R(n71)
(where R is the mean:variance of number of nucleotide
changes, number of changes equals branch lengths multi-
plied by number of base pairs, and n is the number of
lineages) (Goldman 1994) is signi?cantly non-random for
all sites and third positions only, both for all taxa and
when the two most dissimilar pairs (?gure 2) are
excluded (p50.001). Concordance between the two inde-
pendent measures of divergence (?gure 2) also points
strongly to non-simultaneous isolation as the most parsi-
monious explanation for this pattern (Knowlton et al.
1993; Bermingham & Lessios 1993; Collins 1996).
Non-simultaneous divergence is also supported by the
fact that species pairs restricted to o?shore islands or
deeper habitats are typically more divergent than those
found along the mainland (Knowlton et al. 1993; this
paper). Moreover, the smallest divergence values were
seen for shrimp pairs from mangroves (?gure 2), as
expected if these were the last habitats separated by the
rising Isthmus.
If the 15 transisthmian sister species did not diverge
simultaneously, then the lowest mtDNA divergence value,
rather than the average divergence of all shrimp pairs,
provides a better estimate of the rate of molecular evolu-
tion. This rate can then be used to reconstruct the timing
of isolation of more divergent pairs (?gure 2). The use of
a geological estimate of ?nal closure of the connection
between the Caribbean and eastern Paci?c approximately
3Ma (Coates & Obando 1996), together with a gamma
correction for COI divergences, yielded estimated dates of
divergence ranging from 3Ma (most similar mangrove
pair) to ca. 18Ma (two most divergent pairs) (?gure 2).
Divergences between most transisthmian pairs fell
within a 3^9-million-year range. This interval is
bracketed at one end by the ?rst crossing of mammals
between North and South America, and at the other by
the onset of glaciation and massive mammal interchange
(?gure 2), lending credence to the reliability of the rate
calibrations. Similarly, a pulse of divergences about
4.5 Ma coincides with several oceanographic measures of
basin isolation (events summarized by Farrell et al.
(1995)). The two most highly divergent pairs fall outside
this interval, however, and presumably represent a pre-
Isthmian event (see below).
4. DISCUSSION
(a) Caveats
Comparison of true sister species is fundamental to our
conclusion that most transisthmian sister taxa are consid-
erably older than three million years, because aberrantly
high divergence values could be caused by misassigning
sister-species pairs or failing to collect them. The concor-
dance of morphological, behavioural and molecular data
sets (Knowlton et al. 1993; this paper) suggests that closest
relatives among the taxa we collected have been identi?ed
correctly. Kim & Abele's (1988) compilation indicates no
potential transisthmian sister species involving eastern
Paci?c shrimp with ranges that do not include Panama,
and based on our collections their monograph is very
complete. Although monographs for the Caribbean are
less comprehensive forAlpheus, our own Caribbean collec-
tions outside Panama (albeit limited) revealed diver-
gences between presumed conspeci?cs of less than 1.5%
(Cunningham & Collins 1994; Knowlton & Weigt 1997;
N. Knowlton and L. A. Weigt, unpublished data). Thus
failure to compare closest transisthmian relatives does not
seem a likely explanation for the overall pattern we
observed.
A second potential source of error is our assumption
that connections between the Caribbean and eastern
Paci?c ended about 3Ma. A few authors have suggested
leakage or a ?nal breach of the Isthmus as recently as
2Ma (Cronin & Dowsett 1996). If we assign a date of
2Ma to the divergence of the most similar mangrove
species, then divergence times for the remaining pairs by
extrapolation would still range from 2^14Ma, with over
half of all pairs having divergences of more than four
million years. However, empirical support for a bio-
logically important breach of the Isthmus this recently is
thin. Indeed, Cronin & Dowsett (1996) suggest that the
barrier began to have substantial e?ects on surface water
?ow by 3.5Ma, but with a reopening of previously closed
connections between 3.1 and 2.8Ma due to especially
high global sea level. This scenario would also match the
temporal pattern of divergences suggested by the mol-
ecular data without having to invoke a 2Ma breach of
the Isthmus.
Finally, our analyses could be confounded by rate
inconstancy, particularly if all genetically similar shrimp
pairs belonged to clades that were on average slowly evol-
ving, and all genetically divergent shrimp pairs belonged
to rapidly evolving clades. This does not seem to be true,
however.We could detect no signi?cant deviations in rates
based on pairwise comparisons (Wu & Li 1985) once
Bonferroni corrections for the number of simultaneous
comparisons were applied (smallest p50.001; that
required for signi?cance with multiple tests is p50.0002).
Morrison (1997) also found that Alpheus (but not Synal-
pheus) exhibited rate constancy in a study of transisthmian
taxa. Thus, although we cannot rule out minor di?er-
ences in rates of molecular evolution within the genus, all
available data suggest that rate variation is not a major
contributor to the general pattern observed in Alpheus.
(b) Implications and generality
There are several processes that could result in non-
simultaneous divergence times between transisthmian
sister taxa (Lessios 1998). The most obvious of these is
variation in the timing of severance of gene ?ow across
the Isthmian region. Although such variation could be
entirely stochastic, ecological di?erences between the
least and most divergent pairs suggest that at least some
of the variation is ecologically based. Various aspects of
larval behaviour and physiology might contribute to such
a pattern. Grosberg (1982), for example, documented that
larval depth strati?cation mirrored that of adult popula-
tions on a very ?ne scale in barnacles. A comparable
situation in snapping shrimp could lead to di?erences in
the timing of the cessation of gene ?ow across the Isthmus
if shallow-water larvae were less likely to be blocked by
reduced water ?ow than deeper-dwelling larvae. Larvae
from adults restricted to o?shore reef environments might
2260 N. Knowlton and L. A.Weigt Molecular divergence across the Isthmus of Panama
Proc. R. Soc. Lond. B (1998)
also be more likely to avoid bodies of turbid, lower-salinity
water over the rising Isthmus. Future experiments with
larvae of extant transisthmian species having di?erent
divergence values could be informative in this regard.
Extinction can also have major e?ects on biogeographic
patterns (Vermeij 1991; Cunningham & Collins 1998),
including those associated with the Isthmus of Panama
(Cunningham & Collins 1994; Lessios 1998). Consider,
for example, the extant quartet of sibling species
belonging to the A. paracrinitus
^
A. rostratus complex
(?gure 1). This quartet consists of two clades, each with a
transisthmian pair, but they are so similar morphologi-
cally that they were only recently distinguished (Kim &
Abele 1988). If the Caribbean member of one clade and
the Paci?c member of the other clade had gone extinct, it
would lead to an apparent pair of much more divergent
transisthmian sister species. Such di?erential extinction
patterns are not unexpected, considering the oceano-
graphic di?erences between the Caribbean and the
eastern Paci?c. Interestingly, the amount of divergence
between these two clades (data in Knowlton et al. (1993))
is very similar to that observed for the two most divergent
extant pairs in this study. The cause of three independent
divergences at about 18Ma (?gure 2) is unclear, but may
be related to oceanographic changes associated with the
closure of theTethyan seaway at about that time.
No evidence for very recent speciation (less than 2Ma)
within either ocean is provided by these data; there are
no sympatric sibling-species pairs with divergences less
than those exhibited by the mangrove transisthmian taxa.
This is similar to the pattern observed for Caribbean
benthic foraminifera, which show comparatively few
originations after 3.5Ma (L. S. Collins et al. 1996).
However, radiations in the past two million years have
been suggested for several other groups, especially in the
Paci?c, based on molecular (see, for example, McMillan
& Palumbi 1995; Palumbi 1996) and fossil (Jackson et al.
1996) data.
Our principal ?nding, that presumed transisthmian
pairs are often older than three million years, is likely to
be true for many marine groups. Evidence for this comes
from both other molecular analyses and the fossil record.
Transisthmian divergence times for three genera of snails
range from 5.3^8.5Ma, based on an independent (non-
Isthmian) calibration by means of the fossil record of a
fourth genus (Collins 1989; T. M. Collins et al. 1996);
these dates are entirely consistent with the pattern for
Alpheus (?gure 2). Four genera of echinoderms yield COI
divergences of 4.7%, 9.8%, 12% and 17.6% (Hart et al.
1997; Metz et al. 1998; Lessios 1998); this span of diver-
gences is remarkably similar to that observed for Alpheus
(4^19% with a two-parameter Kimura correction),
although the taxonomic range, gene regions, and correc-
tions for these studies are not identical. Data for ?sh show
even greater variability, with COI transisthmian diver-
gences ranging over 60-fold (from 0.18^12.4%) across 17
genera (Bermingham et al. 1997). If we turn to the fossil
record, most neotropical lineages of strombinid gastropod
were restricted to either the Caribbean or eastern Paci?c
basins by 5Ma (Jackson et al. 1993). L. S. Collins et al.
(1996) argue that palaeoceanographic circulation models,
isotopic evidence and foraminiferan originations all point
to divergence of the basins beginning as early as 7^8Ma.
These results have important implications for using the
Isthmus to c`alibrate' molecular c`locks'. Worries about the
noisiness of molecular data have encouraged authors to
focus on average values, particularly when they are
consistent with past estimates. For example, two ?sh
genera (Abudefduf and Anisotremus) each have two trans-
isthmian pairs: in both genera, one pair has a divergence
of 1.5% and the other 4.5% (Bermingham et al. 1997). In
the context of the wide variation exhibited across all ?sh
taxa and rate estimates from other studies, the 4.5%
divergences were chosen as the best overall value for cali-
brations (Bermingham et al. 1997). It seems more likely,
however, that (as with shrimp) the smaller divergences
give better estimates based on the date of ?nal closure,
particularly when they have phylogenetic support (as is
the case for Abudefduf, where only the pair with 1.5%
divergence are sister species). This would imply a rate of
change for this portion of the COI gene almost three
times faster for shrimp than for ?sh (1.4% and 0.5% per
million years, respectively).
In general, the shrimp data suggest that it is easy to
overestimate rates of molecular evolution by failing to
sample the appropriate transisthmian pairs. Taxa
restricted to o?shore, reef environments are likely to
provide especially poor estimates of rates of molecular
evolution, because divergence is likely to have preceded
the closure of the Isthmus by an unknown amount of
time. The best taxa for such estimates are mangrove
associates, species routinely found in turbid, inshore
waters, or high intertidal and quasiterrestrial marine
organisms (see, for example, Sturmbauer et al. 1996;
Schubart et al. 1998). COI divergence values for the trans-
isthmian estuarine crabs analysed by Schubart et al.
(1988; 0.04, 0.06) were remarkably similar to those
obtained for two estuarine Alpheus species (?gure 2).
Similarly, in the two ?sh genera that each have two trans-
isthmian pairs, the lower divergences within each genus
were found for pairs typically found on more inshore
reefs (Allen & Robertson 1994). Even ecologically
promising transisthmian taxa may individually yield
unexpectedly high divergence values, however; although
the mangrove-associated snails analysed by Collins
(1989) had the lowest divergence values of the three pairs
analysed, they still had values that suggested isolation
before the close of the Isthmus based on rate calibrations
for Nucella (Cunningham & Collins 1994).
Good estimates are important because temporal infor-
mation allows one to associate cladogenesis with the
physical history of the Earth and estimate the speed with
which major evolutionary transitions take place. Our
data suggest that some evolutionary transitions are likely
to be considerably slower than several past transisthmian-
based estimates would suggest. For example, Hart et al.
(1997) used a transisthmian calibration for Oreaster
(17.6%, three million years) to suggest that two major
developmental changes in star?sh may have occurred
within the past two million years. However, as Hart et al.
(1997) noted, if the calibration is incorrect (perhaps by
over threefold, based on other echinoderm data), then
these rapid evolutionary transitions are in fact consider-
ably slower. Similarly, dates estimated for reconstructions
of the biogeographic history of Central American fresh-
water ?sh (Bermingham et al. 1997) are too young if the
Molecular divergence across the Isthmus of Panama N. Knowlton and L. A.Weigt 2261
Proc. R. Soc. Lond. B (1998)
true rate of sequence divergence is 0.5% rather than
1.2%, the latter ?gure being derived from pairs that may
have diverged well before closure of the Isthmus.
Finally, studies of transisthmian taxa support the idea
that isolation in the sea can often occur without
impermeable physical barriers (Palumbi 1994; Miya &
Nishida 1997; Hellberg 1998). Where partial barriers do
exist, biological responses to them are likely to be
complex and drawn out in time, on land as well as sea,
leading to phylogeographic patterns termed `pseudocon-
gruence' (Collins & Cunningham 1994). For example,
divergence dates for North American birds traditionally
attributed to the Late Pleistocene appear to range from
200 000 to over ?ve million years based on mtDNA
sequences (Klicka & Zink 1997). Similar ?ndings of non-
simultaneous divergences have also been reported across
barriers in Australian and Brazilian rain forests (Joseph
et al. 1995; Patton & da Silva 1998) and for terrestrial
taxa separated by the opening of the Strait of Gibraltar
(Busack 1986). Thus lessons learned from the Isthmus
have broad applicability.
We thank Javier Jara and Adam Gerstein for assistance in the
?eld, and Eyda Gomez for assistance in the laboratory. Steve
Palumbi, Andrew Martin, Jeremy Jackson, Haris Lessios and an
anonymous reviewer made helpful comments on earlier versions
of this manuscript. Weibe Kooistra and Andrew Martin helped
with analysis of rate constancy. The government of Panama
(Recursos Marinos) and the Kuna Nation gave permission for
?eldwork and collections, and the Smithsonian Institution and
Field Museum provided ?nancial support. This work is based in
part upon data collected on a sequencer purchased with support
from the US National Science Foundation (BIR-9419732)
located in the Field Museum's Biochemistry Laboratories, reno-
vated with support from the NSF (STI-921446) and operated
with support from an endowment by the Pritzker Foundation.
REFERENCES
Allen, G. R. & Robertson, D. R. 1994 Fishes of the tropical eastern
Paci?c. Bathhurst, UK: Crawford House Press.
Bermingham, E. & Lessios, H. A. 1993 Rate variation of
protein and mitochondrial DNA evolution as revealed by sea
urchins separated by the Isthmus of Panama. Proc. Natn. Acad.
Sci. USA 90, 2734^2738.
Bermingham, E., McCa?erty, S. S. & Martin, A. P. 1997 Fish
biogeography and molecular clocks: perspectives from the
Panamanian Isthmus. In Molecular systematics of ?shes (ed. T. D.
Kocher & C. A. Stepien), pp. 113^128. San Diego: Academic
Press.
Burton, R. S. & Lee, B.-N. 1994 Nuclear and mitochondrial
gene genealogies and allozyme polymorphism across a major
phylogeographic break in the copepod Tigriopus californicus.
Proc. Natn. Acad. Sci. USA 91, 5197^5201.
Busack, S. D. 1986 Biogeographic analysis of the herpetofauna
separated by the formation of the Strait of Gibraltar. Natl
Geogr. Res. 2, 17^36.
Chace, F. A. Jr 1972 The shrimps of the Smithsonian-Bredin
Caribbean expeditions with a summary of the West Indian
shallow-water species (Crustacea: Decapoda: Natantia).
Smithson. Contrib. Zool. 98, 1^179.
Chenoweth, S. F., Hughes, J. M., Keenan, C. P. & Lavery, S.
1998 When oceans meet: a teleost shows secondary intergra-
dation at an Indian
^
Paci?c interface. Proc. R. Soc. Lond. B 265,
415^420.
Coates, A. G. & Obando, J. A. 1996 The geological evolution of
the Central American Isthmus. In Evolution and environment in
tropical America (ed. J. B. C. Jackson, A. F. Budd & A. G.
Coates), pp. 21^56. University of Chicago Press.
Collins, L. S., Budd, A. F. & Coates, A. G. 1996 Earliest evolu-
tion associated with closure of the Tropical American Seaway.
Proc. Natn. Acad. Sci. USA 93, 6069^6072.
Collins, T. M. 1989 Rates of mitochondrial DNA evolution in trans-
isthmian geminate species. PhD dissertation,Yale University, New
Haven, Connecticut.
Collins, T. 1996 Molecular comparisons of transisthmian species
pairs: rates and patterns of evolution. In Evolution and environ-
ment in tropical America (ed. J. B. C. Jackson, A. F. Budd &
A. G. Coates), pp. 303^334. University of Chicago Press.
Collins, T. M., Frazer, K., Palmer, A. R., Vermeij, G. J. &
Brown, W. M. 1996 Evolutionary history of northern hemi-
sphere Nucella (Gastropoda, Muricidae): molecular,
morphological, ecological, and paleontological evidence.
Evolution 50, 2287^2304.
Cronin, T. M. & Dowsett, H. J. 1996 Biotic and oceanographic
response to the Pliocene closing of the Central American
Isthmus. In Evolution and environment in tropical America (ed.
J. B. C. Jackson, A. F. Budd & A. G. Coates), pp. 76^
104. University of Chicago Press.
Cunningham, C. W. & Collins, T. M. 1994 Developing model
systems for molecular biogeography: vicariance and inter-
change in marine invertebrates. In Molecular ecology and
evolution: approaches and applications (ed. B. Schierwater, B.
Streit, G. P. Wagner & R. DeSalle), pp. 405^433. Basle,
Switzerland: Birkh?user .
Cunningham, C. W. & Collins, T. M. 1998 Beyond area rela-
tionships: extinction and recolonization in molecular marine
biogeography. In Molecular approaches to ecology and evolution (ed.
R. DeSalle & B. Schierwater). Basle, Switzerland: Birkha
?
user.
(In the press.)
Farrell, J. W., Ra?, I., Janecek, T. R., Murray, D. W., Levitan,
M., Dadey, K. A., Emeis, K.-C., Lyle, M., Flores, J.-A. &
Hovan, S. 1995 Late Neogene sedimentation patterns in the
eastern equatorial Paci?c Ocean. Proc. Ocean Drill. Prog., Scient.
Res. 138, 717^756.
Goldman, N. 1994 Variance to mean ratio, R(t), for Poisson
processes on phylogenetic trees. Molec. Phylog. Evol. 3, 230^239.
Grosberg, R. K. 1982 Intertidal zonation of barnacles: the in?u-
ence of planktonic zonation of larvae on vertical distribution
of adults. Ecology 63, 894^899.
Hart, M.W., Byrne, M. & Smith, M. J. 1997 Molecular phylo-
genetic analysis of life-history evolution in asterinid star?sh.
Evolution 51, 1848^1861.
Hellberg, M. E. 1998 Sympatric sea shells along the sea's shore:
the geography of speciation in the marine gastropod Tegula.
Evolution. (In the press.)
Jackson, J. B. C., Jung, P., Coates, A. G. & Collins, L. S. 1993
Diversity and extinction of tropical American mollusks and
emergence of the Isthmus of Panama. Science 260, 1624^1626.
Jackson, J. B. C., Jung, P. & Fortunato, H. 1996 Paciphillia
revisited: transisthmian evolution of the Strombina group
(Gastropoda: Columbellidae). In Evolution and environment in
tropical America (ed. J. B. C. Jackson, A. F. Budd & A. G.
Coates), pp. 234^270. University of Chicago Press.
Joseph, L., Moritz, C. & Hugall, A. 1995 Molecular support for
vicariance as a source of diversity in rainforest. Proc. R. Soc.
Lond. B 260, 177^182.
Kim, W. & Abele, L. G. 1988 The snapping shrimp genus
Alpheus from the eastern Paci?c (Decapoda: Caridea:
Alpheidae). Smithson. Contrib. Zool. 454, 1^119.
Klicka, J. & Zink, R. M. 1997 The importance of recent ice
ages in speciation: a failed paradigm. Science 277, 1666^1669.
2262 N. Knowlton and L. A.Weigt Molecular divergence across the Isthmus of Panama
Proc. R. Soc. Lond. B (1998)
Knowlton, N. & Mills, D. E. K. 1992 The systematic impor-
tance of color and color pattern: evidence for complexes of
sibling species of snapping shrimp (Caridea: Alpheidae:
Alpheus) from the Caribbean and Paci?c coasts of Panama.
Proc. San Diego Soc. Nat. Hist. 18, 1^5.
Knowlton, N. & Weigt, L. A. 1997 Species of marine inverte-
brates: a comparison of the biological and phylogenetic
species concepts. In Species: the units of biodiversity (ed. M. F.
Claridge, H. A. Dawah & M. R. Wilson), pp. 199^219.
London: Chapman & Hall.
Knowlton, N., Weigt, L. A., Solo
?
rzano, L. A., Mills, D. K. &
Bermingham, E. 1993 Divergence in proteins, mitochondrial
DNA, and reproductive compatibility across the Isthmus of
Panama. Science 260, 1629^1632.
Lessios, H. A. 1979 Use of Panamanian sea urchins to test the
molecular clock. Nature 280, 599^601.
Lessios, H. A. 1998 The ?rst stage of speciation as seen in
organisms separated by the Isthmus of Panama. In Endless
forms: species and speciation (ed. D. Howard & S. Berlocher).
Oxford University Press. (In the press.)
McMillan, W. O. & Palumbi, S. R. 1995 Concordant evolu-
tionary patterns among Indo-west Paci?c butter?y?shes. Proc.
R. Soc. Lond. B 260, 229^236.
Martin, A. P., Naylor, G. J. P & Palumbi, S. R. 1992 Rates of
mitochondrial DNA evolution in sharks are slow compared
with mammals. Nature 357, 153^155.
Metz, E. C., Go? mez-Gutie? rrez, G. & Vacquier, V. D. 1998
Mitochondrial DNA and bindin gene sequence evolution
among allopatric species of the sea urchin genus Arbacia.
Molec. Biol. Evol. 15, 185^195.
Miya, M. & Nishida, M. 1997 Speciation in the open sea. Nature
389, 803^804.
Morrison, C. L. 1997 Patterns of nucleotide substitution and lineage-
speci?c rates of divergence among snapping shrimp separated by the
Isthmus of Panama. PhD dissertation, Florida State University,
Tallahassee, Florida.
Palumbi, S. R. 1994 Genetic divergence, reproductive isolation,
and marine speciation. A. Rev. Ecol. Syst. 25, 547^572.
Palumbi, S. R. 1996 What can molecular genetics contribute to
marine biogeography? An urchin's tale. J. Exp. Mar. Biol. Ecol.
203, 75^92.
Patton, J. L. & da Silva, M. N. F. 1998 Rivers, refuges, and
ridges: the geography of speciation of Amazonian mammals.
In Endless forms: species and speciation (ed. D. Howard & S.
Berlocher). Oxford University Press. (In the press.)
Schubart, C. D., Diesel, R. & Hedges, S. B. 1998 Rapid evolu-
tion to terrestrial life in Jamaican crabs. Nature 393, 363^365.
Strimmer, K., Goldman, N. & van Haeseler, A. 1997 Bayesian
probabilities and quartet puzzling. Molec. Biol. Evol. 14,
210^211.
Sturmbauer, C., Levinton, J. S. & Christy, J. 1996 Molecular
phylogeny analysis of ?ddler crabs: test of the hypothesis of
increasing behavioral complexity in evolution. Proc. Natn.
Acad. Sci. USA 93, 10 855^10 857.
Vawter, A. T., Rosenblatt, R. & Gorman, G. C. 1980 Genetic
divergence among ?shes of the eastern Paci?c and the
Caribbean: support for the molecular clock. Evolution 34,
705^711.
Vermeij, G. J. 1991 When biotas meet: understanding biotic
interchange. Science 253, 1099^1104.
Wu, C.-I. & Li,W.-H. 1985 Evidence for higher rates of nucleo-
tide substitution in rodents than in man. Proc. Natn. Acad. Sci.
USA 82, 1741^1745.
As this paper exceeds the maximum length normally permitted,
the authors have agreed to contribute to production costs.
Molecular divergence across the Isthmus of Panama N. Knowlton and L. A.Weigt 2263
Proc. R. Soc. Lond. B (1998)