PUBLICATIONS

Paleoceanography
RESEARCH ARTICLE A mechanism for freshening the Caribbean Sea
10.1002/2013PA002515
in pre-Ice Age time
Key Points: Alberto M. Mestas-Nuñez1 and Peter Molnar2
• Caribbean to Pacific moisture transport
decreases during El Niño events 1
Department of Physical and Environmental Sciences, Texas A&M University-Corpus Christi, Corpus Christi, Texas, USA,
• El Niño-like SSTs could account for a less 2
saline Caribbean before Ice Age time Department of Geological Sciences and Cooperative Institute for Research in Environmental Sciences, University of
• These results are consistent with an Colorado Boulder, Boulder, Colorado, USA
Isthmus of Panama long before 4 Ma

Abstract Many believe that the Central American Seaway closed near 4 Ma and that that closure led
to increased salinity in the Caribbean Sea and stronger Meridional Overturning Circulation in the Atlantic,
Correspondence to:
A. M. Mestas-Nuñez,
which facilitated the waxing and waning of ice sheets in the Northern Hemisphere. We offer an alternative
alberto.mestas@tamucc.edu explanation for Caribbean salinification. The atmosphere transports approximately 0.23 Sv (1 Sv = 106 m3s 1)
of fresh water (moisture) from the Caribbean to the Pacific today, but that amount varies by >20% during
Citation:
El Niño–Southern Oscillation events. Regressions of moisture transport against the Niño-3 index, a measure
Mestas-Nuñez, A. M., and P. Molnar of the sea surface temperature in the eastern tropical Pacific, show less moisture transport from the Caribbean
(2014), A mechanism for freshening the during El Niño events than average. Abundant evidence indicates that at 3–4 Ma the eastern tropical Pacific
Caribbean Sea in pre-Ice Age time,
Paleoceanography, 29, 508–517,
was 3.5–4°C warmer than today, and if so, an extrapolation of such regressions suggests that smaller moisture
doi:10.1002/2013PA002515. transport across Central America might account for paleoceanographic inferences of a smaller salinity difference
between the Caribbean and Pacific at that time. Accordingly, that decreased salinity difference at ~3–4 Ma
Received 27 MAY 2013 would not require blockage of relatively fresh Pacific water at ~2–4 Ma by the closure of the Central American
Accepted 9 MAY 2014
Accepted article online 14 MAY 2014
Seaway, but rather would be consistent with a transition from El Niño to La Niña-like conditions in the eastern
Published online 11 JUN 2014 tropical Pacific around that time.

1. Introduction
Many discussions of both modern climate and paleoclimate assign a key role to the Atlantic meridional
overturning circulation (AMOC), and the high salinity of surface water exiting the Caribbean Sea and flowing
northward into the North Atlantic seems to be an essential ingredient of this circulation. Accordingly,
hypotheses for climate changes on virtually all timescales appeal to processes that affect that salinity. Many
have inferred that a closing of the Central American Seaway (1) blocked fresher water from the Pacific and
allowed saline water to accumulate in the Caribbean (Figure 1a) before moving northward, and then (2) the
change from fresher to more saline surface water in the Caribbean played a crucial role in the transition from
essentially no Northern Hemisphere ice sheets to recurring ice ages since ~ 3 Ma [e.g., Bartoli et al., 2005;
Berggren and Hollister, 1974; Haug and Tiedemann, 1998; Kaneps, 1979; Keigwin, 1982; Sarnthein et al., 2009;
Weyl, 1968]. In this paper we present an alternate mechanism for the salinification of the Caribbean ~3–4 Ma
that does not depend on a simultaneous closing of the Central American Seaway. We first review the
paleoceanographic evidence for the emergence of the Seaway. Then we discuss the evolution of the
temperature and salinity differences between the Pacific and the Atlantic Oceans. Finally, we review the
relationship between El Niño–Southern Oscillation (ENSO) and interocean moisture fluxes and use it to
account for the evolution of the interocean salinity contrast.

2. Timing for the Emergence of the Central America Seaway

A wealth of evidence indeed shows changing conditions between the eastern Pacific Ocean and Caribbean
Sea since ~5 Ma. These include divergence, including first and last appearances, of taxa on either side of
Central America [e.g., Allmon et al., 1993, 1996; Budd et al., 1996; Coates et al., 1992; L. S. Collins, 1996; Collins
et al., 1996a, 1996b; T. Collins, 1996; Jackson et al., 1993, 1996; Kaneps, 1979; Keigwin, 1978; Keller et al., 1989;
Kirby and Jackson, 2004; Knowlton et al., 1993; McDougall, 1996; Saito, 1976]. Second, much evidence also
indicates changes in the properties of water in the two oceans as manifested by oxygen isotopes with or
without measurements of sea surface temperatures (SSTs) in both plankton [e.g., Groeneveld et al., 2006,
2008; Haug et al., 2001; Keigwin, 1982; Keller et al., 1989; Steph et al., 2006a, 2006b, 2010] and molluscs

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 508
 Paleoceanography 10.1002/2013PA002515

[e.g., Teranes et al., 1996]. Collectively,

these data call for not only increasing
salinity on the Caribbean but also
decreasing salinity on the Pacific side
[e.g., Groeneveld et al., 2006]. Third,
the range of zooid sizes of cheilostome
bryozoans correlates well with seasonal
differences in SSTs [e.g., O’Dea and
Jackson, 2002; O’Dea and Okamura, 2000;
Okamura et al., 2011], and O’Dea et al.
[2007] showed that size distributions
from the southwestern Caribbean
show a marked decrease since ~4 Ma,
suggesting a decrease in seasonal
variability of SSTs and a diminution of
upwelling. This contrasts with the larger
variation in eastern Pacific SSTs in
present-day conditions associated with
ENSO variability. O’Dea et al. [2012] also
inferred that in Pliocene time seasonal
upwelling on the Pacific side was modest
but became strong in Pleistocene time.
One of us, Molnar [2008], reviewed
much of the evidence discussed above,
and he concluded that both taxonomic
and isotopic changes occurred over a
Figure 1. (a) Mean salinity of the Caribbean and adjacent oceans based
on data by Conkright et al. [1998]. Image obtained from the website of the protracted interval from 6 to 2 Ma,
Physical Sciences Division, Earth System Research Laboratory, NOAA, Boulder, which made it difficult to assign closure
Colorado (http://www.esrl.noaa.gov/psd/). (b) Calculated flow rates in Sv of the seaway, or even shoaling of a
6 3 1
(1 Sv = 10 m s ) versus depth of gaps in the Central American Seaway for deep seaway to a precise time. There
different GCM runs. Data are from a compilation by Zhang et al. [2012].
seems to be little doubt, however, that
Different symbols are coded to reflect different widths of gaps: ~200 km
[Steph et al., 2006a, 2006b], 450 km [Lunt et al., 2008; Schneider and Schmittner, the contrast in salinity between the
2006], and ~1000 km [Nisancioglu et al., 2003; Prange and Schulz, 2004; Zhang Pacific and Caribbean increased since
et al., 2012], and different rates of vertical mixing as parameterized by a ~4 Ma [see also Sarnthein et al., 2009],
vertical diffusivity, with open circles showing runs with large values [e.g., which is the fact that motivates us here,
Schneider and Schmittner, 2006]. Other differences include an allowance for
as well as changes in seasonality in the
water to pass through the Bering Strait [e.g., Nisancioglu et al., 2003],
which most models do not include, and different latitudes of the gap, for southwestern Caribbean [O’Dea et al.,
Nisancioglu et al. [2003] used a gap across Mexico and northern Central 2007] and eastern Pacific [O’Dea et al.,
America (1000 m deep), but most assume it to have been near Panama. 2007], which are not our focus.
Recent work challenges the connection
between changes in properties of Caribbean surface water and the hypothesized emergence of the Isthmus
of Panama at ~3–4 Ma. First, Coates et al. [1992]; Coates and Obando [1996] inferred from geologic study in
Panama that a barrier to deep water had formed before 7 Ma. More recently, Montes et al. [2012a, 2012b]
argued that essentially all of the rock cropping out in Panama, including terrigenous sedimentary rock, was
in place perhaps as long ago as 15–20 Ma. Fossils of quite large terrestrial vertebrates dating from ~20 Ma
have been found in both the Panama Canal Zone and North America, suggesting that a land bridge was
present at that time [MacFadden, 2006; MacFadden et al., 2010]. Thermochronological studies from Panama
and Colombia are consistent with eastern Panama being contiguous with northwestern South America at this
time [Farris et al., 2011]. Finally, other commonly quoted observations used as evidence for the emergence of
the isthmus near 3 Ma can be interpreted differently and leave open the date of the Isthmus’s formation (see
summary by Molnar [2008]). Although it is virtually impossible to demonstrate conclusively that there have
been no shallow gaps through which water could cross Central America or Panama since 20 Ma, the bulk of
recent evidence makes the absence of such gaps, especially wide deep ones, a viable working hypothesis.

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 509
 Paleoceanography 10.1002/2013PA002515

3. Evolution of SST in the Eastern Tropical Pacific

Many relate the shift from equable Northern Hemisphere climates before ~4 Ma to recurring ice ages since
3 Ma to a concurrent shift in sea surface temperatures in the tropical Pacific [e.g., Barreiro et al., 2006;
Fedorov et al., 2006; Huybers and Molnar, 2007; Molnar and Cane, 2002, 2007; Philander and Fedorov, 2003;
Ravelo et al., 2004, 2006]. At ~4 Ma, the mean sea surface temperature in the eastern Tropical Pacific
was ~3.5–4°C warmer than it is today [e.g., Chaisson and Ravelo, 2000; Dekens et al., 2007; Lawrence et al.,
2006; Wara et al., 2005], and the thermocline seems to have been deeper before 3–4 Ma than today [e.g.,
Cannariato and Ravelo, 1997; Dekens et al., 2007; Steph et al., 2006a, 2006b; Wara et al., 2005]. Moreover,
the results of O’Dea et al. [2012] can be interpreted to suggest that the seasonal upwelling that typifies
present-day conditions in the Gulf of Panama and the typical background La Niña state was absent in Pliocene
time and developed only later in Pleistocene time.
Several have pointed out that these various observations are consistent with the image that the eastern
tropical Pacific before Ice Age time resembled conditions that prevail during El Niño events [e.g., Fedorov
et al., 2006; Molnar and Cane, 2002, 2007; Philander and Fedorov, 2003; Ravelo et al., 2004, 2006]. Both analyses
of modern data and calculations using general circulation models (GCMs) of the atmosphere show that a
warmer eastern tropical Pacific will lead to longer summers in Canada that, in turn, should inhibit, if not
prevent, the growth of continental ice sheets there [e.g., Barreiro et al., 2006; Brierley and Fedorov, 2010;
Huybers and Molnar, 2007; Shukla et al., 2009, 2011; Vizcaíno et al., 2010]. Although still controversial, if correct,
this explanation would obviate the need to relate high-latitude Northern Hemisphere climate to a change in
AMOC via increased salinity of North Atlantic surface water, and by extension to the closing of the Central
American Seaway [e.g., Bartoli et al., 2005; Sarnthein et al., 2009; Steph et al., 2010].
The discussion above ignores the question of what processes led to the demise of a warm eastern tropical
Pacific Ocean. As this question is open, and not pertinent to the question that we ask here, we merely refer
readers to studies that have addressed possible processes [Cane and Molnar, 2001; Dayem et al., 2007; Fedorov
et al., 2006, 2010, 2013; Philander and Fedorov, 2003].

4. Salt Fluxes and the Evolution of the Interocean Salinity Difference

Even if one ignores the possible role that increasing salinity might have on Ice Age climate, numerous
paleoceanographic inferences of salinity suggest that the Caribbean became more saline than the Pacific
between 5 and 3 Ma [e.g., Groeneveld et al., 2006, 2008; Haug et al., 2001; Keigwin, 1982; Steph et al., 2006a,
2006b, 2010; Teranes et al., 1996]. For surface water in the North Atlantic to be atypically saline, either
relatively saline water must enter from the South Atlantic or the atmosphere must transport moisture
(fresh water) from the North Atlantic. Zaucker et al. [1994] pointed out that the net transport of ~20 Sv
(1 Sv = 106 m3 s 1) in the AMOC with a salinity of 0.6‰ higher than average (or 12 Sv ‰) can be balanced
by a much smaller rate of freshwater transport with salinity ~35‰ less than the saline ocean; freshwater
transport of 0.34 Sv yields ~12 Sv ‰, which is similar to the flux of freshwater from the Caribbean into the
Pacific today of ~0.23 Sv (or ~8.1 Sv ‰) [Mestas-Nuñez et al., 2007]. Zaucker et al. [1994] found that in model
runs, the strength of the MOC did not scale linearly with such flux, and they concluded that a blockage of
0.23 Sv of freshwater transport should not shut such circulation down. By contrast, Schmittner et al. [2000]
inferred that a decrease of 0.2 Sv of moisture transport from the Atlantic would suffice to bring AMOC to a
halt. Although a simple scaling of moisture flux to the strength of the AMOC cannot be made, the moisture
flux across Central America and Panama could play a principal role in the Atlantic-Pacific salinity difference.
Those who consider the closure of the Central American Seaway to be essential to a strengthening of the
AMOC postulate that an earlier transport of relatively fresh water from the Pacific prevented the Atlantic from
becoming sufficiently saline to sink at high latitudes, but controversy pervades this topic. Using Godfrey’s
[1989] “island rule,” Nof and Van Gorder [2003] argued that winds like those of today would force flow of
Atlantic water across an open Central American Seaway to the Pacific; only if the AMOC were large, would
water flow from the Pacific across an open Central American Seaway. Calculations made with general
circulation models (GCMs), however, either forced by present-day winds or coupled to an atmospheric GCM,
invariably not only show flow from the Pacific to the Atlantic but also show a weaker AMOC with an open
seaway than with it closed [Lunt et al., 2008; Nisancioglu et al., 2003; Prange and Schulz, 2004; Schneider and
Schmittner, 2006; Steph et al., 2006a, 2010; Zhang et al., 2012].

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 510
 Paleoceanography 10.1002/2013PA002515

Calculated Pacific-to-Atlantic flow ranges

from 5 to 16 Sv [e.g., Zhang et al., 2012]. If
the salinity difference were as large as
that today, ~1‰ [Haug et al., 2001], then
the closing of the seaway would increase
Atlantic salinity at a rate of 5–16 Sv ‰,
comparable to the transport by the
atmosphere across the Central America
and Panama today, ~8 Sv ‰. If the
Pacific were only 0.5‰ fresher than the
Atlantic, then shutting down flow through
the Central American Seaway would
increase salinity in the Atlantic at a rate of
2.5–8 Sv ‰. The range of 5 to 16 Sv of
calculated Pacific-to-Atlantic flow results
largely from different assumed depths of
gaps in Central America (Figure 1b) and to
a lesser extent in widths of such gaps.
Because the flux of moisture transported
from the Atlantic to the Pacific by the
atmosphere today is comparable to the
hypothesized dilution of saline Atlantic
water by an influx of fresher Pacific
water, we extend results of Mestas-Nuñez
et al. [2007] showing that moisture flux
from the Caribbean to the Pacific varies
with ENSO.

5. Moisture Fluxes Between

Figure 2. (top) December–March and (bottom) June–September averages the Atlantic and Pacific
of the monthly vertically integrated water vapor flux vectors calculated
from the daily 1960–2003 NCEP-NCAR reanalysis. The magnitudes of the Mestas-Nuñez et al. [2007] showed that
water vapor flux vectors are color contoured. The thick solid (dashed) variations in the transport of moisture
lines indicate the eastern and western (northern and southern) Intra- from the Intra-Americas Sea (composed
Americas Sea boundary segments used by Mestas-Nuñez et al. [2007]. of the Gulf of Mexico and Caribbean Sea)
(From Mestas-Nuñez et al. [2007])
across Central America is associated
with the variability of the atmospheric
Caribbean low-level jet, an easterly jet just north of the South American coast (Figure 2). The westward
continuation of this jet transports moisture across the Caribbean to Central America (Figure 2). Unlike
teleconnections to most regions, the ENSO teleconnection to the Atlantic varies with season [e.g., Giannini et al.,
2000]. Consistent with variations in moisture transport across Central America, this jet is stronger in summer
and winter than in spring and autumn [e.g., Cook and Vizy, 2010; Durán-Quesada et al., 2010; Mestas-Nuñez et al.,
2007; Muñoz et al., 2008; Wang, 2007]. During El Niño summers, the jet is stronger than average [e.g., Whyte
et al., 2008], but like other correlations in the region [e.g. Giannini et al., 2000], the opposite holds for El Niño
winters [Wang, 2007].
In this study, we used the monthly anomalies (i.e., deviations from climatology) of the “to west” moisture
flux of Mestas-Nuñez et al. [2007] for the period January 1979 to December 2003. These fluxes represent
anomalous westward moisture transport across the western boundary of the Intra-Americas Sea (western
thick solid white line in Figure 2).
Mestas-Nuñez et al. [2007] showed that anomaly fluxes in winter months correlate negatively with ENSO sea
surface temperature (SST) anomaly indices, like Niño-3 (5°S–5°N, 150°–90°W) and Niño-3.4 (5°S–5°N, 170°–120°W),
and those in summer correlate positively. We extend that analysis by regressing moisture flux anomalies
with monthly Niño-3 SST anomalies for all calendar months and find significant negative correlations

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 511
 Paleoceanography 10.1002/2013PA002515

1
during December–March and significant positive
correlations during July–August (Figure 3). Results
0.5
for the Niño-3.4 index are similar.
Estimates of the linear regression coefficients

Correlation
between Atlantic-to-Pacific moisture flux
0
anomalies in g s 1 and Niño-3 SST anomalies in
°C using data for the months with significant
correlations (see Figure 3) are shown in Figure 4.
−0.5
The regression coefficient for December–March
is 4.59 ± 1.88 × 1010 g s 1 °C 1 (Figure 4a) and
−1 for July–August is 4.52 ± 1.68 × 1010 g s 1 °C 1
J F M A M J J A S O N D (Figure 4b). The overall regression coefficient
Figure 3. Correlation coefficients between Atlantic-to-Pacific
estimated by combining data from the
monthly moisture flux anomalies and Niño-3 monthly SST December–March and July–August periods is
anomalies for all calendar months. Correlations that are significant 2.45 ± 2.02 × 1010 g s 1 °C 1 (Figure 4c). A
with greater than 95% confidence are indicated with solid circles. similar value of the overall regression coefficient
( 2.12 × 1010 g s 1 °C 1) can be obtained by first
computing the regression coefficients for each of the individual months involved (December–March, July,
and August) and then taking their average. The value 0.0245 Sv °C 1 for Niño-3 when multiplied by ~3.5–4°C
is ~0.086–0.098 Sv, which is roughly half of the 0.2 Sv inferred by Schmittner et al. [2000] for moisture
transport from the entire Atlantic (20°S–20°N) and hence not exclusively across Central America, during
El Niño events. Prange et al. [2010] exploited
11 DJFM similar logic to explain changes in salinity
x 10
Moisture flux (g sec−1)

2 since the Last Glacial Maximum.

The flux of moisture from the Caribbean
0
to the Pacific is not the only flux that
−2 correlates with ENSO. Mestas-Nuñez et al.
[2007] showed correlations of fluxes across
−4
a
the southern boundary of the Caribbean
−2 −1 0 1 2 3 with northern South America, across the
JA northern boundary with North America,
x 1011
and across the eastern boundary with the
Moisture flux (g sec−1)

2
Atlantic. These latter three fluxes, however,
0 do not affect the salinity of the Atlantic, for
all of this moisture does not represent a
−2 net loss/gain of Atlantic’s water. Moreover,
b the high flux of surface water through
−4
the Caribbean, ~28 Sv [Johns et al., 2002],
−2 −1 0 1 2 3
suggests that surface water ~150 m deep
x 1011 DJFMJA and ~1000 km wide entering the Caribbean
Moisture flux (g sec−1)

2 through the Lesser Antilles would travel

with an average speed of 0.19 m s 1. Thus,
0 to traverse a sea 2000 km wide, before
entering the Gulf of Mexico, would require
−2
only ~4 months. In fact, surface water in
−4
c the southwestern Caribbean, near Panama,
−2 −1 0 1 2 3 circulates in small, nearly closed gyre and
Nino 3 (deg C) might become more saline than that of the
average North Atlantic [Johns et al., 2002].
Figure 4. Linear regression between Atlantic-to-Pacific monthly moisture
In any case, on climate timescales the only
flux anomalies and Niño-3 monthly SST anomalies for (a) December–
March, (b) July–August, and (c) December–March and July–August
flux of moisture that affects salinity of the
combined. The respective correlation coefficients are 0.57, 0.58, and North Atlantic, and hence Caribbean, is that
0.31, and they are all significant with greater than 95% confidence. across Central America [e.g., Weyl, 1968].

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 512
 Paleoceanography 10.1002/2013PA002515

Paleoceanographic data from the Ceara Rise, ~4°N and north of the mouth of the Amazon, suggests that
near 4 Ma, that region became less saline [Billups et al., 1998; Ravelo and Wara, 2004; Steph et al., 2006a].
Consistently, using Xie and Arkin’s [1996] precipitation data set, Saravanan and Chang [2000] showed that
during El Niño events a band of enhanced precipitation lies north of an arid region over and east of the
Amazon Basin. One might imagine that the freshening north of the mouth of the Amazon near 4 Ma and
the northward flow of surface water might reduce salinity of the North Atlantic. Estimates of the stream
function for flow of surface water through the Caribbean, however, suggest that most of the water entering
the sea comes from the North Atlantic [Johns et al., 2002]. Thus, the freshening of equatorial water near the
Amazon Basin does not seem likely to have affected the salinity of the North Atlantic.

6. Implications for Caribbean Freshening and Paleoclimate

With abundant evidence that at 3–4 Ma the eastern tropical Pacific was warmer, and hence, more El Niño like,
than now, we exploit these regressions to examine how moisture transport at that time might have differed
from that today. If the eastern tropical Pacific were 2° or 4°C warmer than today, these regressions suggest
that less fresh water would be transported from the Caribbean by 0.049 or 0.098 Sv. Salinity in the Caribbean
would have been diminished at a rate of 1.7 or 3.4 Sv ‰ or approximately 21 or 43% smaller than the current
~8 Sv ‰ amount. These rates are comparable to those of model calculations, ~5 Sv, of eastward transport
across a shallow Central American Seaway (Figure 1b) and an assumed difference in salinity of 0.5‰ between
the Pacific and Caribbean at 4 Ma, or 2.5 Sv ‰. Depending on one’s point of view, a warm eastern tropical
Pacific, like that during large El Niño events, could account for the difference in salinity between the Caribbean
and Pacific at 4 Ma smaller than today’s, especially if one accepts the view that by 7 Ma, any opening must have
been shallow [Coates and Obando, 1996; Coates et al., 1992]. Alternatively, the small contribution to that salinity
difference might be taken as evidence that a seaway deeper than 100–200 m must have existed. With the
geologic evidence limiting the depth of the seaway and allowing for an isthmus since perhaps as long ago as
20 Ma [e.g., Montes et al., 2012a, 2012b], we favor the former.
We realize that we take a risk in assuming that the El Niño state in modern climate can be transferred to
past climate when average eastern tropical Pacific SSTs resembled those during El Niño events. This is
rendered complicated in part because paleoceanographic evidence cannot yet discriminate between a
climate similar to that of the present but with more common El Niño events, including more large events, or a
state in which past variability was small, and the mean SST in the eastern Pacific was several °C warmer than
today’s. Some GCM runs with a permanently warm eastern Pacific SST do reveal a background state with
features associated with transient El Niño events of modern climate [e.g., Barreiro et al., 2006; Brierley and
Fedorov, 2010; Shukla et al., 2009, 2011; Vizcaíno et al., 2010] and that replicate differences between Pliocene
and present-day climates [e.g., Molnar and Cane, 2002, 2007]. At the same time, some teleconnections
associated with El Niño events might, if they became permanent, affect moisture transport across Central
America differently from the direct effect that we report here.
In a study of heating throughout the tropics during the 1997–1998 El Niño, Chiang and Lintner [2005]
showed that the Central Atlantic surface water warms by around 1°C, in the spring following the peak
of the event in the Pacific. Enfield and Mayer [1997] noted that this Atlantic warming is a typical feature
of ENSO variability. The 1°C amplitude estimated by Chiang and Lintner for the 1997–1998 ENSO,
however, is an upper bound, since generally the amplitude of the Atlantic warming is about 10–30% of
the eastern tropical Pacific amplitude of about 1–2°C [e.g., see Mestas-Nuñez and Enfield, 2001, Figure 1].
Nevertheless, by drawing analogy with transient ENSO events, we ignore the possible impact of this
warming, which would reduce the temperature gradient between the central Atlantic and eastern
Pacific, and which might weaken the Caribbean low-level jet. Furthermore, Xie et al. [2008] showed that
with imposed cooling of the North Atlantic, a high-pressure center developed over the Caribbean,
which strengthened northeasterly winds across Central America. Presumably the opposite, a warming
as occurs following El Niño events, would lead to a weakening of such winds and a reduction in moisture
transport. We hesitate to suggest that the warming inferred by Chiang and Lintner [2005] during El Niño
events would serve as a positive feedback, for attempts with GCMs to simulate the location and height of
the Caribbean low-level jet, and especially its seasonal variability, have resulted in only limited success
[Martin and Schumacher, 2011]. Thus, we merely note that the link with a permanent El Niño-like climate
carries a speculative aspect.

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 513
 Paleoceanography 10.1002/2013PA002515

In this study, the paleo-observation that a salinity gradient between the Pacific and Atlantic developed
around ~4.0 Ma is based on the assumption that the Caribbean salinity is a good representation of the
tropical Atlantic. As such, our results seem to be in contrast with those of Schmittner et al. [2000], who showed
enhanced rather than reduced freshwater flux out of the Atlantic during warm ENSO events. Schmittner et al.,
however, based their inference on surface freshwater flux (precipitation minus evaporation) estimates from
the National Centers for Environmental Prediction (NCEP)-National Center for Atmospheric Research (NCAR)
reanalysis [Kalnay et al., 1996], which are known to be of poor quality. Indeed, Mestas-Nuñez et al. [2005]
showed that although the freshwater fluxes from the NCEP-NCAR reanalysis were not useful for moisture
budget studies of the Caribbean and Gulf of Mexico, the vertically integrated moisture fluxes from the same
reanalysis were acceptable, and those are the ones used in the present study. In addition, Schmittner et al.
did not consider the seasonality of the ENSO relationships.

7. Conclusions
During El Niño events, atmospheric moisture transport from the Caribbean Sea to the Pacific Ocean decreases
in winter and increases in summer, but such that less moisture leaves the Caribbean during El Niño than
during most years. Regressions of such moisture transport against the Niño-3 SST anomaly index quantify
these relationships (Figure 4).
Paleoceanographic evidence suggests that at 3–4 Ma the eastern tropical Pacific Ocean was warmer by ~4°C
than it commonly is today [e.g., Dekens et al., 2007; Lawrence et al., 2006; Wara et al., 2005], but comparable to
what it is during large El Niño events, and that the difference in salinity between the Pacific and Caribbean
was smaller by ~1‰ [e.g., Steph et al., 2006a]. Extrapolations of the regressions to Niño-3 indices of 2°C and 4°C
suggest that moisture transports would have been 21–42% less than the present-day average. The effects of
such moisture transport on Caribbean salinity are similar in magnitude to those estimated from ocean GCM
runs with only shallow Central American Seaways. Reasonable extrapolations of the regressions, to 0.049 or
0.098 Sv, do not yield such a large value of atmospheric moisture transport, but they show them to be of the
order of magnitude that might yield a large difference in climate.
Recent geologic studies suggest that the Isthmus of Panama may have formed as long ago as 15–20 Ma [Montes
et al., 2012a, 2012b]. Lower moisture transport across that region because of a warmer eastern Tropical Pacific
at ~3–4 Ma might account for the smaller salinity difference between the Caribbean Sea and Pacific Ocean at
that time. A later transition to La Niña-like conditions and associated increase in moisture transport from the
Caribbean to the Pacific could account for the salinification of the Caribbean ~2–4 Ma without invoking a
simultaneous closing of the Central American seaway. Our proposed mechanism for Caribbean freshening in
pre-Ice Age times may bear on future climates. In a warming world, the zonal interocean exchange of moisture
at low latitudes may contribute to the slowdown of the AMOC.

Acknowledgments References
One of the authors (P.M.) was stimulated
Allmon, W. D., G. Rosenberg, R. W. Portell, and K. S. Schinder (1993), Diversity in Atlantic coastal plain mollusks since the Pliocene, Science,
by discussions with C. Montes and R.F.
260, 1626–1629.
Stallard. A.C. Ravelo provided advice on
Allmon, W. D., G. Rosenberg, R. W. Portell, and K. Schindler (1996), Diversity of Pliocene-recent mollusks in the western Atlantic: Extinction,
how to structure our arguments, and
origination, and environmental change, in Evolution and Environment in Tropical America, edited by J. B. C. Jackson, A. F. Budd, and
she, J.C.H. Chiang, and an anonymous
A. G. Coates, pp. 271–302, Univ. of Chicago Press, Chicago, Ill.
reviewer offer constructive criticism of
Barreiro, M., G. Philander, R. Pacanowski, and A. Fedorov (2006), Simulations of warm tropical conditions with application to middle Pliocene
the manuscript. The research was
atmospheres, Clim. Dyn., 26, 349–365.
supported in part by the National
Bartoli, G., M. Sarnthein, M. Weinelt, H. Erlenkeuser, D. Garbe-Schönberg, and D. W. Lea (2005), Final closure of Panama and the onset of
Science Foundation (NSF) under grant
northern hemisphere glaciation, Earth Planet. Sci. Lett., 237, 33–44.
AGS-1136466 and while the other author
Berggren, W. A., and C. D. Hollister (1974), Paleogeography, paleobiogeography and the history of circulation in the Atlantic Ocean, in Studies
(A.M.) was serving at NSF. Any opinion,
in Paleo-Oceanography, Spec. Publ. No. 20, edited by W. W. Hay, pp. 126–186, Soc. Econ. Paleontol. Mineral., Tulsa, Okla.
findings, and conclusions or recommen-
Billups, K., A. C. Ravelo, and J. C. Zachos (1998), Early Pliocene climate: A perspective from the western equatorial Atlantic warm pool,
dations expressed in this publication
Paleoceanography, 13, 459–470, doi:10.1029/98PA02262.
are those of the authors and do not
Brierley, C. M., and A. V. Fedorov (2010), Relative importance of meridional and zonal sea surface temperature gradients for the onset of the
necessarily reflect the views of NSF.
ice ages and Pliocene-Pleistocene climate evolution, Paleoceanography, 25, PA2214, doi:10.1029/2009PA001809.
Budd, A. F., K. G. Johnson, and T. A. Stemann (1996), Plio-Pleistocene turnover and extinctions in the Caribbean reef-coral fauna, in Evolution and
Environment in Tropical America, edited by J. B. C. Jackson, A. F. Budd, and A. G. Coates, pp. 168–204, Univ. of Chicago Press, Chicago, Ill.
Cane, M. A., and P. Molnar (2001), Closing of the Indonesian seaway as a precursor to east African aridification around 3–4 million years ago,
Nature, 411, 157–162.
Cannariato, K. G., and A. C. Ravelo (1997), Plio-Pleistocene evolution of eastern tropical Pacific rate of circulation and thermocline depth,
Paleoceanography, 12, 805–820, doi:10.1029/97PA02514.

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 514
 Paleoceanography 10.1002/2013PA002515

Chaisson, W. P., and A. C. Ravelo (2000), Pliocene development of the east–west hydrographic gradient in the tropical Pacific, Paleoceanography,
15, 497–505, doi:10.1029/1999PA000442.
Chiang, J. C. H., and B. R. Lintner (2005), Mechanisms of remote tropical surface warming during El Niño, J. Clim., 18, 4130–4149.
Coates, A. G., and J. A. Obando (1996), The geologic evolution of the Central American Isthmus, in Evolution and Environment in Tropical America,
edited by J. B. C. Jackson, A. F. Budd, and A. G. Coates, pp. 21–56, Univ. of Chicago Press, Chicago, Ill.
Coates, A. G., J. B. C. Jackson, L. S. Collins, T. M. Cronin, H. J. Dowsett, L. M. Bybell, P. Jung, and J. A. Obando (1992), Closure of the Isthmus of
Panama: The near-shore marine record of Costa Rica and western Panama, Geol. Soc. Am. Bull., 104, 814–828.
Collins, L. S. (1996), Environmental changes in Caribbean shallow waters relative to the closing tropical American seaway, in Evolution and
Environment in Tropical America, edited by J. B. C. Jackson, A. F. Budd, and A. G. Coates, pp. 130–167, Univ. of Chicago Press, Chicago, Ill.
Collins, L. S., A. G. Coates, W. A. Berggren, M.-P. Aubry, and J.-J. Zhang (1996a), The late Miocene Panama isthmian strait, Geology, 24, 687–690.
Collins, L. S., A. F. Budd, and A. G. Coates (1996b), Earliest evolution associated with closure of the Tropical American seaway, Proc. Natl. Acad.
Sci. U.S.A., 93, 6069–6072.
Collins, T. (1996), Molecular comparisons of transisthmian species pairs: Rates and patterns of evolution, in Evolution and Environment in
Tropical America, edited by J. B. C. Jackson, A. F. Budd, and A. G. Coates, pp. 303–334, Univ. of Chicago Press, Chicago, Ill.
Conkright, M. E., et al. (1998), World Ocean Database 1998 Documentation and Quality Control, National Oceanographic Data Center,
Silver Spring, Md.
Cook, K. H., and E. K. Vizy (2010), Hydrodynamics of the Caribbean Low-Level Jet and its relationship to precipitation, J. Clim., 23, 1477–1494.
Dayem, K. E., D. C. Noone, and P. Molnar (2007), Tropical western Pacific warm pool and maritime continent precipitation rates and their
contrasting relationships with the Walker Circulation, J. Geophys. Res., 112, D06101, doi:10.1029/2006JD007870.
Dekens, P. S., A. C. Ravelo, and M. D. McCarthy (2007), Warm upwelling regions in the Pliocene warm period, Paleoceanography, 22, PA3211,
doi:10.1029/2006PA001394.
Durán-Quesada, A. M., L. Gimeno, J. A. Amador, and R. Nieto (2010), Moisture sources for Central America: Identification of moisture sources
using a Lagrangian analysis technique, J. Geophys. Res., 115, D05103, doi:10.1029/2009JD012455.
Enfield, D. B., and D. A. Mayer (1997), Tropical Atlantic SST variability and its relation to El Niño–Southern Oscillation, J. Geophys. Res., 102,
929–945, doi:10.1029/96JC03296.
Farris, D. W., C. Jaramillo, G. Bayona, S. A. Restrepo-Moreno, C. Montes, A. Cardona, A. Mora, R. J. Speakman, M. D. Glascock, and V. Valencia (2011),
Fracturing of the Panamanian Isthmus during initial collision with South America, Geology, 39, 1007–1010.
Fedorov, A. V., P. S. Dekens, M. McCarthy, A. C. Ravelo, P. B. deMenocal, M. Barreiro, R. C. Pacanowski, and S. G. Philander (2006), The Pliocene
Paradox (Mechanisms for a Permanent El Niño), Science, 312, 1485–1489.
Fedorov, A. V., C. M. Brierley, and K. Emanuel (2010), Tropical cyclones and permanent El Niño in the early Pliocene epoch, Nature, 463, 1066–1070.
Fedorov, A. V., C. M. Brierley, K. T. Lawrence, Z. Liu, P. S. Dekens, and A. C. Ravelo (2013), Patterns and mechanisms of early Pliocene warmth,
Nature, 496, 43–49.
Giannini, A., Y. Kushnir, and M. A. Cane (2000), Interannual variability of Caribbean rainfall, ENSO, the Atlantic Ocean, J. Clim., 13, 297–311.
Godfrey, J. S. (1989), A Sverdrup model of the depth-integrated flow for the ocean allowing for island circulations, Geophys. Astrophys. Fluid
Dyn., 45, 89–112.
Groeneveld, J., S. Steph, R. Tiedemann, D. Garbe-Schönberg, D. Nürnberg, and A. Sturm (2006), Pliocene mixed-layer oceanography for Site 1241,
18
using combined Mg/Ca and δ O analyses of Globigerinoides sacculifer, in Proc. ODP, Sci. Results, vol. 202, edited by R. Tiedemann et al.,
pp. 1–27, Ocean Drilling Program, College Station, Tex.
Groeneveld, J., D. Nürnberg, R. Tiedemann, G.-J. Reichart, S. Steph, L. Reuning, D. Crudeli, and P. Mason (2008), Foraminiferal Mg/Ca increase in
the Caribbean during the Pliocene: Western Atlantic Warm Pool formation, salinity influence, or diagenetic overprint?, Geochem. Geophys.
Geosyst., 9, Q01P23, doi:10.1029/2006GC001564.
Haug, G. H., and R. Tiedemann (1998), Effect of the formation of the Isthmus of Panama on Atlantic thermohaline circulation, Nature,
393, 673–676.
Haug, G. H., R. Tiedemann, R. Zahn, and A. C. Ravelo (2001), Role of Panama uplift on oceanic freshwater balance, Geology, 29, 207–210.
Huybers, P., and P. Molnar (2007), Tropical cooling and the onset of North American glaciation, Clim. Past, 3, 549–557.
Jackson, J. B. C., P. Jung, A. G. Coates, and L. S. Collins (1993), Diversity and extinction of the tropical American mollusks and emergence of the
Isthmus of Panama, Science, 260, 1624–1626.
Jackson, J. B. C., P. Jung, and H. Fortunato (1996), Paciphilia revisited: Transisthmian evolution of the Strombina group (Gastropoda:
Columbellidae), in Evolution and Environment in Tropical America, edited by J. B. C. Jackson, A. F. Budd, and A. G. Coates, pp. 234–270,
Univ. of Chicago Press, Chicago, Ill.
Johns, W. E., T. L. Townsend, D. M. Fratantoni, and W. D. Wilson (2002), On the Atlantic inflow to the Caribbean Sea, Deep Sea Res., Part I,
49, 211–243.
Kalnay, E., et al. (1996), The NCEP/NCAR 40-Year Reanalysis Project, Bull. Am. Meteorol. Soc., 77, 437–471.
Kaneps, A. G. (1979), Gulf Stream: Velocity fluctuations during the late Cenozoic, Science, 204, 297–301.
Keigwin, L. D., Jr. (1978), Pliocene closing of the Isthmus of Panama, based on biostratigraphic evidence from nearby Pacific Ocean and
Caribbean Sea cores, Geology, 6, 630–634.
Keigwin, L. D., Jr. (1982), Isotope paleoceanography of the Caribbean and east Pacific: Role of Panama uplift in late Neogene time, Science,
217, 350–353.
Keller, G., C. E. Zenker, and S. M. Stone (1989), Late Neogene history of the Pacific-Caribbean gateway, J. South Am. Earth Sci., 2, 73–108.
Kirby, M. X., and J. B. C. Jackson (2004), Extinction of a fast-growing oyster and changing ocean circulation in Pliocene tropical America,
Geology, 32, 1025–1028.
Knowlton, N., L. A. Weigt, L. A. Solórzano, D. K. Mills, and E. Bermingham (1993), Divergence in proteins, mitochondrial DNA, and reproductive
compatibility across the Isthmus of Panama, Science, 260, 1629–1632.
Lawrence, K. T., Z.-H. Liu, and T. D. Herbert (2006), Evolution of the eastern tropical Pacific through Plio-Pleistocene glaciation, Science,
312, 79–83.
Lunt, D. J., P. J. Valdes, A. Haywood, and I. C. Rutt (2008), Closure of the Panama Seaway during the Pliocene: Implications for climate and
Northern Hemisphere glaciation, Clim. Dyn., 30, 1–18.
MacFadden, B. J. (2006), North American Miocene land mammals from Panama, J. Vert. Paleontol., 26, 720–734.
MacFadden, B. J., M. X. Kirby, A. Rincon, C. Montes, S. Moron, N. Strong, and C. Jaramillo (2010), Extinct Peccary “Cynorca” Occidentale
(Tayassuidae, Tayassuinae) from the Miocene of Panama and correlations to North America, J. Paleontol., 84, 288–298.
Martin, E., and C. Schumacher (2011), The Caribbean Low-Level Jet and its relationship with precipitation in IPCC AR4 Models, J. Clim.,
24, 5935–5950.

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 515
 Paleoceanography 10.1002/2013PA002515

McDougall, K. (1996), Benthic foraminiferal response to the emergence of the Isthmus of Panama and coincident paleoceanographic changes,
Mar. Micropaleontol., 28, 133–169.
Mestas-Nuñez, A. M., and D. B. Enfield (2001), Eastern equatorial Pacific SST variability: ENSO and non-ENSO components and their climatic
associations, J. Clim., 14, 391–402.
Mestas-Nuñez, A. M., C. Zhang, and D. B. Enfield (2005), Uncertainties in estimating moisture fluxes in the Intra-Americas Sea, J. Hydrometeorol.,
6, 696–709.
Mestas-Nuñez, A. M., D. B. Enfield, and C. Zhang (2007), Water vapor fluxes over the Intra-Americas Sea: Seasonal and interannual variability
and associations with rainfall, J. Clim., 20, 1910–1922.
Molnar, P. (2008), Closing of the Central American Seaway and the Ice Age: A critical review, Paleoceanography, 23, PA2201, doi:10.1029/
2007PA001574.
Molnar, P., and M. A. Cane (2002), El Niño’s tropical climate and teleconnections as a blueprint for pre-Ice Age climates, Paleoceanography,
17(2), 1021, doi:10.1029/2001PA000663.
Molnar, P., and M. A. Cane (2007), Early Pliocene (Pre-Ice Age) El Niño-like Global Climate: Which El Niño?, Geosphere, 3(5), 337–365.
Montes, C., et al. (2012a), Evidence for middle Eocene and younger emergence in Central Panama: Implications for Isthmus closure,
Geol. Soc. Am. Bull., 124, 780–799.
Montes, C., G. Bayona, A. Cardona, D. M. Buchs, C. A. Silva, S. Morón, N. Hoyos, D. A. Ramírez, C. A. Jaramillo, and V. Valencia (2012b),
Arc-continent collision and orocline formation: Closing of the Central American seaway, J. Geophys. Res., 117, B04105, doi:10.1029/
2011JB008959.
Muñoz, E., A. J. Busalacchi, S. Nigam, and A. Ruiz-Barradas (2008), Winter and summer structure of the Caribbean Low-Level Jet, J. Clim.,
21, 1260–1276.
Nisancioglu, K. H., M. E. Raymo, and P. H. Stone (2003), Reorganization of Miocene deep water circulation in response to the shoaling of the
Central American Seaway, Paleoceanography, 18(1), 1006, doi:10.1029/2002PA000767.
Nof, D., and S. Van Gorder (2003), Did an open Panama Isthmus correspond to an invasion of Pacific water into the Atlantic?, J. Phys. Oceanogr.,
33, 1324–1336.
O’Dea, A., and J. B. C. Jackson (2002), Bryozoan growth mirrors contrasting seasonal regimes across the Isthmus of Panama, Palaeogeogr.
Palaeoclimatol. Palaeoecol., 185, 77–94.
O’Dea, A., and B. Okamura (2000), Intracolony variation in zooid size in cheilostome bryozoans as a new technique for investigating
palaeoseasonality, Palaeogeogr. Palaeoclimatol. Palaeoecol., 162, 319–332.
O’Dea, A., J. B. C. Jackson, H. Fortunato, J. T. Smith, L. D’Croz, K. G. Johnson, and J. A. Todd (2007), Environmental change preceded Caribbean
extinction by 2 million years, Proc. Natl. Acad. Sci. U. S. A., 104, 5501–5506.
O’Dea, A., N. Hoyos, F. Rodríguez, B. Degracia, and C. De Gracia (2012), History of upwelling in the tropical eastern Pacific and the paleogeography
of the Isthmus of Panama, Palaeogeogr. Palaeoclimatol. Palaeoecol., 348–349, 59–66.
Okamura, B., A. O’Dea, and T. Knowles (2011), Bryozoan modular growth and the retrospective analysis of environments, Mar. Ecol. Prog. Ser.,
430, 133–146.
Philander, S. G., and A. V. Fedorov (2003), Role of tropics in changing the response to Milankovich forcing some three million years ago,
Paleoceanography, 18(2), 1045, doi:10.1029/2002PA000837.
Prange, M., and M. Schulz (2004), A coastal upwelling seesaw in the Atlantic Ocean as a result of the closure of the Central American Seaway,
Geophys. Res. Lett., 31, L17207, doi:10.1029/2004GL020073.
Prange, M., S. Steph, M. Schulz, and L. D. Keigwin (2010), Inferring moisture transport across Central America: Can modern analogs of climate
variability help reconcile paleosalinity records?, Quat. Sci. Rev., 29, 1317–1321.
Ravelo, A. C., and M. W. Wara (2004), The role of the tropical oceans on global climate during a warm period and a major climate transition,
Oceanography, 17(3), 32–41.
Ravelo, A. C., D. H. Andreasen, M. Lyle, A. Olivarez Lyle, and M. W. Wara (2004), Regional climate shifts caused by gradual global cooling in the
Pliocene Epoch, Nature, 429, 263–267.
Ravelo, A. C., P. S. Dekens, and M. McCarthy (2006), Evidence for El Niño–like conditions during the Pliocene, GSA Today, 16(3), 4–11.
Saito, T. (1976), Geological significance of coiling direction in the planktonic foraminifera Pulleniatina, Geology, 4, 305–309.
Saravanan, R., and P. Chang (2000), Interaction between tropical Atlantic variability and El Niño–Southern Oscillation, J. Clim., 13, 2177–2194.
Sarnthein, M., G. Bartoli, M. Prange, A. Schmittner, B. Schneider, M. Weinelt, N. Andersen, and D. Garbe-Schönberg (2009), Mid-Pliocene shifts
in ocean overturning circulation and the onset of Quaternary-style climates, Clim. Past, 5, 269–283.
Schmittner, A., C. Appenzeller, and T. F. Stocker (2000), Enhanced Atlantic freshwater export during El Niño, Geophys. Res. Lett., 27, 1163–1166,
doi:10.1029/1999GL011048.
Schneider, B., and A. Schmittner (2006), Simulating the impact of the Panamanian seaway closure on ocean circulation, marine productivity
and nutrient cycling, Earth Planet. Sci. Lett., 246, 367–380.
Shukla, S. P., M. A. Chandler, J. Jonas, L. E. Sohl, K. Mankoff, and H. Dowsett (2009), Impact of a permanent El Niño (El Padre) and Indian Ocean
Dipole in warm Pliocene climates, Paleoceanography, 24, PA2221, doi:10.1029/2008PA001682.
Shukla, S. P., M. A. Chandler, D. Rind, L. E. Sohl, J. Jonas, and J. Lerner (2011), Teleconnections in a warmer climate: The pliocene perspective,
Clim. Dyn., 37, 1869–1887.
Steph, S., R. Tiedemann, M. Prange, J. Groeneveld, D. Nürnberg, L. Reuning, M. Schulz, and G. H. Haug (2006a), Changes in Caribbean
surface hydrography during the Pliocene shoaling of the Central American Seaway, Paleoceanography, 21, PA4221, doi:10.1029/
2004PA001092.
Steph, S., R. Tiedemann, J. Groeneveld, A. Sturm, and D. Nürnberg (2006b), Pliocene changes in tropical east Pacific upper ocean stratification:
Response to tropical gateways?, in Proc. ODP, Sci. Results, vol. 202, edited by R. Tiedemann, et al., pp. 1–51, Ocean Drilling Program,
College Station, Tex.
Steph, S., R. Tiedemann, M. Prange, J. Groeneveld, M. Schulz, A. Timmermann, D. Nürnberg, C. Rühlemann, C. Saukel, and G. H. Haug (2010),
Early Pliocene increase in thermohaline overturning: A precondition for the development of the modern equatorial Pacific cold tongue,
Paleoceanography, 25, PA2202, doi:10.1029/2008PA001645.
Teranes, J. L., D. H. Geary, and B. E. Bemis (1996), The oxygen isotope record of seasonality in Neogene bivalves from the Central American
Isthmus, in Evolution and Environment in Tropical America, edited by J. B. C. Jackson, A. F. Budd, and A. G. Coates, pp. 105–129,
Univ. of Chicago Press, Chicago, Ill.
Vizcaíno, M., S. Rupper, and J. C. H. Chiang (2010), Permanent El Niño and the onset of Northern Hemisphere glaciations: Mechanism and
comparison with other hypotheses, Paleoceanography, 25, PA2205, doi:10.1029/2009PA001733.
Wang, C. (2007), Variability of the Caribbean Low-Level Jet and its relations to climate, Clim. Dyn., 29, 411–422.

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 516
 Paleoceanography 10.1002/2013PA002515

Wara, M. W., A. C. Ravelo, and M. L. Delaney (2005), Permanent El Niño-like conditions during the Pliocene warm period, Science, 309, 758–761.
Weyl, P. K. (1968), The role of the oceans in climatic change: A theory of the ice ages, Meteorol. Monogr., 8, 37–62.
Whyte, F. S., M. A. Taylor, T. S. Stephenson, and J. D. Campbell (2008), Features of the Caribbean low level jet, Int. J. Climatol., 28, 119–128.
Xie, P.-P., and P. A. Arkin (1996), Analyses of global monthly precipitation using gauge observations, satellite estimates, and numerical
model predictions, J. Clim., 9, 840–858.
Xie, S.-P., Y. Okumura, T. Miyama, and A. Timmermann (2008), Influences of Atlantic climate change on the tropical Pacific via the Central
American Isthmus, J. Clim., 21, 3914–3928.
Zaucker, F., T. F. Stocker, and W. S. Broecker (1994), Atmospheric freshwater fluxes and their effect on the global thermohaline circulation,
J. Geophys. Res., 99, 12,443–12,457, doi:10.1029/94JC00526.
Zhang, X., et al. (2012), Changes in equatorial Pacific thermocline depth in response to Panamanian seaway closure: Insights from a multi-model
study, Earth Planet. Sci. Lett., 317–318, 76–84.

MESTAS-NUÑEZ AND MOLNAR ©2014. American Geophysical Union. All Rights Reserved. 517